2015 Archaeological Survey of Ritidian Latte Village Complex

FINAL REPORT Archaeological resource survey of 27.02 hectares at the eastern portion of the Ritidian Unit of Guam National Wildlife Refuge Prepared by Mike T. Carson, Ph.D. James M. Bayman, Ph.D. Charmaine Ledesma, B.A. Jacy Moore, M.A. Hiro Kurashina, Ph.D. Sandra Monton-Subias, Ph.D. John A. Peterson, Ph.D. Document prepared for U.S. Fish and Wildlife Service (USFWS) Micronesian Area Research Center (MARC) University of Guam (UOG) UOG Station Mangilao, Guam 96923 December 2015 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 1 ABSTRACT In an area of 27.02 hectares (66.77 acres) within the eastern portion of the Ritidian Unit of Guam National Wildlife Refuge, an archaeological resource survey documented nine distinct “Site Clusters,” each with a number of component features related to ancient use of the landscape during the period of approximately A.D. 1000 through 1700. The site components consist of the remains of stone-pillar housing structures (locally known as latte), other stonework constructions, and concentrations of artifacts and midden. An additional four “Areas of Interest” consists of natural environmental features that likely functioned as parts of the larger inhabited landscape. The site-specific findings reveal information about the ancient village structure and use of the landscape during the locally defined latte period beginning perhaps as early as A.D. 1000, followed by a series of developments and eventually the historically known encounters with Spanish colonialists during the late 1600s. The identified resources can be understood as contributing elements of this larger context, here proposed as eligible in the National Register of Historic Places (NRHP) as a Historic District. While more effort yet will be necessary for complete survey of the entire Historic District that extends outside the current survey boundary, the existing survey results can support the formulation of an eventual NRHP nomination. Moreover, the results can be situated within a larger body of information about the cultural use of the Ritidian area continually since at least 1500 B.C, thus increasing general archaeological knowledge while enhancing the potential for public appreciation, education, and interpretation. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 2 TABLE OF CONTENTS Introduction ………………………….. 4 Natural Setting ……………………….. 9 Cultural and Historical Setting ………. 12 Archaeological Setting ………………. 15 Investigative Procedures ……………... 25 Overall Survey Findings ……………... 30 Areas of Interest A, B, and C ………... 37 Area of Interest D ……………………. 39 Site Cluster 01 ……………………….. 40 Site Cluster 02 ……………………….. 55 Site Cluster 03 ……………………….. 95 Site Cluster 04 ……………………….. 101 Site Cluster 05 ……………………….. 107 Site Cluster 06 ……………………….. 140 Site Cluster 07 ……………………….. 143 Conclusions and Recommendations …. 156 References …………………………… 158 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 3 01. INTRODUCTION This document contains the technical results of an archaeological resource survey of an area of 27.02 hectares (66.77 acres) within the Ritidian Unit of Guam National Wildlife Refuge (GNWR), Territory of Guam (Figures 1, 2, and 3). The reported work fulfills contract F15PC0048, issued by U.S. Fish and Wildlife Service (USFWS) to University of Guam (UOG). The new survey followed Section 110 of the National Historic Preservation Act (NHPA), specifically with the goal of allowing USFWS to provide identification and evaluation of historic resources within its jurisdiction. All activities accorded with regulations of USFWS, including but not limited to compliance with the Archaeological Resources Protection Act (ARPA), the NHPA, and the U.S. Secretary of the Interior’s Standards for archaeological and historical work. Activities in Guam furthermore followed the laws and guidelines of the Government of Guam and of the Guam Historic Preservation Office (GHPO). The survey identified and documented nine “Site Clusters” (numbered 1 through 9) and four other “Areas of Interest” (labeled A through D), here considered as contributing elements of a Historic District that may be eligible in the National Register of Historic Places (NRHP). Documentation in this report includes detailed maps, text descriptions, scaled illustrations, photographs, analysis of surfacecollected materials, and analysis of limited test excavations. This information is presented with goals of assessing the significance of each identified resource and of the larger landscape toward archaeological research, eligibility in the NRHP, and potential for public education and appreciation. The survey during April and June 2015 covered 27.02 hectares (66.77 acres), leaving more area farther to the east side of the Refuge yet to be surveyed, where very likely more contributing elements of the proposed Historic District may yet be documented. The original plan was to survey approximately 30 hectares (74.13 acres), but field conditions after Typhoon Dolphin did not allow exploration of this full extent. In May 2015, Typhoon Dolphin (system 07W) brought sustained winds of 135 km/h (84 mph) and gusts of 171 km/h (106 mph) to the Ritidian area, resulting in widespread fallen trees, debris, and other obstacles against a formal survey effort. Additional damage was sustained from the coastal flooding and storm-surge of the extreme low-pressure system, plus 236 mm (9.3 inches) of rain within a 12-hour period. The archaeological resource survey followed the goals of the 2009 Comprehensive Conservation Plan (CCP) of GNWR (USFWS 2009), specifically regarding the issues of Cultural Resources in Appendix I of the 2009 CCP. The 2009 CCP was developed in consultation with GHPO, Department of Chamorro Affairs, University of Guam (UOG), and other interested parties as listed in the CCP documentation. Regarding the cultural resources, the 2009 CCP concluded to act on three primary objectives. 1) Implement a proactive cultural resource management program that focuses of meeting the requirements of the National Historic Preservation Act, including consultation, identification, inventory, evaluation, and protection of cultural resources. 2) Develop, in partnership with preservation partners, a program for the education and interpretation of cultural resources of the Refuge. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 4 3) Create and utilize a Memorandum of Agreement (MOA) with Guam SHPO to address inadvertent discoveries of human remains. In congruence with these stated objectives, the 2015 survey aimed to build a substantive basis for satisfying the first goal, specifically pertaining to resource identification, inventory, and evaluation. These results, as presented in this report, allow substantive basis to work on the other objectives, such as related to education and interpretation, as well as eventually to develop the management of inadvertent discoveries of human remains. The field survey was conducted in April and June 2015, with continued data analysis and input from Government review through September 2015. The principal investigator Dr. Mike T. Carson (Visiting Associate Professor at the Richard F. Taitano Micronesian Area Research Center, University of Guam) was responsible as the project leader and coordinator. He personally performed all aspects of the investigation as reported here, except as noted for test excavations at “Site Cluster 5,” follow-up crosssurvey of the northwest portion of the survey area, and prior work during 2010 at Feature A of Site Cluster 01. Figure 1. Guam and the Mariana Islands within the Asia-Pacific region. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 5 Figure 2. Approximate survey area within the Ritidian Unit of Guam National Wildlife Refuge, shown on a portion of U.S. Geological Survey (USGS) 1:24,000 Topographic Series Map, Ritidian Quadrangle. The boundary of the Refuge has been updated here. At “Site Cluster 5,” Dr. James M. Bayman (Professor of Anthropology, University of Hawaii) was responsible for test excavations and data analysis, with input from Dr. Hiro Kurashina (Emeritus Director of the Richard F. Taitano Micronesian Area Research Center, University of Guam), Dr. Sandra Monton Subias (Research Professor of Humanities, Universitat Pampeu Fabra in Barcelona), and Dr. John A. Peterson (Assistant Vice President of Graduate Students and Special Research Programs, University of Guam). Under Dr. Bayman’s direct supervision, some of the field and laboratory tasks were performed by Ms. Charmaine Ledesma (graduate student, Department of Anthropology, University of Hawaii) and Ms. Jacy Moore (M.A. recipient, Leiden University). Other student participation was supervised closely as part of a field training program at University of Guam (students Enea Bellelis, Alden Cabero, Logan Oplinger, Gia Righetti) and University of Hawaii (students Jason Chinundet, Hattie Gerrish, Karen Honda, Isaiah LeDonne, Marlene McVey, Kristen Sakai, Melissa Smith, and Caylee Stein). In the northwest portion of the 2015 survey area, a repeated cross-survey was conducted in order to check for the possibility of caves in the limestone terrain, as Dr. Hiro Kurashina recalled a visit to a cave containing water but no archaeological materials in this area in 1987. Although the area already had been surveyed thoroughly for archaeological sites earlier in June 2015, it was cross-surveyed in new samples of transects later in 2015 by Dr. Mike T. Carson with Dr. Hiro Kurashina and Ms. Jacy Moore, specifically for the possibility of identifying caves, rockshelters, or other cavities. No such features were discovered to contain archaeological materials or pools of water, although the possibility still exists that the remembered water cave’s entrance may have become obscured by rock-fall or other debris over time. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 6 Figure 3. Approximate survey area within the Ritidian Unit of Guam National Wildlife Refuge, shown on a portion of 2005 Quickbird Satellite imagery of Guam in semi-transparency overlay with base of 2008 LIDAR dataset imagery. Prior surface documentation and test excavation had been conducted during 2010 at one archaeological feature, here designated as Feature A of Site Cluster 01. The new work in 2015 involved fully detailed recording, fuller coverage of the surrounding site resources, additional materials collection, more test excavation, and comprehensive analysis. The 2010 study was part of an archaeological field training program at University of Guam, supervised by Dr. Mike T. Carson and involving several students including Vanessa Cabrera, Dee Cruz, Tonilynn Dotts, Asa Gonzales-Peterson, Bradley Halbert, Chris Halligan, Alex Laungowa, John Mercado, Jacy Moore, Nina Peck, Alyssa Posadas, Sandy Sidwell, and Brett Storie. All aspects of this project were conducted in partnership with USFWS, and several key personnel must be acknowledged for their contributions. GNWR Manager Mr. Joseph Schwagerl and Administrative Officer Ms. Latasha Hawk developed the Government Contract F15PC0048 for the 2015 survey, and their ongoing support contributed to the success of the project. In June 2015, Ms. Laura Beauregard began duties as Refuge Manager and continued the high level of professional support and guidance for the completion of the 2015 survey. Throughout all activities, Law Enforcement Officer Mr. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 7 Gabriel Cruz ensured the safety of the 2015 survey personnel. Maintenance Technician Mr. Brian Leon Guerrero worked closely with the survey crew for identifying site areas, maintaining clear and safe access, and clearing invasive vegetation within USFWS guidelines. Meanwhile, several other GNWR staff provided guidance about management of cultural resources, native plants and wildlife, and general concerns of the project, including Park Ranger Ms. Emily Sablan, Biologist Ms. Jennifer Cruce, and Biologist Ms. Lauren Gutierrez. The 2015 survey was made possible by several days of post-typhoon clean-up labor by USFWS staff, following the extensive damage of Typhoon Dolphin (system 07W) in May 2015. The professional crew worked extended hours for several consecutive days in May through June 2015, clearing roads and trails, cutting and removing fallen vegetation, clearing other debris, and ensuring safe access to the 2015 survey area. The crew members included Maintenance Technician Mr. Brian Leon Guerrero of GNWR, Lead Forestry Technician Mr. Thaddeus Niebel of Malheur National Wildlife Refuge, Fire Engine Captain Mr. Collin McCabe of Mid-Columbia River National Wildlife Refuge, Fire Engine Captain Mr. Mark Rakeshaw of Sacramento National Wildlife Refuge, and Forestry Technician Mr. Keith Mayer of San Luis National Wildlife Refuge. This technical report contains information about the project context, investigative procedures, findings at each identified feature, interpretations of research significance, and recommendations of resource management. The report details are presented in a standardized format for ease of finding specific types of information and potentially making cross-comparisons. While this report satisfies the need for a record of the primary data, additional efforts will continue with academic publications and public outreach in coordination with USFWS. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 8 02. NATURAL SETTING The survey area is situated in extensive limestone plateau terracing, seaward of a steep limestone cliff, bordered by scattered patches of sandy beach along the shoreline (Figures 4, 5, and 6). In this northernmost point of Guam, the area known as Ritidian or Litekyan occupies part of the widespread limestone terrain that distinguishes the north portion of the island from the volcanic terrain of the south portion of the island. The northern limestone terrain appears to be composed of a series of karstic plateaus of Pliocene or Pleistocene origin, wherein each plateau or terrace represents ancient reef formations exposed by lowering sea level or raised tectonically above sea level (Randall 1990:41–42). Calcareous sand accumulated along the seaward edge of the raised limestone mass, gradually forming a wide coastal plain. At the exposed northern tip of the island, the rate of calcareous sand accumulation exceeds that at sheltered embayments and other protected locales. The high-energy beachfront at Ritidian is notably subjected to mass deposition of calcareous sand and marine detritus during periodic storm surges, so that ongoing gradual shoreline progradation may be punctuated by occasional episodes of rapid sand accumulation. Figure 4. Overview of the Ritidian area in June 2015, showing the shoreline and coastal plain in relation to the landward limestone formations. This view is to the north, across the Rota Channel, with the 2015 survey area to the right-hand (east) side of the image. Figure 5. View of the interior limestone cliff at Ritidian in June 2015, looking landward the south, with the 2015 survey area to the left-hand (east) side of the image. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 9 Figure 6. View of the eastern end of the 2015 survey area at Ritidian, showing one patch of sandy coastal plain in relation to the interior limestone formations. This view is to the east, near “Site Cluster 07.” For the portions of Ritidian consisting of coastal plain landforms, archaeological deposits may have been emplaced in the dynamic environment of prograding sandy beaches. Some portions of sandy beach contain no archaeological deposits at all, whereas others retain records of layered occupations. Especially in seaward areas, lenses or layers of storm-surge deposits may have interrupted cultural occupation layers. These interruptions vary in thickness, up to 1 m. Within these sandy beach units, the oldest cultural deposits have been found more than 2.5 m deep, as much as 3500 years old. In the western side of the Refuge, outside the 2015 survey area, archaeological layers in an extensive sandy coastal plain landform span the full range of the last 3500 years, in stratigraphic order of their deposition. In the scattered pockets of sandy beaches in the eastern side of the Refuge and in the 2015 survey area, the potential appears much more limited for finding buried archaeological layers. For the portions of Ritidian in the elevated limestone terrain, as in most of the 2015 survey area, archaeological deposits were emplaced in a low-energy accumulation of clay and silt sedimentary units. These natural sedimentary units generally were thinner than 50 cm over the basal limestone bedrock, and most of the archaeological components appear in surface-visible or near-surface contexts. In these kinds of shallow and rocky silts and clays, the surviving archaeological materials tend to date mostly within the last 1000 years and in rare cases as much as 2000 years old. Nonetheless, possibilities exist for even older archaeological deposits in places with suitable preservation conditions. Caves, rockshelters, and other cavities are abundant in the limestone formations. Some but not all of these natural geological features contain archaeological resources, such as cultural deposits and ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 10 rock art panels. In these cases, the cultural deposits can be highly variable in terms of depth, contents, and age. Rock art panels similarly are varied in terms of their composition, placement, and possibly age. The present vegetation at Ritidian includes a number of historic introductions, invasive species, and secondary-growth taxa, but efforts are underway to revitalize native flora and habitat. In the year 1819, de Freycinet (2003:248) described Ritidian as one of two places in Guam (the other is noted as Orote) with “the finest building timber” in the form of trees that he identified as “agau, daok, djoga, dugdug, gago, hadju-lago, kamatchile, nonak, ufa, unik, putting, rima, and talissai.” Plant growth is rapid at Ritidian, as expected in a humid tropical setting, but root systems are weak in the calcareous sandy deposits and in the shallow silts and clays overlaying limestone bedrock. Root systems are mostly within 20 cm of the ground surface, and trees are overturned or removed during powerful storms and typhoons. Reports of observations in January 1988 indicated that “many large trees intact with root systems had been toppled” after a typhoon (Randall 1990:43). Similar observations were made in January 1998 shortly after another typhoon had affected the Ritidian area. Most recently, severe damage was widespread throughout Ritidian after Typhoon Dolphin (system 07W) in May 2015. Present-day animal life at Ritidian most visibly includes coconut crabs, monitor lizards, sea turtles, pigs, deer, and brown tree snakes. The latter three animals were introduced after European contact in the Mariana Islands, and they pose threats to the native habitat. Other animals live in the area but in less visible presence. Native animals, like the native plants, are under the management of the USFWS. Populations of birds and fruit bats have declined dramatically at Ritidian as throughout Guam, due to predatory snakes and degradation of habitat. Bones of these animals may be expected in some archaeological deposits, offering clues about their population history. Diverse and abundant marine life presents a reliable resource base for subsistence and other purposes. Marine food remains are expected to be abundant in archaeological deposits at Ritidian, as at nearly any coastal area. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 11 03. CULTURAL AND HISTORICAL SETTING The place-name of Ritidian most likely was Litekyan or another similar word prior to intensive Spanish influence in the late A.D. 1600s. Literal and figurative meanings of the place-name are unclear at present, but the name most likely relates to a “place of stirring.” Today, the name refers to a former village and the general locale at the north end of Guam. A former village at Ritidian or Litekyan is evidenced by substantial pre-Spanish artifacts and by historical references to a village in the 1670s. A map produced by Alonso Lopez in 1671 indicates “Ritidyan” as a village with a Christian church. In an overview of Ritidian ethnohistory, Stephenson (1990:18) summarized the very few available written documents about Ritidian as a native Chamorro village: Although apparently an important Chamorro village prior to Spanish times, reference research shows that the Ritidian site was abandoned early in the Spanish era …, most likely in 1682. Furthermore, knowledge of village life at Ritidian cannot be gathered … . During Spanish efforts to control native Chamorro populations in the late 1600s, Ritidian and many other villages were abandoned. According to Fritz (1904:31; translation in Fritz 1984:9), Quiroga (former Spanish governor of Guam) in 1680 “established six parish villages, besides Hagatnia: Pago, Juapsan, Juarahan, Merizo, Humatag and Agat. He destroyed all other settlements and forced the natives to settle around the churches.” Resistance to Quiroga’s relocation campaign was punished severely. The campaign was more or less complete by 1683 for Guam, yet it was re-established in 1694 through 1698 to complete Quiroga’s efforts in other islands. Rogers (1995:67–68) notes that the church at Ritidian was operating in 1681, when a priest attempted to flee to Rota but was caught and killed. Presumably referring to the same event, Fritz (1904:32; translation in 1984:10) reported that a priest named Angelis was killed at Ritidian, but the date appears to have been 1683. This event may have been viewed by Quiroga and others as an example of how the Spanish forces could not survive in numerous dispersed outposts without substantial protection. In response, the native Chamorro populations were removed from Ritidian and other places, re-located in fewer villages for more effective foreign control. Very few early Spanish-era documents refer to the Ritidian area, and these references are constrained to sparse notes within a very short period of the 1672 through 1676: 1) In August 1672, a group of Christian missionaries visited a village called “Retian,” but no further details are provided (Lévesque 1995:143). 2) The village of “Retiyan” is mentioned without elaboration in 1674 (Lévesque 1995:221). 3) In the Jesuit annual report for 1674 to 1675 (Lévesque 1995:198-299), Alonso Lopez noted the construction of a church at “Ritidyan,” dedicated to Francis Xavier. 4) In the same annual report for 1674 to 1675 (Lévesque 1995:303), Lopez noted: “About 400 attending mass at Ritidian church of San Xavier, as compared to 500 at ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 12 San Ignacio in Agana, 300 at Santa Rosa in Tupungan, 700 at San José in Upi, and more than 500 at San Miguel in Tarague.” These numbers suggest that Ritidian supported a population equivalent to other large villages in the 1670s. 5) Again in the same annual report (Lévesque 1995:303), Lopez noted: “The Father in charge of San Xavier de Ritidian was so well disposed that he undertook to build a college for children (dedicated to the glorious Archangel St. Gabriel) where twice a day 22 children meet … .” 6) In a later document dated 1676 (Lévesque 1995:357), “Rittian” is mentioned as the location of the “residence of St. Francis Xavier” as well as “a house or college dedicated to the glorious St. Sabina where the girls of that residence were collected for their education …” 7) The same 1676 document (Lévesque 1995:357-358) describes the death of Brother Díaz at “the residence of San [Francisco] Xavier of Rittian.” According to this account, Brother Díaz was killed after an incident wherein a group of young men invaded the girls’ college at night. The next day, Brother Díaz confronted the group, and a rapidly escalating conflict ended in his death. Working with the above documents and also including oral accounts and personal observations, le Gobien (1700:86–89) described the villages of Ritidian and Tarague as maintaining an ongoing rivalry in the 1670s. The origin of the rivalry is unknown. The nature of the rivalry is not specified, but it was not necessarily violent. Working further with the historical documents, Jalandoni (2011, 2012, 2014) has addressed two problems in the apparent modern perceptions of the Spanish-era presence at Ritidian. First, the historical documents refer to a building or set of buildings operated by Jesuit missionaries, contrary to the apparently popular present-day reference to a casa real (literally “royal house” or implying a “government building”). Second, the Jesuit church was built and re-built a number of times at Ritidian, as well as at adjacent areas that may or may not have been part of the traditional village of Ritidian, despite the conflation of all of these areas as if they represented a unified settlement. Sustained residential occupation did not occur at Ritidian from the 1680s until perhaps the early 1900s, but other kinds of land-use occurred in the context of periodic resource collection and attempts of agricultural developments. Like many areas outside village settlements, Ritidian at that time may have been productive for various natural resources, such as forest trees and plants, birds and fruit bats, and marine life. For instance, in 1819, de Freycinet (2003:248) described Ritidian as supporting notably abundant tree growth suitable for construction timber. During the early through middle 1800s, Ritidian was one of a number of places in Guam that was used for Spanish-operated agricultural development. According to the historical documents of that period, Madrid (2014) has concluded that large areas of Ritidian had been converted into agricultural fields, primarily in the area that today is the western side of the Ritidian Unit of Guam National Wildlife ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 13 Refuge. In this western portion of the Refuge, several large crater-shaped water-well features and scattered stacked-rock features very likely relate to the early-middle 1800s patterns of land use. Under U.S. rule from 1898 through 1941, Guam was effectively a U.S. Naval station, but Ritidian and other areas were prized by Japanese and American business entities for coconut plantations and other enterprises. Today, remnants of a coconut plantation are evident in a portion of the westernside coastal plain at Ritidian, but records have not been found about its creation and operation. By comparison, however, similar plantation areas were established at Tarague in the first decades of the 1900s (Liston 1996:47). For practical purposes, some of these plantations likely were constructed in places of former Spanish-operated agricultural developments, where the land already had been cleared and in some cases with water-wells already installed. During the Japanese occupation of Guam in 1941 through 1945, no WW II military engagements are known to have occurred specifically at Ritidian. However, at least some Japanese soldiers probably were in the Ritidian area at one time. Archaeological test excavations recovered “the barrel and firing mechanism of a Japanese type 99 short rifle” (Kurashina 1990:98). After the re-establishment of U.S. rule in 1945, Ritidian was a very low priority during efforts to build a new infrastructure for Guam. Naval facilities were commissioned at Ritidian in 1968, and the Ritidian Unit of Guam National Wildlife Refuge (GNWR) was created in 1993. Buildings and remnants of structures from this era still are surviving today, and two of those buildings continue to be used. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 14 04. ARCHAEOLOGICAL SETTING The archaeological record at Ritidian, as in the Mariana Islands overall, extends at least as early as 1500 B.C. During this time span of at least 3500 years, several spatial patterns and chronological trends are evident in site size and distribution, pottery forms, stone and shell artifacts, food remains, housing structures, human-environment relations, and other factors. The archaeological resources at Ritidian relate to this complete scope of information, represented in surface-visible ruins and artifact concentrations, subsurface archaeological layers, and caves with cultural deposits and rock art panels. The first reported archaeological study at Ritidian was part of an archipelago-wide effort by Hans Hornbostel in the 1920s to evaluate the potential for archaeology in the region. This pioneering work involved cursory surface inspection and expedient test excavations throughout the region. By today’s standards, the results are useful as a reconnaissance, aimed at documenting the vicinities of major concentrations of surface-visible archaeological ruins and monuments. Hornbostel (1925) was concerned primarily with latte sites as the material remains of the last time when a native Chamorro cultural system operated prior to the historically known period of Spanish conquest. Latte consist of stone pillars (haligi) with corresponding capital stones (tasa) that once had supported wooden house structures (Figure 7), as has been confirmed in historical documents, ascertained through evaluation of architectural engineering characteristics, and found to be consistent with the associated archaeological deposits, as reviewed by Laguana et al. (2012) and Thompson (1940). Among the numerous references concerning latte, an updated general summary is available (Carson 2012a), and studies of the latte at Ritidian specifically have generated new substantive and theoretical knowledge within the last few years (Bayman et al. 2012a, 2012b; Carson 2014a). Figure 7. Architectural rendering of a latte structure, composed by John Aguon and originally published by Laguana et al. (2012). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 15 Figure 8. Latte sites of Guam, documented by Hans Hornbostel (1925) and redrawn for clarity by Erik Reed (1952). Copy of a map at the Richard F. Taitano Micronesian Area Research Center, University of Guam. According to Hornbostel’s (1925) island-wide survey of Guam, Ritidian was one of the many places in Guam bearing evidence of a former latte village complex, and in fact it was illustrated as an area of “dense latte” (Figure 8). The entire northern tip of Guam, including Ritidian and adjacent areas, was depicted as containing more or less continuous latte-related archaeological materials (Figure 9). On the surface of the sandy coastal plain near the present-day Nature Center at the Ritidian Unit of GNWR, Hornbostel (1925) recorded two prominent surface-visible archaeological ruins (Figure 10). Near the shoreline, the remains of a Spanish “block house” were noted. Farther landward (southward) toward the base of the limestone cliff, ruins were visible of a large set of latte remains, especially noting one prominent pillar (haligi or halege) about 1.2 m in height. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 16 Figure 9. Areas of latte-related archaeological remains in northern Guam, recorded by Hornbostel (1925) and redrawn by Thompson (1925). Figure 10. Prominent archaeological remains at Ritidian, documented by Hornbostel (1925), with clarification of measurements added by Kurashina (1990). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 17 Working with Hornbostel’s unpublished notes and collections, Thompson (1932:20) described the first known archaeological excavation at Ritidian: Excavation below the surface of the floor of a cave at Ritidian, Guam, revealed the following stratification: a layer of bat manure 1.5 feet thick superimposed upon a 1-foot layer of fragments from the walls of the cave; under these fragments, a 1.5-foot stratum of sand containing scattered human bones, potsherds, stone and shell implements, and fragments of charcoal. No drawings were found in this cave. Hornbostel’s (1925) original notes indicate that the cave was “in the wall of the second step of the limestone cliff,” accessible by a climbing a “perpendicular wall of some forty feet,” and containing an internal space up to “forty feet” in height. At least two caves could fit this description, and both are found outside the 2015 survey area. In any case, the prior results can suggest that archaeological deposits are likely present and perhaps buried beneath more recent debris inside caves of the limestone terrain. Shortly after WW II, Obsorne (1947a, 1947b) surveyed archaeological sites of Guam, aiming for an island-wide synthesis and update, including Ritidian as just one minor component in a much larger undertaking. This work entailed surface reconnaissance of much of the island. Limited subsurface testing was undertaken in selected locations, but no excavation was reported at Ritidian. Of relevance to Ritidian, Osborne (1947a:47) noted: From Oruno around the whole northern to northeastern coast on the island there is a continual archaeological area. It has been generally disturbed by recent copra plantations but remnants of sufficient size from archaeologic examination remain. At Ritidian are several small latte groups in a poor state of preservation. They are generally so covered by the prolific limonchina that no more than location could be done. Here too is the Casa Real ruin of the Jesuit mission which was established there in the early years of evangelization. During the 1950s, archaeologists in many places around the world became increasingly involved in excavating deep beneath the ground surface, exploring the potential to learn about ancient cultural layers pre-dating the surface-visible ruins and monuments such as the latte sites of the Mariana Islands. Among this world-wide trend, Alexander Spoehr (1957) promoted stratigraphic excavations, chronological sequencing, and radiocarbon dating in the Mariana Islands, where he demonstrated a long cultural sequence of at least 3000 years and by far pre-dating the latte-associated materials of the last 1000 years. Spoehr conducted his research in the islands of Rota, Tinian, and Saipan with clear implications for the archaeologist who were active at that time in Guam. Later efforts to compile standard site recording for Guam largely followed the pre-existing baseline data of surface-visible latte ruins, for example as reported in the Ritidian area. Surface-guided surveys by Erik Reed (1952) and later by Fred Reinman (1968, 1974, 1977) concentrated on documenting latte sites, in some cases augmented by excavations of cultural deposits associated with the latte. According to these island-wide surveys of Guam, the Ritidian area consistently was described as: ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 18 a) containing substantial prehistoric artifacts on the surface; b) possibly containing disturbed latte remains that could not be re-identified; and c) possibly containing disturbed remains of the Spanish church or other structure no longer in evidence. According to the standardized site records for Guam (Reinman 1974, 1977), the Ritidian area coincides roughly with Sites 66-08-0012 and -0013. The sites appear to be defined as two large areas where prehistoric artifacts are visible on the surface, almost certainly relating to a village settlement just prior to abandonment in the late 1600s. Within these two generalized site areas, several specific portions may be discerned as containing variable density, type, and age of archaeological materials. In other words, the site records from the 1970s surely can be updated with refining the boundaries, definitions, and potential significance of the archaeological resources. The current survey area corresponds roughly with Fred Reinman’s 1965–66 survey notes of the area of “Ritidian Point to Pajon Point” in his unpublished report to the U.S. National Science Foundation, quoted here in its entirety (Reinman n.d.: 8–9): From Ritidian Point to the central portion of Tarague Beach the coast swings around Ritidian Point and runs nearly in a northwest/southeast direction. On this easterly side of Ritidian construction has removed most of any evidence of occupation for approximately the first 1000 meters. Beyond this area the occupation remains again become fairly heavy. From a point about half way between Ritidian Point and Pajon Point a heavy occupation area exists from the 10 meter contour line all the way back to the cliffs which begin about 250 meters from the beach. Eight sets of latte still survive in this area nearly all of which are 8 stone sets. Most of these are located between 25 and 100 meters back from the beach; two, however, were found only about 75 meters from the cliffs. Several of the latte structures located here also had low stone walls on the down hill (seaward) side of the late as well as a smaller, 4 stone structure adjacent to the larger structure. The uprights of the main latte are relatively flat in cross section with a truncated triangular outline. Pottery, broken tools, and other parts of the now badly broken or destroyed latte are also to be found throughout the area. Coconut, pandanus, and papaya are the principal trees in the area. The latte as described here would appear to match with those as documented in this report, although the current documentation provides somewhat different details. Excavation-focused archaeological research at Ritidian began to gain momentum with a project by Hiro Kurashina (1990). This work involved several test pits and monitoring of construction work trenching, as part of the closure of the U.S. Navy operations of NAVFAC Ritidian Point, prior to the eventual transfer of the property to the current management of USFWS. In this setting, the archaeological investigation necessarily concentrated on the footprint of immediate impact in the vicinity of the government office buildings and parking lots. The archaeological study by Kurashina (1990) confirmed widespread archaeological deposits in the sandy coastal plain landform. These deposits contained remnants of habitation layers, variable ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 19 concentrations of artifacts and midden, possible traces of house structures or other features, and a number of burial features. Radiocarbon dating verified human habitation perhaps as old as A.D. 1000, consistent with the earliest date range of the latte period in the Mariana Islands. Since the 1990s, several archaeological projects were conducted within the Ritidian Unit of GNWR, necessarily in compliance with U.G. Government regulations and in consultation with the Guam HPO. 1) First was a study by Boyd Dixon (2000), eastward of the current government office buildings, where excavations revealed signs of mechanical ground-clearing and grounddisturbance. 2) Second was an emergency recovery of human remains from a locality in the western portion of the Refuge property, apparently disturbed from a latte period cultural deposit (Carson 2005). 3) Third was archaeological monitoring during the rehabilitation of a septic system, leading to documentation of human burial features of the latte period, found in a portion of the parking lot near the eastern government office building (Walth 2014). 4) Fourth was a study of the parking lot areas, prior to new construction improvements, where scattered remnants of latte period cultural deposits were documented (Carson 2014b). 5) Fifth was a series of surface reconnaissance surveys, mostly in the eastern portion of the Ritidian Unit of GNWR, for the purpose of identifying archaeological resources prior to decision-making about constructing a possible military firing-range in the vicinity, leading to the documentation of several artifact concentrations, latte ruins, and caves or rockshelters with apparent cultural deposits, although these detailed findings are not yet available for public access (Dixon et al. 2011; Dixon, Jalandoni et al. 2013). Research-driven archaeological investigations have been ongoing at Ritidian since 2005, in partnership with USFWS and in consultation with Guam HPO. 1) Carson (2012b, 2014c, 2014d) documented a 3500-year sequence of the evolving natural and cultural landscape, based on extensive survey and subsurface explorations in the western portion of the Refuge. 2) Peterson (2014) reported varied studies of the traditional cultural use of the area, potential for studies of the Spanish Jesuit period, and ability to conserve the rock art inside caves. 3) Jalandoni (2011, 2012, 2014) further explored the potential of examining the Spanish Jesuit period, specifically through excavations at the locality of the expected missionary settlement of the late 1600s. 4) Bayman et al. (2012a, 2012b) excavated large blocks of two adjacent latte sets in the western area of the Refuge, concluding that each latte likely was associated with a specific set of past activities reflecting the economic and social organization of the 1600s. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 20 5) Carson (2014a) began exploring the potential for studying the remnants of the latte period landscape still visible on the surface of the eastern portion of the Refuge, also recognizing the possibility of evidence pertaining to older periods. The archaeological findings at Ritidian further have been significant toward larger research, such as: a) reliable radiocarbon dating of certain types of marine shells found in archaeological deposits (Carson 2010); b) proving similar reliability of radiocarbon dating of algal bioclasts, often associated with coastal archaeological sites (Carson and Peterson 2012); c) understanding the relationships of ancient cultural deposits within the natural geological formation processes of solidified calcrete found in many beach sites (Carson and Peterson 2011); d) placing ancient archaeological layers within a long-term chronological sequence of landscape evolution at an island-wide scale (Carson 2011, 2014e); e) identifying where to search for the most ancient cultural habitation layers in the Marianas region as a whole (Carson and Hung 2015); f) situating the archaeological materials of the Mariana Islands within a larger Asia-Pacific context, in fact verifying Ritidian and other sites of the Marianas as the oldest known of any human settlement in the entire Remote Oceanic region (Carson 2013, 2014f, 2014g; Carson and Kurashina 2012; Carson et al. 2013; Hung et al. 2011); and g) illustrating the potential for studies of archaeological landscapes as heritage resources suitable for public access and appreciation, using Ritidian as an example (Carson 2014d). The most recent synthesis of archaeological research at Ritidian reviews the findings of archaeological deposits of each time period in a 3500-year-long sequence, coordinated with the chronological change sea level and coastal geomorphology (Carson 2014c). This kind of synthesis has been possible through detailed information about local and regional sea-level history, precise and reliable radiocarbon dating of archaeological and natural layers, and an extensive program of surface and subsurface exploration. The new proposed survey work very likely will enable a stronger synthesis of the Ritidian area at an even larger scale than so far has been possible. At present, the western portion of Ritidian has been studied much more intensively than the eastern portion. The western portion of Ritidian may be described as containing nearly continuous archaeological remains of surface-visible ruins and subsurface deposits, spanning 3500 years of human presence, distributed over most of the coastal plain landform and inside caves of the adjoining limestone cliff face (Figure 11). In the eastern portion of Ritidian, as well as in the upper limestone plateau landforms, the archaeological remains appear to occur in several clusters of varying size and density of mostly surface-visible components of the latte period, although more details can be learned through the current 2015 survey. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 21 Figure 11. Distribution of archaeological test pits and key resource elements in the coastal zone of the western portion of the Ritidian Unit of Guam National Wildlife Refuge, modified from Carson (2014c). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 22 Figure 12. Schematic profile of landform chronology at Ritidian, reprinted from Carson (2014c). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 23 Figure 13. Approximate zones of archaeological resources at Ritidian, reprinted from Carson et al. (2014). Based on the prior findings in the western portion of Ritidian, a chronological sequence can be proposed for illustrating how the landscape changed over time and situating archaeological deposits within this framework (Figure 12). This framework was based on the known locations, depths, and ages of archaeological findings within a sequence of natural stratigraphic layers, coordinated with change in sea level and coastal morphology over the last several thousands of years, as reported in detail elsewhere (Carson 2011, 2012b, 2014c, 2014e). Although a full 3500-year-long sequence is evident when considering the Ritidian area as a whole, the specific findings in any single location may relate to a limited scope of the larger chronology. Issues have not yet been resolved about how to designate the archaeological site boundaries, contents, ages, and statements of significance at Ritidian. While a more or less continuous area of archaeological resources has been documented in the coastal zones of the western portion of the Ritidian Unit of GNWR, the level of survey coverage so far is not yet sufficient to justify specific site boundaries in the eastern portion of the Ritidian Unit or in the upper limestone plateau landforms. The 2015 survey now offers new information toward resolving some of these issues. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 24 As reviewed by Carson et al. (2014), Ritidian overall includes a set of zones that each appear to have a different potential for containing archaeological resources (Figure 13). Some zones are expected to contain little or no archaeological materials, whereas others are expected to contain variable densities of surface-visible and subsurface archaeological components. The newest survey effort during 2015 was guided partially by this framework, although naturally the model needed to be refined according to the new discoveries of field research. Specifically in the 2015 survey area at Ritidian, the associated landforms appear to consist mostly of limestone plateau formations, unlike the sandy coastal plain of the western portion of Ritidian. In the areas of limestone formations, the potential generally is considered low for finding archaeological deposits preserved from more than 1000 years ago, and instead the preserved materials most likely relate to the latte period or more recent contexts. Nonetheless, older cultural deposits may be preserved inside caves or rockshelters, as well as in pockets of unexpectedly deep sedimentary units. In the locations of known archaeological resources, new efforts may concentrate on refining scientific knowledge of the material contents, associated cultural contexts, and time periods. These results can be significant toward building programs of appreciation, education, and interpretation. These efforts can be pursued immediately for many of the localities of known resources in the coastal zone of the western portion of Ritidian, while similar efforts in other parts of the Refuge will require more development in accordance with the new and future survey findings. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 25 05. INVESTIGATIVE PROCEDURES All of the survey tasks were coordinated with USFWS. Permission was requested from the Refuge Manager prior to any activities. Opportunities were available at all times for involvement of USFWS staff and for members of the public. Prior to the field survey activity, the survey personnel consulted with USFWS and staff at GNWR, for three purposes: 1) verifying the specific locations of survey activities with an agreeable schedule of days and hours; 2) clarifying the property boundaries of the Ritidian Unit of GNWR; and 3) ensuring that all personnel were fully aware of the special rules that apply during their activities, for example regarding safety, use of vehicles and equipment, and treatment of all natural and cultural resources. The general area of the 2015 survey was identified in November 2014, when USFWS employee Mr. Brian Leon Guerrero and Dr. Mike T. Carson observed several latte sets in the heavily forested zone of limestone terrain in the eastern side of the Refuge. Based on non-intrusive visual observations only, the surrounding area was found to contain the remains of an ancient village complex that had not yet been reported in detail in any known archaeological survey report. After consultation with USFWS and a December 2014 field visit by personnel of the Guam Historic Preservation Office (Figure 14), a survey contract was developed by USFWS in order to accomplish the resource survey as reported here. Figure 14. Field visit at the Ritidian Unit of GNWR, dated 22 December 2014, by personnel of the Guam Historic Preservation Office, including staff archaeologist Mr. John Mark Joseph (left) and technical consultant Mr. Richard K. Olmo (right), with GNWR Refuge Manger Mr. Joseph Schwagerl (center). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 26 The government contract was developed as Contract F15PC0048, issued by U.S. Fish and Wildlife Service (USFWS) to University of Guam (UOG). The scope of work called for an archaeological resource survey as described in this report, conferring all permissions such as ARPA permit and other considerations for professional archaeological work in support of Section 110 of the NHPA. Following the official award of the contract on 11 March 2015, a fully detailed work plan was provided to the Guam Historic Preservation Office (GHPO) for review on 16 March 2015. Although post-dating the requested 30-day review period, GHPO provided comments on 12 May 2015 in support of proceeding with the survey as planned. The new survey area was drawn approximately on a map (see Figures 2 and 3) to encompass a boundary substantially larger than the archaeological features that were observed in November 2014. The proposed survey area included the location of a latte set that had been studied and reported previously in 2010 as part of archaeological field training at University of Guam (Carson 2014a), although new survey effort would be necessary for situating the prior findings within the context of the larger site complex. Other surveys may have overlapped with the same area, although site-specific details are not yet disclosed (Dixon et al. 2011; Dixon, Jalandoni et al. 2013). The new survey effort began by walking over the area in transects separated by no more than 5 m. This close spacing allowed for reasonable visibility of the ground surface. This initial transecting entailed identification of the localities with surface-visible archaeological elements, as well as those locations where the setting suggested the possibility of subsurface archaeological deposits. During this initial effort, the locations of archaeological resource potential were marked by flagging tape for later reidentification and detailed documentation, and they were recorded by a survey-grade (sub-meter accuracy) global positioning system (GPS). Following the initial resource identification, vegetation-clearing was necessary for a reasonable view and documentation of the vicinity of findings. All vegetation-clearing was conducted in accordance with USFWS regulations, especially for protecting any endangered or potentially endangered species. Limited vegetation-clearing was accomplished in April 2015, but a major labor investment was required after the extensive tree-falling and other debris of Typhoon Dolphin in May 2015. In June 2015, resource documentation proceeded with the complete field-walking transects, mapping, photographs, and text recording. Scaled plan view maps showed the salient characteristics of individual archaeological features, their material contents, spatial distribution, and density. Photography conveyed more of these characteristics and documented the actual field conditions. Text recording documented the spatial dimensions, individual feature component materials, associated cultural context, associated time period, and general information that ordinarily would be required in a nomination for the U.S. National Register of Historic Places (NRHP) and the Guam Historic Property Inventory (GHPI). The above-described procedures of documentation were sufficient for sites with surface-visible components, such as latte ruins, variable concentrations of artifacts and midden, and caves containing rock art or other signs of past cultural activity. However, the potential for subsurface materials still needed to be assessed in these locations, as well as in other locations that conceivably could contain buried archaeological materials. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 27 In places where artifacts and midden were evident on the present ground surface, these instances were mapped, photographed, and described as in the above-noted procedures, augmented by limited sample collections. Material collections and subsequent data analysis followed the procedures as outlined below. In those locations where subsurface archaeological components were expected, small test units were excavated for the purpose of collecting baseline survey data. Each such unit was measured in cm of plan view dimensions, no more than 100 cm in maximum length of any single unit, thus sampling only small portions of the larger site resources. Excavations proceeded in artificial 10-cm levels within natural stratigraphic layers. All contents were sieved through fine (1/8-inch or 3.175-mm) wire mesh to ensure maximum recovery of archaeological materials. Layers were recorded by scaled profile illustrations and photographed, and they were described in terms of depth, thickness, natural and cultural contents, matrix, color, texture, boundary, and other characteristics. In a single event of encountering human remains during text excavation at “Site Cluster 5,” the specific findings were documented as fully as possible without causing any unnecessary disturbance. According to Guam Executive Order 89-24: “Scientific, medical or other study shall be restricted to the minimum necessary to ascertain cultural or ethnic associations, and to address significant research questions.” This law was upheld, along with the guidelines of the Guam Historic Resources Division regarding the treatment of human remains. The location was recorded by survey-grade GPS and indicated on a scaled map with reference to key geographic points in the surroundings. The discovered skeletal elements were documented in terms of their location, stratigraphic association, number of elements, percentage of visible portions of those elements, and any discernable anatomical parts. The location was secured and carefully covered by the natural sediment that had been excavated. USFWS staff were notified of the specific location for clarity of management. Further considerations are outside the scope of this survey report, but the reported information will be important toward consultations between USFWS and Guam HPO about long-term management of burial features. All mapping data were incorporated in a single geographic information system (GIS). The data were encoded in Universal Transverse Mercator (UTM) Zone 55 North, using the World Geodetic Survey (WGS) datum of 1984. The data are available to USFWS. The recovered artifacts, food remains, and charcoal were washed and sorted by material type. Field catalog numbers and provenience information were maintained for each separated material group in clearly labeled bags. The procedures of analysis were the same for excavated materials as for surfacediscovered items. Pottery fragments were counted and weighed, and representative pieces were photographed or illustrated. Additional observations, although not always reported here, allowed recording of size (sq cm), thickness (mm), major non-plastic inclusions (volcanic, calcareous, mixed, etc.), edge condition (eroded or angular), vessel part (rim, body, base, etc.), evidence of manufacture (coil-built, slab-built, etc.), and surface treatment (slipped, burnished, incised, etc.). For pieces diagnostic of vessel size and form, extra measurements accounted for body curvature, opening diameter, vessel height, or base width. Although only the basic information is presented in this report, the detailed findings are retained for further study. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 28 Stone, shell, and other non-ceramic artifacts were described in terms of their material type and form. Attempts were made to assess artifact functions based on knowledge of others in the region. All artifacts were counted and weighed by provenience, and these values were entered in standard tabular format for comparison of location and stratigraphic context. Representative pieces were illustrated or photographed. Further studies of density of findings and spatial distributions may be based on the numbers of items or weights per liter in excavations, whereas these calculations can be formulated for items or weights per sq m of surface-discovered materials. Faunal remains were separated into major taxonomic categories, counted, and weighed by provenience and recorded in standard tabular form. Shellfish remains were identified to family or genus if possible, but some materials could be identified only in higher-level taxonomic groupings. Terrestrial gastropod (land snail) shells were identified to genus when possible. Vertebrate animal bones were identified as major categories of fish, bird, etc. Pottery offered the primary means but certainly not the only means to establish a relative chronology in conjunction with the site stratigraphy. Certain forms of stone tools, shell ornaments, and other artifacts also served as relative time-markers. Radiocarbon dating was performed for two samples of carbonized coconut endocarps, thus establishing absolute dates for key points in the relative sequence. Additional studies may yet be considered for research outside the scope of the requirements of a resource identification survey. Toward these potential goals, USFWS will be consulted, and permission will be requested for any specific type of analysis. A permanent curation facility is not yet available that can accept new archaeological collections in Guam pursuant to 36 CFR 79. In this case, the newly recovered archaeological materials were stored temporarily at the Richard F. Taitano Micronesian Area Research Center (MARC) of the University of Guam (UOG). All materials are available upon any request made by USFWS. The materials have been prepared to return to USFWS as the responsible managing party. The U.S. National Park Service (NPS) maintains one curation facility at the War in the Pacific National Park (WAPA) compatible with the standards of 36 CFR 79, but it does not have the space or the capacity to accept new materials while already operating at its practical limits. Potentially, NPS may consider to develop an inter-agency agreement with USFWS. Although this option remains open, it is beyond the scope of performing the present survey report. Guam Museum currently is in process of constructing a new facility, but the potential for curation has not yet been ascertained. Construction has been delayed over the last few years, but possibly it can be completed before the end of 2015. If ever the facility can be evaluated as meeting the curation standards of 36 CFR 79, then Guam Museum could be considered by USFWS for long-term curation of archaeological materials from GNWR. The Archaeology Office at MARC includes a self-contained curation facility specifically designated for archaeological materials. It has not yet been inspected for confirming whether or not its current condition will satisfy the curation standards of 36 CFR 79, but an inspection is expected before the end of the year 2015. The separate rooms in the office suite are designated for: a) main office space; b) equipment storage; c) materials wet processing; d) materials dry processing; and e) curation room. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 29 The curation room has its own fire-shield door, theft alarm system, video camera monitoring, rolling metal shelving storage, archival boxes and plastic bags, database management system for collection, air conditioned environment, fire-suppressant system, pest control, and other protocols in place. Given the concerns of a permanent curation facility in Guam, the field research minimized the collections of materials needing storage or curation. In-field analysis was performed to the extent practical for surface-associated findings that could left in place. Additionally, the surface collections and test excavations overall were representative yet decidedly limited in extent, thus avoiding the accumulation of large material collections. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 30 06. OVERALL SURVEY FINDINGS The 2015 survey documented several contributing elements of a proposed Historic District, so far represented in nine “Site Clusters” and another four “Areas of Interest” (see Figure 3). The full spatial extent of the proposed Historic District very likely extends outside the boundary of the 2015 survey area, where continued survey effort will be necessary. As summarized in Table 1, each identified resource can be characterized by its material contents, interpreted cultural functions, associated time period, contribution to research significance, and recommended further management actions. Supporting documentation is provided for each identified resource in the next report sections. All of the identified archaeological resources appear to relate to a unified sense of an ancient latte village complex, considered significant for understanding the native Chamorro past. In terms of eligibility as contributing elements of a proposed Historic District, these resources retain integrity as archaeological ruins, and they contain substantive material information about local and regional cultural history. The latte village complex may be observed and appreciated as embodying the distinctive characteristics of a type, period, and method of construction that in total conveys a sense of the traditional village settlement as a significant and distinguishable entity of cultural history. The existing survey results already have contributed important new information, and each identified resource is likely to contribute further information through continued studies and through public interpretation programs. The represented time period of the apparent latte village complex is estimated as having lasted some centuries, approximately within the range of A.D. 1000 through 1700. The area of Ritdian is attested in historical records as supporting a native Chamorro population in the late 1600s, and the local population was forced to re-locate before 1700, as described elsewhere in this report. Also as described elsewhere in this report, the potential earliest time range of the latte period is estimated around A.D. 1000 generally in the Mariana Islands and specifically in other parts of Ritidian. The surface-related materials relate to the last time when the latte village was inhabited, apparently during the late 1600s. Additionally, from two of the 2015 text excavations, radiocarbon dates indicate that the surface-associated midden layers had accumulated primarily during the 1600s (Table 2). Furthermore, artifacts of known historical age of the late 1600s included a piece of high-fired porcelain at Site Cluster 01 and a large iron nail at Site Cluster 05. In total, the identified resources offer important information about the functioning of a traditional latte village complex, with multiple inter-related contributing elements all preserved in place. The cohesiveness of the site complex is a rare occurrence in Guam. Although each individual resource can be mapped with a distinctive and separate boundary, part of the resource significance overall relates to the unity of a visible latte village landscape as a proposed Historic District. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 31 Table 1. Summary of archaeological resources, identified in the 2015 survey area. Material Contents Cultural Function Time Period Research Significance Recommended Management Resource Feature A - Coral limestone formation Possible latte quarry AD 1000– 1700 Latte-building process and context Preserve in place; further study B - Coral limestone formation Possible latte quarry AD 1000– 1700 Latte-building process and context Preserve in place; further study C - Coral limestone formation Possible latte quarry AD 1000– 1700 Latte-building process and context Preserve in place; further study D - Freshwater seep Probable water source AD 1000– 1700 Cultural use of water sources Preserve in place; further study A Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study C Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study 01 02 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 32 Resource Feature Material Contents Cultural Function Time Period Research Significance Recommended Management C Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study D Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study E Upright boulders Unknown AD 1000– 1700 Part of latte village complex Preserve in place; public interpretation; further study F Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study G Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study H Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study I Stone-filled platform Unknown AD 1000– 1700 Part of latte village complex Preserve in place; public interpretation; further study J Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 33 Resource Feature Material Contents Cultural Function Time Period Research Significance Recommended Management K Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study L Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Cave and midden Temporary camp or other unknown activity AD 1000– 1700 Unknown relation with latte village complex Preserve in place; further study B Midden Temporary camp or other unknown activity AD 1000– 1700 Unknown relation with latte village complex Preserve in place; further study A Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study C Structural ruins and Residential AD 1000– Creation and development of Preserve in place; public 03 04 05 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 34 Resource Material Contents midden Cultural Function habitation Time Period 1700 Research Significance latte village complex Recommended Management interpretation; further study D Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study E Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study F Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study C Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Structural ruins and midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study Feature 06 07 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 35 Resource Feature Material Contents Cultural Function Time Period Research Significance Recommended Management C Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study B Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study C Structural ruins Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study D Midden Residential habitation AD 1000– 1700 Creation and development of latte village complex Preserve in place; public interpretation; further study A Structural elements in manufacture Probable latte workshop AD 1000– 1700 Latte-building process and context Preserve in place; public interpretation; further study B Structural elements in manufacture Probable latte workshop AD 1000– 1700 Latte-building process and context Preserve in place; public interpretation; further study C Structural elements in manufacture Probable latte workshop AD 1000– 1700 Latte-building process and context Preserve in place; public interpretation; further study 08 09 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 36 Table 2. Radiocarbon dating results from the 2015 survey area. 13C/12C Sample Beta-# 414210 Provenience Site Cluster 02, Feature A,TU-2 Material Carbonized nutshell, cf. coconut endocarp, 0.7 g Measured Age (years B.P.) 250 ± 30 ratio (‰) -25.3 Conventional years (B.P.) 250 ± 30 Calibrated calendar years A.D. (95.4%) 1521–1575 (14.6%) 1585–1590 (0.4%) 1626–1679 (55.2%) 1764–1801 (21.3%) 1939–1950 (4%) 414214 Site Cluster 05, Feature A, TU-2 Carbonized nutshell, cf. coconut endocarp, 0.6 g 300 ± 30 -24.7 300 ± 30 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 1489–1604 (69.6%) 1611–1654 (25.8%) 37 07. AREAS OF INTEREST A, B, AND C These “Areas of Interest” refer to natural formations of coral limestone, possibly used as quarries for making the stone pillar and capital elements of latte structures. They consist of fossilized coral that once had lived during a period of slightly higher sea level. The coral limestone retains visible fragments of seashells, branch corals, and other organisms of the lagoon-reef habitat where the coral platform originally had grown. In the latte structural remains of nearby sites, the quarried and cut stone elements possess the same visual characteristics of the coral limestone formation as seen in these noted places (A, B, and C) along the shoreline. Furthermore, these instances of coral limestone contain natural cracks and fissures that reveal a tendency to fracture into tabular shapes which are typical of the latte elements in the nearby archaeological sites. This breakage pattern follows the original growth of the coral in horizontal platforms or terraces. Considerably older limestone formations in many cases no longer retain the internal layering, but these particular instances along the shoreline at Ritidian do retain this type of internal structure. The observed coral limestone formations each occur along the edge of the present shoreline, only partially exposed above the high-tide level and fully exposed at the low-tide level (Figures 15 and 16). The lengths along the shoreline were measured as 50 m for Area A, 13 m for Area B, and 20 m for Area C. The visible landward-seaward widths in each case vary 1.5 to 3 m, depending on localized conditions of beach sand, noting that the beach sand is easily moved during strong tidal events and storm-surge episodes. Figure 15. Portion of a possible latte quarry in Area of Interest B, view to northwest. Scale bar is in 20-cm increments. Photograph by Dr. Hiro Kurashina. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 38 Figure 16. Portion of a possible latte quarry in Area of Interest C, view to northwest. Scale bar is in 20-cm increments. Diagnostic traits of tool-cutting marks were not observed that otherwise could confirm the existence of latte quarrying. Nevertheless, in recognition that such evidence may yet be discovered, these instances are noted as “possible latte quarry areas.” The most convincing evidence potentially has been covered under beach sand or has been otherwise obscured from modern view. More convincing evidence of workshop preparing of latte elements was found at “Site Cluster 09,” as described elsewhere in this report. This more convincing workshop area, however, was not an in situ quarry where the stone elements were carved directly from a geological formation. Rather, those stone elements already had been moved into a new position for finer working. In a review of latte quarries in the Mariana Islands, April (2004:75–83) described a quarry along the shoreline of Urunao, adjacent to the west of Ritidian in northern Guam. The Urunao latte quarry occurred in a natural formation of coral limestone found along the present-day shoreline, similar to the areas as noted here at Ritidian. The natural horizontal layering of the old coral platform apparently contributed to the ability to remove large tabular-shaped pieces. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 39 08. AREA OF INTEREST D This freshwater seep occurs near the shoreline, close to “Site Cluster 05” (Figure 17). The water flows from the natural aquifer of the island, running out to discharge just above sea level at this location. Due to its elevation just above sea level, the seep is most clearly visible during times of middle-tide or low-tide levels. The seep flows through an area of deep beach sand that offers little or no resistance against the flow of the water, situated between rocky limestone formations that by comparison offer more resistance against the water flow. Given that all human groups require sources of fresh water for basic survival, the presence of a freshwater seep very likely relates to the distribution of residential habitation sites in this area. The nearest such habitation evidence is at “Site Cluster 05”, within 50 m of this freshwater seep. Other habitation remnants as identified during the 2015 survey are situated within 300 m. Additional seeps may yet be discovered, plus pools of clean water may have been accessed inside caves in the limestone terrain. Figure 17. Freshwater seep in Area of Interest D. Photograph by Dr. Hiro Kurashina. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 40 09. SITE CLUSTER 01 Site Cluster 01 is comprised of the stone remains of at least three latte structures, each with an associated midden concentration (Figures 18, 19, and 20). The three identifiable components are referenced here as Features A, B, and C. The stone structural ruins, thickened-rim pottery, and other materials all are consistent with the interpretation of a latte residential habitation. The site materials are distributed over a length of about 50 m northwest to southeast, mostly along an elevation contour between 6 m and 7 m above present sea level. This distribution is parallel with the coast and about 50–60 m inland from today’s shoreline. A few remnants of stone rubble are distributed farther seaward (northeast) near Feature A, within the range of 5 m to 6 elevation, resembling the disturbed remnants of latte structures that no longer can be defined clearly. The feature components are situated at the interface of the rough limestone terrace formation with a fringe of clay silt deposit. The latte structures were based at the downslope or seaward (northeast) edge of the limestone formation. Portions of the associated midden concentrations have accumulated at least partially in the adjacent downslope zone of clay silt. Figure 18. Overall map of Site Cluster 01. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 41 Figure 19. Map detail 1 of 2, Site Cluster 01. Feature 01-A The area of Feature A had been studied as part of an archaeological field training program supervised by Dr. Mike T. Carson at University of Guam in 2010 (Carson 2014a), whereas the full documentation as reported here was accomplished in April and June 2015. In preparing for the 2010 training program, Feature A was identified as the largest and best preserved of the latte sets in this general vicinity. Although limited in scope, the 2010 study at Feature A provided a controlled setting and substantive training opportunities for several University of Guam students (Figures 21, 22, and 23). The latte structure of Feature A appears to have been constructed as a set of four paired rows of pillars or haligi, each topped by a disc-shaped capital or tasa. Each pillar originally stood about 1 m tall, topped by a large tasa disc more than 1 m in diameter (Figure 24). Today, most of the stone components are toppled, broken, and displaced to some degree, but the original design remains identifiable. The original design is observed most clearly by one haligi still in upright position at the southeast end of the original structure footprint (Figure 25). Following from this identification in the corner of the original latte structure, the other haligi can be observed in paired rows distributed northwest-southeast along an elevation contour between 6 m and 6.5 m. The tasa, originally positioned atop each pillar or haligi, now are toppled and broken, except for two instances of intact large discs that now are toppled but remain intact. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 42 Figure 20. Map detail 2 of 2, Site Cluster 01. Figure 21. Recording surface findings at the midden concentration of Feature A, Site Cluster 01, during 2010. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 43 Figure 22. Recording surface findings at the latte structural remains of Feature A, Site Cluster 01, during 2010. Figure 23. Sieving to recover artifacts and midden from TU-1 at Feature A, Site Cluster 01, during 2010. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 44 Figure 24. View of latte structural ruins of Feature A, Site Cluster 01, as seen in June 2015. View is to the east. Figure 25. Standing haligi within the latte structural ruins of Feature A, Site Cluster 01, as seen in June 2015. View is to the southwest, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 45 The design of the latte at Feature A resembles the most popular occurrence of a four-paired set of haligi, but the size ratio appears unusual for the relationship between the haligi and tasa. At least within the immediate surrounding area at Ritidian, the design of tasa discs larger than 1 m diameter appears uniquely constrained to this one reported instance at Feature A. Most other tasa at Ritidian measure considerably less than 1 m diameter. Additionally, the shapes of individual haligi at Feature A are noticeably more straight-sided than others at Ritidian, without a significant upward-tapering profile, perhaps so that the broader and straight-sided haligi could more effectively support the larger sizes of tasa as seen at Feature A. In contrast to the occurrence of less than one pottery fragment per sq m at the latte footprint of Feature A, a visibly dense midden concentration is visible on the surface directly adjacent to the east and northeast. Most of the midden material consists of broken pottery in densities exceeding 10 items per sq m in most instances (Figure 26), but stone and shell tools also are present (Figures 27 and 28), along with discarded fragments of marine shells presumably from meals. The distribution of this midden material coincides with a blackened soil (Munsell color 10 YR 2/1). Whereas the 2010 field training exercises had involved in-place examination, the 2015 survey included limited collections of representative artifacts within the midden area (Figures 29, 30, 31, and 32), including a volcanic stone flaking core (collection 01), a Tridacna sp. shell adze fragment (collection 02), and a flaked volcanic stone tool with edge wear (collection 03), and a representative cluster of broken pottery (collection 04). Pottery fragments included one thickened rim with combing (67.5 g), eight plain thickened rims (639.3 g), five combed body pieces (106.1 g), and ten plain body pieces (151 g). Figure 26. Representative area of broken pottery in the midden concentration of Feature A at Site Cluster 01, as viewed in June 2015 with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 46 Figure 27. Basalt adze, found in 2010 on the surface of the midden concentration of Feature A, Site Cluster 01. Figure 28. Flaked chert artifact, found in 2010 on the surface of the midden concentration of Feature A, Site Cluster 01. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 47 Figure 29. Volcanic stone flaking core, 141.6 g, collection 01 from the midden concentration of Feature A, Site Cluster 01. Scale bar is in 1-cm increments. Figure 30. Fragment of Tridacna sp. shell adze, 31.4 g, collection 02 from the midden concentration of Feature A, Site Cluster 01. Scale bar is in 1-cm increments. Figure 31. Flaked volcanic stone with use-wear edge, 8.2 g, collection 03 from the midden concentration of Feature A, Site Cluster 01. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 48 Figure 32. Examples of thickened-rim pottery fragments, 2015 collection 04 at the midden concentration of Feature A, Site Cluster 01. Photograph at top with scale bar in 1-cm increments. Corresponding tracings of rim profiles at bottom. Although the footprint of the latte structure at Feature A did not appear to contain a substantial midden deposit, test excavations in 2010 aimed to clarify the structural foundation base. Two test units were excavated as individual 1 by 1 m units, here labelled as TU-1 and TU-2, exposing a thin layer of 30 cm of very rocky clay silt directly over the limestone bedrock (Figure 33). The sediment here contained very few pieces of broken pottery, stone flakes, marine shells, fish bones, and one tiny fragment of highfired porcelain (Table 3). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 49 Figure 33. Contiguous exposed profile of TU-1 (left) and TU-2 (right) at Feature A, Site Cluster 01. View is to the southwest, with scale bar in 20-cm increments. Table 3. Summary of findings at TU-1 and TU-2. Provenience Broken pottery High-fired porcelain Stone flake Marine shell Fish bone 1 4 1 4 3 4 2 4 2 TU-1 0–10 cm 2 10–20 cm 3 20–30 cm 2 1 TU-2 0–10 cm 2 1 10–20 cm 2 5 2 20–30 cm 2 3 2 ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 50 Figure 34. Discovery of tiny piece of high-fired porcelain at the base of TU-1 at Feature A, Site Cluster 01. Inset (lower right corner) shows closer detail. Scale bars are in 20-cm increments. The tiny fragment of high-fired porcelain measured only 9 mm in maximum length, and it was found in contact with the limestone bedrock beneath the latte structure (Figure 34). This finding provides strong evidence that the latte of Feature A was constructed at some time after the availability of high-fired porcelain in the Ritidian area, necessarily following the arrival of Spanish ships from overseas with Asian and European cargo. Prior to the Spanish colonial period, no high-fired porcelain was available in Guam, so the tiny piece at Feature A reveals a dating most likely in the late 1600s when a Jesuit missionary outpost existed at Ritidian. Feature 01-B Roughly along the same axis alignment of the latte of Feature A, another latte set’s remains are visible at Feature B. The latte elements at Feature B include the broken bases of haligi in their original positions, along with the displaced and broken pieces of other haligi and tasa. Those remnant base pieces of haligi today stand only 30–45 cm in height, although the surrounding broken pieces suggest original heights of 85–95 cm. The remnant bases further allow recognition of an original four-paired set of haligi. The downslope (northeast) side of the Feature B latte is marked by an alignment of limestone cobbles and small boulders. This construction resembles a front porch or patio at the outer edge of the original housing footprint. This design of a downslope-fronting paving alignment is seen in other latte sets of Site Clusters 02, 05, and 07. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 51 Also on the downslope (northeast) side of the Feature B latte, a concentration of broken pottery occurs on the silty sand surface of an area of about 7 sq m. The broken pottery appears consistent with other pieces described at Site Cluster 01, typical of the latte period. Diagnostic traits include thickenedrim profiles, combed exterior surfaces, and coarse-sized temper inclusions. Feature 01-C Near the northwest end of Site Cluster 01, Feature C includes the disturbed remains of a latte structure, along with a midden concentration. The original position and design of the latte cannot be reconstructed from the present-day condition of the ruins. The stone elements today are distributed mostly over a surface of rough limestone, yet some of the pieces extend over the adjacent downslope deposit of clay silt. The midden concentration is found entirely within the clay silt deposit. Stone rubble of Feature C is distributed over an area about 1.5–2 m wide and over a length of about 6 m northwest-southeast, within which at least two displaced haligi and one displaced tasa can be recognized. The one recognizable rectangular-shaped tasa exhibits a socketed base, likely enhancing the ability to be held in its original place over a tapered-top haligi (Figure 35). The two discernable haligi measure about 60 cm in height, noticeably shorter than the haligi of Features A and B. The midden concentration of Feature C is identifiable by a blackened soil surface (Munsell color 10 YR 2/1) with variable amounts of broken pottery and discarded fragments of marine shells. Within an area of approximately 45 sq m, the most visibly dense concentration of pottery on the surface was collected in a 1 by 1 m area (collection 05), consisting of two plain thickened rims (80.5 g), one combed body piece (10.8 g), and seven plain body fragments (80.7 g) (Figure 36). Scattered marine shell fragments were identifiable as pieces of Turbo spp. and Strombus sp. specimens. Figure 35. Tasa with socketed base at Feature C of Site Cluster 01. Scale bar is in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 52 Figure 36. Examples of pottery fragments, collection 05 at Feature C of Site Cluster 01. Scale bar is in 1-cm increments. Photograph at top with scale bar in 1-cm increments. Corresponding traced profiles at bottom. A 1 by1 m test unit (TU-3) was excavated adjacent to the surface-material collection 05 of the Feature C midden, revealing a single cultural layer in the upper 10 cm of a rocky clay silt deposit over the limestone bedrock (Figures 37 and 38). The clay silt had accumulated about 30 cm thick over the limestone bedrock, within which the lower 20 cm contained considerable broken limestone rubble, and the upper 10 cm contained the archaeological materials of past cultural use of the area. The archaeological materials within the 1 by 1 m excavation included 43 pieces of broken pottery and 8 pieces of marine shell (Table 4). In particular, the combed exteriors of two pottery fragments appeared consistent with the overall character of the pottery seen generally at Site Cluster 01 and considered diagnostic of the latte period (Figure 39). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 53 Figure 37. Excavation of TU-3 at Feature C of Site Cluster 01. View is to the southwest, with scale bar in 20-cm increments. Figure 38. Northwest profile of TU-3 at Feature C of Site Cluster 01. Figure 39. Pottery fragments with external combing, from 0–10 cm of TU-3 at Feature C, Site Cluster 01. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 54 Table 4. Summary of findings fromTU-3 at Feature C of Site Cluster 01. TU-3, depth (cm) 0–10 Marine shell Pottery fragments 3 Turbo sp. opercula (18.6 g) 2 combed body pieces (23.5 g) 2 Turbo sp shell fragments (22.7 g) 41 plain body pieces (407.5 g) 2 Trochus sp. shell fragments (18.8 g) 1 Strombus sp. shell (3.2 g) 10–20 none None 20–30 none None Site Cluster 01 Summary The findings at Site Cluster 01 confirm the presence of a latte residential habitation, consisting of at least three latte structures and associated concentrations of midden. The surface-visible ruins evidently were used for the last time as a residential site in the late 1600s, according to the apparent date of a piece of high-fired porcelain at the structural base of the Feature A latte. Test excavations, although limited, showed only a single cultural layer, associated with the surface-visible remains and not containing anything noticeably different from the surface-collected materials. The spatial clustering and similar alignments suggest that the latte of Site Cluster 01 functioned contemporaneously, and they likely were used by a closely interacting group of people. Each latte set here is associated with a midden concentration that appears typical of residential habitation, possibly with different kinds of specific activities in each location. Stone and shell tools appear more numerous at Feature A than at Features B and C. Likewise, the density of material overall appears more heavily concentrated at Feature A as compared to Features B and C. Further studies may yet be designed to explore the inter-relationships of the features. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 55 10. SITE CLUSTER 02 Site Cluster 02 is the largest of the clusters identified in the 2015 survey, within which 12 feature designations are distributed more or less evenly over an area of approximately 3000 sq m, between 13 m and 17 m above present sea level (Figures 40, 41, 42, 43, 44, and 45). The component features all occur in pocket localities bearing variable amounts of clay silt deposit, within a larger formation of rough limestone terrain. The placement of cultural activity in this landscape appears to have been determined by the availability of sedimentary deposits, whereas the surrounding expanse of rough limestone did not support the kinds of cultural activities that would result in substantial archaeological remains. Figure 40. Overall map of Site Cluster 02. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 56 Figure 41. Map detail 1 of 5, Site Cluster 02. The functional relationships among these features could be interpreted in variable ways. In any case, they likely were used contemporaneously with each other during the latte period, as indicated by the uniform characteristics of the archaeological materials described here. Potentially, a number of activity areas may be interpreted. Features A through L all can be characterized as related to the latte period, given their contents of latte structural remains and pieces of thickened-rim and combed pottery typical of the latte period. Most of the features consist of latte structural remains and associated midden concentrations. Two notable exceptions include a set of upright boulders at Feature E and a stone-filled platform at Feature I, adding more complexity to the potential interpretation of the surrounding archaeological resources. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 57 Figure 42. Map detail 2 of 5, Site Cluster 02. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 58 Figure 43. Map detail 3 of 5, Site Cluster 02. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 59 Figure 44. Map detail 4 of 5, Site Cluster 02. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 60 Figure 45. Map detail 5 of 5, Site Cluster 02. Feature 02-A Feature A refers to an area of nearly flat-surfaced and blackened soil (Munsell color 10 YR 2/1), bordered on its downslope (north) side by a pile of cobble and boulder rubble including at least two haligi (Figures 46 and 47). The rubble apparently had been cleared from the surface of black soil and piled in a roughly linear fashion at the edge of the cleared space, so that the cleared space now contains no cobble-sized or boulder-sized stone material. Instead the cleared space now consists of black clay silt with scattered pebbles and gravel, broken pottery, and discarded marine shell fragments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 61 Figure 46. Feature A of Site Cluster 02, view to northeast. Scale bar is in 20-cm increments. Figure 47. Displaced haligi with a upward-tapered shape, within Feature A of Site Cluster 02. Scale bar is in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 62 Figure 48. Thickened-rim pottery fragment, collection 01 from surface of Feature A in Site Cluster 02. Scale bar is in 1-cm increments. The area of black clay silt at one time probably supported a latte structure as the origin of the two haligi and other probable latte elements now re-deposited inside the pile of rubble. The remaining area of blackened soil measures about 7 m east-west by 4 m north-south, roughly comparable to the footprint of most of the latte structural remains in the general vicinity at Ritidian. The displaced haligi exhibit upward-tapered shape, and they would have stood about 45–50 cm high. Near the north edge of the blackened soil area, two earthenware potsherds were collected (collection 01) as representative of the pottery scattered over the dark soil deposit. These two pieces included one thickened rim of 99.1 g and one plain body piece of 35.5 g (Figure 48). The thickened-rim profile is characteristic of the latte period generally. Although not collected, pieces of marine shells included fragments of Turbo spp., Isognomon sp., and Tridanca sp. specimens. Adjacent to the surface-collection, an area of 1 m by 1 m was excavated as a formal test unit (TU-1), revealing a single cultural layer within a deposit of black clay silt over the limestone bedrock (Figures 49 and 50). The black clay silt had accumulated in a sedimentary unit about 35 cm thick. Within the upper 20 cm, it contained cultural material of broken pottery and marine shell fragments as seen on the surface (Table 5). Within the 1 by 1 m excavation of TU-1, recovered materials included several pieces of broken pottery and marine shells. The pottery fragments overall were redundant of others as seen on the surface, showing a thickened-rim profile, but some of these pieces showed exterior combing treatment that was not seen in the surface-visible pottery (Figure 51). The marine shells in TU-1 similarly were consistent with the surface-visible observations of the dark soil area at Feature A. No significant differentiation was evident from the top to bottom of the excavation, but rather all of the cultural material appeared to relate to a single depositional unit. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 63 Figure 49. Excavation of TU-1 at Feature A in Site Cluster 02. View is to the northeast, with scale bar in 20-cm increments, Figure 50. East profile of TU-1 at Feature A in Site Cluster 02. Table 5. Summary of materials recovered from TU-1 at Feature A in Site Cluster 02. TU-1, depth (cm) 0–10 Marine shell 2 Turbo sp. opercula (10.4 g) 6 Isognomon sp. shell fragments (7.5 g) 1 Trochus sp. shell (18.1 g) Pottery fragments 2 plain thickened rims (161.1 g) 9 combed body pieces (243.5 g) 14 plain body pieces (202.6 g) 10–20 1 Trochus sp shell fragment (6.7 g) 7 combed body pieces (212.3 g) 10 plain body pieces (143 g) 20–30/35 none None ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 64 Figure 51. Examples of pottery fragments from TU-1 at Feature A in Site Cluster 02. Top shows selected pottery from 0–10 cm, with tracing of rim profiles and scale bar in 1-cm increments. Bottom shows selected pottery from 10–20 cm, with scale bar in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 65 Radiocarbon dating was performed for a concentration of carbonized broken nutshells (resembling coconut endocarp fragments) near the base of the cultural layer in TU-1, yielding a result of highest probability in the late A.D. 1600s (see Table 2). According to historical knowledge that the Ritidian area had been abandoned as a traditional residential habitation by A.D. 1700, the post-1700 portions of the radiocarbon dating result can be disregarded, thus adding even greater probability of a dating in the late 1600s. Overall, the dating result meets the expectations that the latte village complex was last occupied during the late 1600s. Feature A provides an example of evident destruction and probable burning of a latte structure during the late 1600s. The reasons for dismantling and possibly burning the latte structure are not directly knowable from the archaeological material, but a number of plausible interpretations may be formulated. Additional similar instances are evident in other features of Site Cluster 01 and other parts of the 2015 survey area, which can be discussed more productively in a comprehensive summary. Feature 02-B Feature B includes one of the more nicely made and preserved latte sets in Site Cluster 02 and in the whole of the 2015 survey area. This particular latte was designed with a rare three-paired set of haligi, each with an accompanying disc-shaped tasa (Figure 52). Additionally, a paving alignment of cobbles and boulders was positioned along the downslope (north) edge of the latte footprint (Figure 53), incidentally constructed precisely along the 16.5 m elevation contour. Figure 52. Remains of latte structure at Feature B of Site Cluster 02. View is to the east, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 66 Figure 53. Portion of the paving alignment at latte structure, Feature B of Site Cluster 02. View is looking down toward the west, with scale bar in 20-cm increments. Although necessarily smaller in size than the more popular four-paired set of haligi, the threepaired example at Feature B exhibits impressively careful shaping of the quarried stone elements (Figures 54 and 55). The haligi each display an evenly upward-tapered shape at 90 cm height, and their surfaces are remarkably smoothed in contrast to the usual roughened surfaces of other examples of quarried pieces of limestone at Ritidian. The tasa likewise are remarkably smoothed, and they display in each case an evenly rounded disc-like shape of 40 cm diameter. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 67 Figure 54. Detail of haligi and tasa within latte structure, Feature B of Site Cluster 02. View is to the east-southeast, with scale bar in 20-cm increments. Figure 55. Detail of haligi and tasa within latte structure, Feature B of Site Cluster 02. View is to the southeast, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 68 Figure 56. Flaked stone, 20.4 g, collection 02 from Feature B of Site Cluster 02. Scale bar is in 1-cm increments. The paving alignment appears to have been partially disturbed in its center portion, leaving a large intact section at the east end and a very small intact portion at the west end. Near the apparently disturbed portion, one piece of flaked volcanic stone (collection 02) was collected from the surface (Figure 56). No other artifacts or midden were observed on the surface on or around the paving alignment. On the ground surface of the Feature B latte footprint, only extremely sparse pottery fragments and shells were visible. The concentrations were measured as less than 1 piece per sq m. These materials included thickened-rim pottery, Tridacna sp. shells, and Turbo spp. shells. The underlying clay silt deposit appeared dark grayish brown (Munsell color 10 YR 4/2) in contrast to the black color (Munsell color 10 YR 2/1) of the denser midden concentrations as observed elsewhere in the 2015 survey area. A small test unit (TU-2) of 50 by 50 cm revealed 25 cm of rocky clay silt over the basal limestone bedrock at Feature B (Figure 57). This test unit was excavated in the apparently disturbed portion of the paving alignment, at the edge of the intact portion of the paving construction. The paving evidently had been emplaced on the top of the thin layer of clay silt, and very sparse cultural material had infiltrated downward into the clay silt (Table 6). Recovered materials included just one piece of combed pottery and five Turbo sp. opercula within the upper 10 cm of the rocky clay silt (Figure 58). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 69 Figure 57. Excavation of TU-2 at Feature B of Site Cluster 02, showing the paving alignment based in shallow rocky clay silt over the limestone bedrock. View is to the south-southeast, with scale bar in 20-cm increments. Figure 58. Materials excavated from 0–10 cm of TU-2, Feature B of Site Cluster 02, showing one combed pottery piece on the left and five Turbo sp. opercula on the right. Scale bars are in 1-cm increments. Table 6. Summary of materials recovered from TU-2, Feature B of Site Cluster 02. TU-2, depth in cm Marine shell Pottery fragments 0–10 cm 5 Turbo sp opercula (36.9 g) 1 combed body piece (6.4 g) 10–20/25 cm none none ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 70 Feature 02-C Feature C consists of the remains of a latte stricture, immediately adjacent with a midden concentration. These materials are distributed roughly along the 17 m elevation contour. Feature C occurs within a small pocket of clay silt among the surrounding expanse of rough limestone terrain. The midden concentration had been stained black (Munsell color 10 YR 2/1), apparently due to burned material and organic carbon, in contrast to the lighter color elsewhere at Feature C described as dark grayish brown (Munsell color 10 YR 4/2). The midden concentration here contained variable densities of pottery, ranging from 6 to 15 pieces per sq m, among similarly variable densities of broken shells of Turbo spp., Trochus sp., Tridacna sp., and Strombus sp specimens. In one representative area of 1 m by 1 m (collection 03), nine pieces of pottery included one thickend rim with a combed exterior (140.6 g), three plain thickened rims (195.5 g), and five combed body fragments (189.5 g) (Figure 59). Additionally, a pumice abrader (collection 04) of 3.1 g was found here (Figure 60). Figure 59. Representative pottery fragments, collection 03 from midden concentration at Feature C of Site Cluster 02. Top shows photographs, and bottom shows tracings of rime profiles. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 71 Figure 60. Pumice abrader, 3.1g, collection 04 from midden concentration at Feature C of Site Cluster 02. Scale bar is in 1-cm increments. The latte remains at Feature C reveal a design of a four-paired set of haligi, each with a corresponding tasa, plus an end-border of a group of aligned boulders (Figures 61 and 62). The upwardtapered haligi originally stood 75 cm high, capped by disc-shaped tasa of about 50 cm diameter. The boulder alignment was positioned at the southeast end of the long axis, so far constituting the only known example of this kind of construction in the 2015 survey area. The boulder alignment at the longaxis end of the Feature C latte contrasts against other stone alignments that were made of flat-surfaced cobbles and small boulders along the downward-slope edge of other latte structure footprints. At the edge of the Feature C latte footprint, a 50 cm by 50 cm test unit (TU-3) revealed extremely sparse cultural material within a thin layer of less than 10 cm of rocky clay silt over the limestone bedrock (Figures 63 and 64). The recovered materials included just two Turbo spp. opercula and two pieces of pottery (Table 7). The two pieces of pottery were plain body fragments without any diagnostic traits of exterior combing or thickened-rim profile, but the thickness (8–11 mm) and coarse paste with large temper inclusions were consistent with observations of pottery fragments at the adjacent midden concentration of Feature C. The boulder alignment within Feature C is of an unknown function, although a number of interpretations are possible. Given the position between the latte structure and the midden concentration, the boulder alignment at least partially may have marked a division between two cultural activity zones, such as to differentiate the living space inside the latte versus the apparent discard of material in the midden concentration. Other possibilities may relate to a step for facilitating access to and from the raised living floor of the latte structure. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 72 Figure 61. Remains of latte structure at Feature C of Site Cluster 02. View is to the southwest, with scale bar in 20-cm increments. Figure 62. Boulder alignment in foreground at southeast end of latte set, at Feature C of Site Cluster 02. View is to the northwest, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 73 Figure 63. Excavation of TU-3 at Feature C, Site Cluster 02, showing shallow rocky clay silt over the limestone bedrock. View is to the south, with scale bar in 20-cm increments. Figure 64. East profile of TU-3 at Feature C, Site Cluster 02. Table 7. Summary of materials recovered from TU-3 at Feature C, Site Cluster 02. TU-3, depth (cm) 0–10 Marine shell 2 Turbo sp. opercula (9.4 g) Pottery fragments 2 plain body pieces (61.3 g) Feature 02-D Feature D includes the remains of a latte set, partially overlapping with a midden concentration. These materials are constrained mostly within an area of 11 m north-south by 5 m east-west, roughly along the 16.5 m elevation contour. A pile of stone rubble is situated about 4 m farther downslope (east), possibly related to a displaced group of disturbed latte elements. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 74 Figure 65. Latte structural remains at Feature D, Site Cluster 02. View is to the east-southeast, with scale bar in 20-cm increments. Figure 66. Socketed tasa in middle foreground of latte set at Feature D, Site Cluster 02. View is down to northwest, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 75 The latte elements of Feature D are nicely shaped, although the original design and layout of the structure is difficult to interpret from the positions of the surviving elements (Figure 65). At least four haligi and four tasa are obvious, although an effective latte structure presumably included more elements that are no longer clearly identifiable. The haligi each have an upward-tapered shape, and they would have stood 80–90 cm high. The disc-shaped tasa each measure about 50 cm in diameter, and one tasa displays a shallow socket (Figure 66) that may have facilitated emplacement over the tapered top of an accompanying haligi. The midden concentration in this case at Feature D occurs in an area of dark grayish brown clay silt (Munsell color 10 YR 4/2). Within the midden concentration, one pumice abrader (collection 05) was found just outside the apparent footprint of the latte structure (Figure 67). A representative area of 1 m by 1 m contained four pieces of broken pottery (collection 06), including one thickened rim with a combed exterior (89.5 g), another plain thickened rim (90.2 g), and two plain body fragments (75.1 g) (Figure 68). A small test unit (TU-4) of 50 by 50 cm revealed denser cultural material than was visible on the surface (Table 8), within the upper 10 cm of a 20-cm-thick layer of rocky clay silt over the limestone bedrock (Figure 69). Although occurring in denser concentration than the material on the surface, the excavated items were overall consistent with the surface-visible findings. Recovered materials included four Turbo sp. opercula, one piece of broken Turbo sp shell, one plain thickened-rim pottery fragment, one combed pottery body piece, and 15 plain body fragments of pottery (Figure 70). Figure 67. Pumice abrader, 23 g, collection 05, from Feature D, Site Cluster 02. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 76 Figure 68. Examples of pottery fragments, collection 06, from Feature D, Site Cluster 02. Top shows photographs, and bottom shows corresponding tracings of rim profiles. Scale bar is in 1cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 77 Figure 69. Excavation of TU-4 at Feature D, Site Cluster 02. View is to the southeast, with scale bar in 20-cm increments. Figure 70. Examples of pottery recovered from TU-4 at Feature D, Site Cluster 02. Scale bar is in 1-cm increments. Table 8. Summary of material excavated from TU-4 at Feature D, Site Cluster 02. TU-4, depth (cm) 0–10 Marine shell Pottery fragments 4 Turbo sp. opercula (24.5 g) 1 plain thickened rim (10.5 g) 1 Turbo sp. shell fragment (1.6 g( 1 combed body piece (8.1 g) 15 plain body pieces (87.6 g) 10–20 none none ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 78 Feature 02-E Feature E is a unique occurrence in the 2015 survey area, consisting of a set of six upright boulders (Figure 71). Unlike the haligi of latte structures, the boulders at Feature E were not quarried or shaped. Rather, they were naturally occurring boulders, evidently moved into upright position at evenly spaced intervals over an area of nearly flat ground between the elevation contours of 15.5 m and 16 m. Furthermore, no accompanying tasa are evident, so the boulders apparently stood alone as upright masses. The individual boulders varied 30 cm to 60 cm in height. The ground surface around the boulders may be described as rocky clay silt of a dark grayish brown color (Munsell color 10 YR 4/2), consistent with the naturally occurring clay silt elsewhere in the vicinity. Very sparse pottery fragments were observed in concentration of less than 1 piece per sq m. Among the visible pottery fragments, no diagnostic traits were evident such as thickened-rim profile or combed exterior surface, but the thickness (8–11 mm), coarse clay paste, and large-sized temper inclusions nonetheless were consistent with other pottery of the latte period in the general vicinity. Test excavation was not attempted at Feature E, due to the apparently very low potential for finding any subsurface cultural materials. Expedient probing with a trowel showed that the clay silt was only 5 cm tick over the underlying limestone bedrock. Additionally, the light sediment color and sparse cultural material on the surface suggested that the underlying clay silt contains similarly sparse material not significantly different from the surface-visible observations. Figure 71. Set of upright boulders, Feature E of Site Cluster 02. View is to the east-northeast, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 79 The function of Feature E is unclear at this time, but it may have related to the marking of the edge of the residential space of Site Cluster 02. Feature D is situated approximately in the middle of the seaward (northeast) side of the total distribution of features in Site Cluster 02. No other archaeological features were found between Feature E and the shoreline about 190 m to the northeast, so accordingly Feature E would mark the first and centrally located portion of Site Cluster 02 when approaching in a direct line from the shoreline. Feature 02-F Feature F refers to a midden concentration of broken pottery and marine shells, covering an area of approximately 6 m north-south by 2.5 m east-west. The underlying ground surface is described as a rocky clay silt of dark brown color (Munsell color 10 YR 3/3), unlike the much darker-stained organic deposits in denser midden concentrations elsewhere in the 2015 survey area. Probing with a trowel revealed that the rocky clay silt was only 5 cm thick over the underlying limestone bedrock. The observed pieces of pottery and shells were typical of the surrounding vicinity and of the latte period generally. The pottery fragments occur in overall low density of 2–4 pieces per sq m, including a few pieces with diagnostic traits of thickened-rim profile, combed exterior surface, coarse clay paste, and large-sized temper inclusions. Scattered marine shells include pieces of Tridacna sp. and Turbo spp. specimens. The midden concentration at Feature F most likely was associated with the latte structural remains at Feature G. These two features are separated by only 3 m. Especially when considering an overhanging floor and roof of a superstructure beyond the edges of the supporting haligi and tasa, the midden materials of Feature F most likely can be associated with the latte at Feature G. Feature 02-G Feature G includes the remains of a latte set, with its long axis aligned northwest-southeast near the 16 m elevation contour. The observable stone elements reveal an original construction of a fourpaired set of haligi, although nearly all of these haligi now are broken with just 30–40 cm of their basal portions remaining (Figure 72). One nearly complete haligi stands at 75 cm height. The tasa are no longer discernable, although presumably their broken fragments may be found among the nearby broken rubble debris. The ground surface at Feature G revealed extremely sparse concentrations of broken pottery and marine shells, occurring in densities of less than 1 item per sq m. The underlying sediment did not display dark coloring of burned material or organic carbon, but rather it consisted of rocky clay silt of dark brown color (Munsell color 10 YR 3/3). The pottery fragments did not include diagnostic pieces of thickened rims or combed bodies, but they were consistent with others of the latte period with 8–11 mm thick bodies, coarse clay paste, and coarse-sized temper inclusions. The marine shells were identified as Tridacna sp. and Turbo spp. specimens. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 80 Figure 72. Latte structural remains at Feature G, Site Cluster 02. View is to the northwest, with scale bar in 20-cm increments. Figure 73. Excavation of TU-5 at Feature G, Site Cluster 02, showing remnant of haligi based at the top of shallow rocky clay silt over the limestone bedrock. View is to the southwest, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 81 Figure 74. Northwest profile of TU-5 at Feature G, Site Cluster 02. Figure 75. Bird bone fragment, recovered from TU-5 at Feature G, Site Cluster 02. Table 9. Summary of materials recovered from TU-5 at Feature G, Site Cluster 02. TU-5, depth (cm) 0–10 Marine shell 1 Turbo sp. operculum (7.5 g) Bird bone 1 fragment, possibly polished (0.3 g) Pottery fragments 1 combed body piece (17.2 g) 2 plain body pieces (34.7 g) 10–12/18 none none none At the downslope (northeast) edge of one haligi remnant base, a test unit (TU-5) of 40 by 60 cm revealed a thin cultural deposit within the rocky clay silt over an undulating base of limestone bedrock (Figures 73 and 74). The deposit was of variable depth over the uneven limestone, ranging 12–18 cm, and cultural materials were confined to the upper 10 cm. As detailed in Table 9, recovered materials included one Turbo sp. operculum, one curved and possibly polished bird bone fragment (Figure 75), one pottery fragment with a combed exterior, and two other examples of plain pottery body fragments (Figure 76). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 82 Figure 76. Pottery fragments from TU-5 at Feature G, Site Cluster 02. Feature 02-H Feature H consists of an area of blackened soil covering approximately 12 m northeastsouthwest by 6 m northwest-southeast, with pile of stone rubble on its downslope (northeast) edge near the 16 m elevation contour. The cobble-sized and boulder-sized rubble most likely had been cleared from the area of dark soil, which now is devoid of any stones larger than the size of gravels or pebbles. The rubble does not contain any diagnostic pieces of latte elements as seen at other similar features in the 2015 survey area, although perhaps any latte elements have been broken to the point of losing their diagnostic traits. The dark soil consists of black (Munsell color 10 YR 2/1) clay silt, with variable densities of broken pottery and marine shells. The pottery fragments vary in density from 2 to 5 pieces per sq m, including a few pieces with thickened-rim profile or combed exteriors typical of the latte period. The marine shells include pieces of Turbo spp., Trochus sp., Tridacna sp., Cypraea sp., and Strombus sp. specimens. The long landward-seaward axis of Feature H differs from the usual expectation of a house structure oriented along the natural elevation contour parallel with the shoreline. The different orientation in this case may relate more to the resulting patterning of the burned debris of a former structure than to the original footprint of such a structure. In any case, no actual structural ruin is evident, as the pile of stone debris at this time contains broken limestone fragments without diagnostic traits of haligi or tasa. Feature 02-I Feature I is unique within the 2015 survey area, consisting of a stone-filled platform and an associated midden concentration (Figures 77 and 78). The platform is square in plan view, about 4.5 m by 4.5 m, and it stands about 50 cm high. The midden contains pottery fragments, stone tools, and marine shells distributed around the perimeter of the stone-filled platform. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 83 Figure 77. Stone-filled platform at Feature I, Site Cluster 02. View is to the southeast, with scale bar in 20-cm increments. Figure 78. Detail of southwest corner of stone-filled platform at Feature I, Site Cluster 02. View is to the north, with scale bar in 20-cm increments. Samples of artifacts from the midden included four separate stone tools (collections 07, 08, 09, and 10), as well as pottery fragments from an area of 50 cm by 50 cm (collection 11). The stone tools are described as three pumice abrader fragments and one piece of dense volcanic stone that may have been a hammerstone (Figures 79, 80, 81, and 82). The pottery fragments were collected from an apparently dense occurrence on the surface outside the downslope (northeast) edge of the stone-filled platform, including five pieces of thickened rims (391.5 g) , four combed body pieces (98.8 g), and five plain body pieces (85.2 g) (Figure 83). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 84 Figure 79. Pumice abrader fragment, 48.1 g, collection 07 from Feature I, Site Cluster 02. Scale bar is in 1-cm increments. Figure 80. Pumice abrader fragment, 30.6 g, collection 08 from Feature I, Site Cluster 02. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 85 Figure 81. Pumice abrader fragment, 168.4 g, collection 09 from Feature I, Site Cluster 02. Scale bar is in 1-cm increments. Figure 82. Dense volcanic stone artifact, 850 g, collection 10 from Feature I, Site Cluster 02. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 86 Figure 83. Examples of pottery fragments, collection 11 from Feature I, Site Cluster 02. Photograph with scale in 1-cm increments at top. Corresponding tracings of rim profiles at bottom. Adjacent to the area of surface-collected pottery, a 50 cm by 50 cm test unit (TU-6) revealed a single cultural layer with a very dense occurrence of pottery and other cultural material (Table 10). Cultural materials were found within the upper 10 cm. The deposit as a whole consisted of a single layer of rocky clay silt directly over the limestone bedrock, varying 15 cm to 25 cm depth over the uneven limestone (Figures 84 and 85). Recovered shell fragments represented two Turbo sp. opercula, three Turbo sp. shells, two Trochus sp. shells, one Tridacna sp. shell, and one Strombus sp. shell. Recovered pottery fragments included one thickened rim with a combed exterior, three plain thickened rims, one combed body piece with lime plaster coating, eight other combed body pieces, and 115 plain body fragments (Figure 86). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 87 Table 10. Summary of material excavated from TU-6 at Feature I, Site Cluster 02. TU-6, depth (cm) 0–10 Marine shell Pottery fragments 2 Turbo sp. opercula (10.5 g) 1 combed thickened rim (58.9 g) 3 Turbo sp. shell fragments (31 g) 3 plain thickened rims (214.6 g) 2 Trochus sp. shell fragments (8.2 g) 1 Tridacna sp. shell fragment (35.8 g) 1 combed body piece with lime plaster (27.7 g) 8 combed body pieces (183.5 g) 115 plain body pieces (1377.3 g) 1 Strombidae shell fragment (16.3 g) 10–15/25 none none Figure 84. Excavation of TU-6 at Feature I, Site Cluster 02, showing the base of the stonefilled platform in the top of a thin layer of rocky clay silt over the limestone bedrock. View is to the southwest, with scale bar in 20-cm increments. Figure 85. Northwest profile of TU-6 at Feature I, Site Cluster 02. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 88 Figure 86. Examples of pottery fragments, excavated from TU-6 at Feature I, Site Cluster 02. The cultural function of Feature I is unclear at this time, but most likely it functioned as part of the surrounding latte village complex. The pottery fragments are diagnostic of the latte period generally, and they are consistent with findings at the known latte structures in the vicinity. Albeit limited in scope, the test excavation of TU-6 revealed a single cultural activity period, associated directly with the construction and use of the stone-filled platform at Feature I, presumably as part of the overall latte period cultural activity in this general area. Clues of the function of Feature I may be seen in the shape and design of the platform, as distribution and density of midden material, and location relative to other features in the surrounding landscape. The stone-filled platform displays a rough surface that could not have been suitable as a living floor, although it may have marked a location of significance that could be identified from a distance. The unusually dense pottery fragments, stone tools, and marine shells are distributed around the perimeter of the platform but not within the stone-filled space, thus indicating that the acts of material deposition deliberately were removed from the platform itself but positioned as close as ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 89 possible to the exterior. The location of Feature I is at least 4 m distant from the next nearest features of Site Cluster 02, definitely within the scope of the site complex but spatially distinctive. The findings at Feature I suggest most likely a place marking a significant event whose details are no longer known precisely. In the past, specifically during the latte period, people may have left commemorative or dedicatory offerings around the edges and especially at the downslope (northeast) side of the platform, but the interior space of the platform was left clear of any such materials. Any more specific details are unknown at this time, although of course several hypothetical interpretations may be formulated. Feature 02-J Feature J is an area of black (Munsell color 10 YR 2/1) clay silt, bordered on its east side by a pile of stone rubble (Figure 87). Within the roughly 50 sq m of the dark soil area, pottery fragments are visible in very low concentration of considerably less than 1 piece per sq m. The densest single occurrence of broken pottery (collection 12) included one thickened-rim piece (83.2 g), one plain body piece (36.5 g), and one other body piece with faint multi-directional combing (53.8 g) (Figure 88). The stone rubble appears to have been cleared from the area of black clay silt and re-deposited in a pile outside the dark soil, leaving very little stone material within the cleared area. One piece of rubble resembles a haligi of 95 cm height, now fallen on its side. Another piece resembles a disc-shaped tasa of 50 cm diameter. A few of the other rubble fragments display portions with quarrying or cutting marks as expected of latte elements, but the broken conditions disallow clear recognition as potentially relating to specific forms of haligi or tasa. Figure 87. Feature J, Site Cluster 02, as seen in April 2015 prior to Typhoon Dolphin of May 2015. View is to the south-southeast, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 90 Figure 88. Pottery fragments, collection 12 from Feature J, Site Cluster 02. Scale bar is in 1-cm increments. Figure 89. Excavation of TU-7 at Feature J, Site Cluster 02, showing thin clay silt over limestone bedrock. View is to south, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 91 Figure 90. North profile of TU-7 at Feature J, Site Cluster 02. Figure 91. Combed pottery fragment, recovered from TU-7 at Feature J, Site Cluster 02. Table 11. Summary of materials recovered from TU-7. Site Cluster 02 TU-7, depth (cm) Marine shell 0–10 1 Strombus sp. shell fragment (3.5 g) 10–21/32 none Pottery fragments 1 combed body piece (7 g) 6 plain body pieces (50.5 g) none A 50 by 100 cm test unit (TU-7) exposed a single cultural layer within the deposit of black clay silt overlaying the limestone bedrock (Figures 89 and 90). The upper 6 cm had been partially disturbed by apparent pig rooting and then by a fallen cycad (fadang or Cycas micronesica) during Typhoon Dolphin in May 2015. The deposit varied 21–32 cm depth over the uneven limestone, and the upper 10 cm contained sparse amounts of shell and pottery (Table 11). The pottery included one combed body fragment and six plain body fragments (Figure 91). Feature J most likely represents the ruins of a former latte structure, most convincingly seen in the displaced latte elements and diagnostic pottery. The former latte house may have been demolished and burned, resulting in the black soil in the area of the original housing footprint. The remaining stone rubble subsequently would have been pushed into a pile outside the original housing footprint. A similar pattern is noticed in other cases in the 2015 survey area, although a conclusive interpretation may require further study. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 92 Feature 02-K Feature K refers to an area of stone rubble with possible displaced latte elements. These materials are distributed over a small area of about 10 sq m near the 14 m elevation contour, about 7–8 m downslope (northeast) from the rubble of Feature J, possibly displaced from Feature J. Three of the pieces of limestone rubble show evidence of quarrying or cutting marks as expected of latte elements, but the broken pieces cannot be identified clearly as any specific forms of haligi or tasa. Figure 92. Examples of pottery fragments, collection 13 from Feature L, Site Cluster 02. Photograph at top, with scale bar in 1-cm increments. Tracings of rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 93 Feature 02-L Feature L is a midden concentration, containing broken pottery in variable density across an area of approximately 55 sq m around the 13.5 m elevation contour. Pottery was collected from a representative area of 1 by 1 m (collection 13), including one thickened rim with combing (64.6 g), four plain thickened rims (171.9 g), four body fragments with exterior combing (82.3 g), and eight plain body fragments (100.3 g) (Figure 93). The underlying sediment consisted of rocky clay silt of dark grayish brown color (Munsell color 10 YR 4/2), and a trowel probe found the limestone bedrock at less than 5 cm beneath the surface. Site Cluster 02 Summary Site Cluster 02 represents a latte period residential occupation, with a number of internally distinctive activity areas described here as Features A through L. A latte period association is strongly evidenced not only by the latte structural remains but also by the associated pottery that is diagnostic of this period. Radiocarbon dating at Feature A confirms cultural activity during the late 1600s, coincident with the Spanish-Chamorro interactions at Ritidian just prior to the forced re-location of the native Chamorro population. At least three instances of former latte structures (at Features A, H, and J) had been burned and dismantled, resulting in areas of blackened soil with adjacent piles of displaced rubble. These features possibly reflect the historically attested events of Spanish missionaries destroying the Chamorro bachelor houses (guma uritao). Other interpretations could involve other reasons for destroying a house, not necessarily due to violence but perhaps due to efforts of formalizing the end of a family’s residence, preparing a space for a new construction, or removing any uncleanliness. Four latte structures (at Features B, C, D, and G) today are found in fair to good condition. Each feature displays its own architectural style in the shapes of stone elements and overall design. The original designs can be observed from the surviving stone ruins, including some haligi still in upright position and others fallen close to their original locations. The only case of not clearly observing the original structural design is in the evidently disturbed remains of Feature D, although the individual haligi and tasa of Feature D still can be appreciated for their fine technical and artistic execution. While nearly all of the identified features relate to former latte structures and associated midden concentrations, Features E and I represent uniquely different activity areas in the forms of a set of upright boulders (Feature E) and a stone-filled platform (Feature I). Both of these features bear materials that are typical of the latte period, and their positions suggest that they functioned as parts of the surrounding latte village complex. Their precise cultural functions, however, are unknown at this time, yet the material findings indicate activities other than the ordinary scope of residential habitation seen at the other archaeological features. Midden concentrations occur primarily in association with the structural remains but also independently, found at Features A, C, D, F, H, I, J, and L. The overall pattern indicates that artifacts and food remains did not accumulate directly at the latte housing footprints, but rather the densest materials had been deposited just a few m distant from the houses. Generally low densities of materials ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 94 were found within the housing footprints, even in the areas of blackened soil where latte likely had been destroyed while some objects and cultural debris still were inside the houses. As the largest identified Site Cluster in the 2015 survey area, the findings at Site Cluster 02 may be viewed as representative of the larger survey area. The major characteristics refer to a latte period village complex, with several individual features that must have been inter-related during the last period of formal habitation in the late 1600s. The surviving components indicate that each latte house had its own architectural style and design, likely reflecting the differential preferences, skills, and other associations of the people who built or used these individual houses. Overall, though, the varied elements functioned together as parts of a village complex that can be observed and appreciated today. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 95 11. SITE CLUSTER 03 Site Cluster 03 (Figure 93) includes a cave with evidence of cultural use (Feature A) and a very large limestone boulder overhanging a midden concentration (Feature B). Both features are situated in rough limestone terrain, near the base of the steep limestone cliff at the landward (southwest) boundary of the 2015 survey area. Unlike the other identified sites in the 2015 survey area, Site Cluster 03 does not include the remains of latte structures, but the observed materials suggest cultural use during the latte period generally. Figure 93. Map of Site Cluster 03. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 96 Figure 94. Feature A cave, Site Cluster 03. View is to the west,-northwest with scale bar in 20-cm increments. Figure 95. Cypraea sp. shell in crevice at cave of Feature A, Site Cluster 03. View is down to the south-southwest, with scale bar in 20-cm increments. Feature 03-A Feature A refers to a cave in the base of the steep limestone cliff (Figure 94). One large domelike chamber has a few speleothem pillars, and a few large boulders have fallen from the ceiling onto the floor. Another smaller side chamber occurs at the east end of the main chamber. Several stalagmites and stalactites continue in active formation with dripping water. In one spot of dripping water near the northwest end of the cave, a small collection of water occasionally is visible in a shallow depression after heavy raining episodes. The shallow depression is about 10 cm in diameter and only 3 cm deep. Adjacent to this spot, a single Cypraea sp. shell was observed inside a tiny crevice (Figure 95). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 97 Near the front edge of the large dome-like chamber, a mortar depression (lusong) had been created in one of the large boulders (Figure 96). The lusong here is positioned about 1.5 m back from the dripline of the cave. This particular occurrence includes a single mortar depression, whereas most other lusong include multiple mortar depressions. Over the floor of the cave’s main chamber, fires resulting in a blackened sediment (Munsell color 10 YR 2/1). The black color suggests some kind of cultural activity, presumably related at least partially to the lusong. Nonetheless, no artifacts or food remains were visible on the surface, but instead the recent bones of at least one pig and one deer littered the northeast portion of the cave. A 50 cm by 60 cm test unit (TU-1) exposed a deposit of powdery silt over the limestone bedrock of the cave (Figures 97 and 98), containing variable amounts of marine shells and no artifacts (Table 12). Depths of the deposit varied over the uneven limestone base of the cave formation, with a maximum depth recorded at 50 cm. Although no pottery fragments or other artifacts were present in TU-1, the marine shells certainly must have been deposited by cultural activity in this location far removed from the coast. In each depth interval of TU-1, most of the shells were assorted bivalves, but the lower depths included a few limpets and Turbo sp. shells. These kinds of bivalves noticeably were uncommon or even entirely absent in the residential features elsewhere in the 2015 survey area, so their occurrence at the cave could be interpreted as reflecting a specific type of activity that did not usually occur in the residential features. The limpets in the lower depths of TU-1 potentially could reflect an older time period prior to the typical over-harvesting of these taxa in most residential sites, or alternatively they could reflect an effort to harvest from increasingly rare habitats. Figure 96. Mortar depression (lusong) in boulder at cave of Feature A, Site Cluster 03. View is down to the north, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 98 Figure 97. Excavation of TU-1 at Feature A, Site Cluster 03. View is to the north-northwest, with scale bar in 20-cm increments. Figure 98. West profile of TU-1 at Feature A, Site Cluster 03. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 99 Table 12. Summary of materials recovered from TU-1 at Feature A, Site Cluster 03. TU-1, depth (cm) 0–10 Marine shell 2 Codakia sp. shells (7.4 g) 1 Periglypta sp. shell (1.8 g) 1 Strombus sp. shell (3.4 g) 10–20 1 Codakia sp. shell (10.4 g) 1 Periglypta sp. shell (1.9 g) 20–30 2 Turbo sp. shell fragments (2.3 g) 1 limpet shell (0.7 g) 30–40 1 Codakia sp. shell (6.3 g) 1 Turbo sp operculum (3.6 g) 1 limpet shell (0.9 g) 40–50 1 Codakia sp. shell (7.1 g) 1 Periglypta sp. shell (1.1 g) 1 limpet shell (1.1 g) Feature 03-B Feature B includes a midden concentration beneath the overhang of a very large limestone boulder (Figure 99). The boulder most likely had tumbled from the nearby limestone cliff, and then the midden concentration has accumulated on the ground beneath the overhanging portion of the boulder on the downslope (east) side. Broken pottery is visible in densities of 3 to 4 pieces per sq m, including a few fragments with thickened-rim profiles and combed exteriors (Figure 100). Shell fragments include pieces of Turbo spp., Isognomon sp., and Cypraea spp. specimens. The potential for archaeological excavation was estimated as very low. Probing with a trowel in five different spots found the underlying limestone bedrock within less than 5 cm of the surface. Likewise, the surrounding terrain consists of mostly exposed rough limestone with no sedimentary accumulation. Site Cluster 03 Summary Site Cluster 03 offers evidence of cultural activities that clearly differed from the ordinary activities at the nearby residential habitations of a latte period village. The precise cultural functions are unknown at this time, but they occurred in a place that was physically separate from the inhabited village. Access was possible by hiking upslope over rough limestone terrain, over a linear distance of about 65 m from the nearest residential features at Site Cluster 02. The activities at Features A and B may or may not have related with each other, as the material remains are quite different. Activities at Feature A occurred within the context of a cave, with a single lusong grinding mortar, an absence of pottery or other artifacts, and deposition of bivalve shells that ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 100 otherwise were uncommon or absent in most residential features. Activities at Feature B occurred partially under a boulder overhang that was categorically different from the cave, with deposition of broken pottery and varied marine shells different from the findings inside the cave. If the activities at Features A and B were inter-related, then they nevertheless involved some degree of material differentiation. The kinds of activities that people performed inside the cave evidently were restricted in scope, whereas the activities outside the cave were more liberal in scope. Possibly, the more generalized-scope activities outside the cave (at Feature B) occurred before, after, or even during the specific episodes of more specialized-scope activities inside the cave (at Feature A). The precise cultural activities at Site Cluster 03 are unknown, but they definitely occurred during the latte period as generally defined. Most likely, the cave and the boulder overhang were known to the people who lived at the nearby latte village complex, such as at Site Cluster 02. Earlier cultural activities very likely occurred here, but they did not result in substantial material evidence at Site Cluster 03. Figure 99. Northwest-facing section view of Feature B, Site Cluster 03. Figure 100. Examples of pottery observed on the surface of Feature B, Site Cluster 03. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 101 12. SITE CLUSTER 04 Site Cluster 04 consists of a single isolated feature (Feature A) of blackened soil with midden and a pile of rubble (Figure 101). These materials are distributed mostly between the elevation contours of 10.5 m and 11 m. The stone rubble occurs on the downslope (north) side of the blackened soil, partially distributed below the 10.5 m elevation contour. Feature 04-A Feature A includes a midden concentration with a pile of limestone rubble (Figure 102). The midden concentration is defined by the presence of black (Munsell color 10 YR 2/1) clay silt over an area of approximately 12 m east-west by 7 m north-south, within which variable densities of pottery and shells are observable. The black soil is bordered on the downslope (north) side by a pile of limestone rubble that most likely had been cleared from the area of black soil. The black soil contains broken pottery generally in low density of 1–3 pieces per sq m, but one spot of 50 by 50 cm near the edge with the rubble (collection 01) contained much higher density of material. Recovered items include three thickened rims with combing (351.6 g), one plain thickened rim (112.8 g), three combed body fragments (147.7 g), and nine plain body fragments (262.1 g) (Figure 103). The black soil also contains scattered Turbo sp. opercula and shells, Isognomon sp. shells, and Tridacna sp. shells. Figure 101. Map of Feature A, Site Cluster 04. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 102 Figure 102. Photograph of Feature A, Site Cluster 04, showing area of black soil, with rubble on the downslope side. View is to the north, with scale bar in 20-cm increments. The limestone rubble includes at least six fragments with indications of quarrying or cutting marks, as expected of latte elements (Figures 104 and 105). A latte structure may have been dismantled in the nearby vicinity, most likely at the area of black soil. The broken pieces do not retain diagnostic indications of specific shapes of haligi or tasa, presumably no longer discernable after the breakage of the rubble. A 40 cm by 60 cm test unit (TU-1) found broken pottery, a stone flake tools, and discarded shell debris (Table 13) within a single sedimentary unit of rocky clay silt over the limestone bedrock (Figures 106 and 107). Depths of the deposit varied 20–30 cm over the uneven limestone. Cultural materials were recovered from the surface all the way through the base in contact with the limestone bedrock. From all recorded depths, broken pottery included diagnostic pieces with thickened-rim profiles and combed exteriors, typical of the latte period (Figure 108). Several Turbo sp. opercula occurred among the broken shells of Turbo sp. and Trochus sp. specimens. Site Cluster 04 Summary Site Cluster 04 most likely reflects the results of burning and dismantling of a latte structure, as seen in a number of other cases in the 2015 survey area. The burning resulted in the blackened soil with variable amounts of broken pottery, at least one stone tools fragment, and assorted shell debris. The dismantling resulted in the pile of rubble on the downslope side of Feature A, within which at least a few pieces of broken limestone resemble the remains of latte elements. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 103 Figure 103. Examples of broken pottery with thickened-rim profiles, collection 01 from Feature A, Site Cluster 04. Top shows photograph, with scale bar in 1-cm increments. Bottom shows rim profiles. Site Cluster 04 appears unusual as an instance of a single isolated feature (Feature A). The next nearest features are at distances of 60 m to the east-southeast (at Site Cluster 05) and 100 m to the southwest (at Site Cluster 02). If other features had existed at one time closer to Feature A of Site Cluster 04, then they are no longer clearly visible today. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 104 Figure 104. Probable displaced latte element within rubble of Feature A, Site Cluster 04. View is down to the northeast, with scale bar in 20-cm increments. Figure 105. Probable displaced latte element within rubble of Feature A, Site Cluster 04. View is to the southwest, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 105 Figure 106. Excavation of TU-1 at Feature A, Site Cluster 04, showing rubble based in rocky clay silt over the limestone bedrock. View is to the northwest, with scale bar in 20-cm increments. Figure 107. West profile of TU-1 at Feature A, Site Cluster 04. The apparent isolation of Site Cluster 04 could suggest that a former latte structure was used by a specific subgroup of people or for a different focused purpose than had occurred at other latte structures of the surrounding village complex. The material remains, however, suggest more or less the same scope of general residential habitation. The artifacts and shell remains are not significantly different from the findings at the other habitation features. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 106 Figure 108. Examples of broken pottery, recovered from 0–10 cm of TU-1 at Feature A, Site Cluster 04. Top shows photograph, with scale bar is in 1-cm increments. Bottom shows rim profiles. Table 13. Summary of materials recovered from TU-1 at Feature A, Site Cluster 04. TU-1, depth (cm) 0–10 Marine shell Pottery fragments 6 Turbo sp. opercula (46.8 g) 1 combed thickened rim (119.1 g) 2 Turbo sp. shell fragments (27.5 g) 8 plain thickened rims (472.5 g) 18 combed body pieces (39.1 g) 48 plain body pieces (516.8 g) 10–20/30 3 Turbo sp. opercula (15.7 g) 4 combed body pieces (63.5 g) 1 Trochus sp. shell fragment (3.3 g) 10 plain body pieces (80.4 g) ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 107 13. SITE CLUSTER 05 Site Cluster 05 consists of a set of latte remains and midden concentrations, designated as Features A through F, over an area of approximately 1650 sq m (Figures 109, 110, 111, and 112). The identifiable features occur in pockets of silty clay within an overall expanse of rough limestone terrain. All of these materials are distributed between the 7 m and 10.5 m elevation contours, about 25 m landward (south) from the inner edge of a small cove that contains the culturally useful features of a possible latte quarry (Area of Interest B) and a freshwater seep (Area of Interest D). The latte remains and other features of Site Cluster 05 were identified in November 2014, during an informal observational tour by Mr. Brian Leon Guerrero and Dr. Mike T. Carson (Figures 113 and 114). Vegetation clearing allowed fuller examination of the features in April 2015, but a much greater labor investment was required for the vegetation-clearing after Typhoon Dolphin in May 2015. Detailed recorded was possible in June 2015. Figure 109. Overall map of Site Cluster 05. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 108 Figure 110. Map detail 1 of 3, Site Cluster 05. During the June 2015 documentation, Dr. James Bayman directed a set of test excavations at Features A and B, informally known as “Brian’s Site” while awaiting the feature and site designations as reported here. This work was conducted as part of an archaeology field training program by Dr. James Bayman for students at University of Hawaii and at University of Guam. The test excavations at Site Cluster 05 therefore were more intensive than at other places in the 2015 survey area. Detailed studies of artifacts types, shell and animal bone remains, and spatial distributions of materials are underway as of July 2015 by Dr. James Bayman and his team. Feature 05-A Feature A consists of a midden concentration, identified by the presence of black (Munsell color 10 YR 2/1) clay silt over an area of 14 m east-west by 4–6 m north-south. The long axis east-west roughly follows the 10 m elevation contour. In the middle portion of the black soil area, cobble and pebble rubble has accumulated in a shallow depression of the ground surface. The west and east sides of the black soil appear to be slightly mounded. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 109 Figure 111. Map detail 2 of 3, Site Cluster 05. The surface of the black soil shows low densities of broken pottery, recorded mostly as 2–3 pieces per sq m, including pieces with thickened-rim profiles and combed exteriors as expected of the latte period. Assorted shells include Tridacna sp., Turbo sp., and Isognomon sp. specimens. Probing with a trowel did not encounter solid bedrock within the upper 10 cm of the deposit, thus indicating good potential for archaeological excavation. Controlled collections from the surface of Feature A targeted two unusual occurrences of pottery rims (collections 01 and 03) and one large piece of a broken stone adze (collection 02). The piece of stone adze had been broken from a noticeably large and heavy tool, made of dense volcanic stone (Figure 117). The recovered pottery rims included two large thickened rims, plus one rare example of an everted rim (Figures 115, 116, and 117). Controlled excavations of six formal 1 m by 1 m test units (TU-1 through TU-6) found a single cultural layer within the black clay silt overlaying the limestone bedrock. Recovered archaeological materials included broken pottery, shells, and non-human animal bones (Table 14). At least a few pieces of pottery exhibited diagnostic traits of thickened-rim profiles and combed exteriors as expected of the latte period. Depths of the deposit varied down to 30 cm over the uneven limestone, within which the lower 10–15 cm contained increasingly greater amounts of broken limestone and lesser amounts of cultural material. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 110 Figure 112. Map detail 3 of 3, Site Cluster 05. Radiocarbon dating was performed for a sample of burned material from TU-2 yielding a result within a most probable range of the 1500s through 1600s (see Table 2). The sample consisted of carbonized pieces of nutshells, most closely resembling coconut endocarp fragments, considered to be short-lived specimens dating very close to the age of the cultural deposit without any significant in-built older age of longer-lived organic specimens. The dating result accords with expectations of some generations of habitation, ending with a final residential occupation at the time of Spanish-Chamorro wars during the late 1600s. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 111 Figure 113. Initial observations by Brian Leon Guerrero at Feature B, Site Cluster 05, during November 2014. View is to southeast, with scale bar in 20-cm increments. Figure 114. Initial observations by Brian Leon Guerrero at Feature B, Site Cluster 05, during November 2014. View is to west, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 112 Figure 115. Broken pottery, 102.1 g, collection 01 from Feature A, Site Cluster 05. Scale bar is in 1cm increments. Figure 116. Fragment of volcanic stone adze, 667.3 g. Collection 02 from Feature A, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 113 Figure 117. Broken pottery, 78.4 g (upper) and 1005. G (lower), collection 03 from Feature A, Site Cluster 05. Scale bars are in 1-cm increments. Table 14. Summary of materials recovered from TU-1 through TU-6 at Feature A, Site Cluster 05. Provenience Pottery Shell Worked shell 0–10 cm 75 (412.5 g) 790 g 5.4 g 10–20 cm 16 (105 g) 75 g 20–30 cm 9 (49 g) 10 g 0–10 cm 29 (50 g) 110 g 3.4 g 10–20 cm 37 (100 g) 105 g 6.6 g 20–30 cm 4 (25 g) 30 g Animal bone TU-1 TU-2 3.4 g ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 114 Provenience Pottery Shell Worked shell Animal bone TU-3 0–10 cm 47 (220 g) 75 g 2.1 g 0.6 g 10–20 cm 16 (125 g) 25 g 21.6 g 2.8 g 20–30 cm 12 (60 g) 25 g 1g 3.5 g 0–10 cm 36 (155 g) 125 g 21.2 g 10–20 cm 27 (112.5 g) 100 g 20–30 cm 5 (3 g) 80 g 0–10 cm 53 (375 g) 75 g 10–20 cm 94 (150 g) 100 g 20–30 cm 13 (50 g) 25 g 0–10 cm 26 (30 g) 180 g 10–20 cm 3 (5 g) 120 g TU-4 TU-5 26.2 g 2.2 g TU-6 20–30 cm 75 g Feature 05-B Feature B refers to one of the more impressive latte sets in the 2015 survey area (Figures 118 and 119). The layout can be discerned of a four-paired set of haligi and accompanying tasa, including two haligi still in upright position with others fallen close to their original locations. Three of the haligi evidently had been recycled from lusong, showing the mortar depressions in their side surfaces (Figure 120). The long axis of the former latte house had been aligned northwest-southeast, between the 9 m and 9.5 m elevation contours. Along the downslope (north) edge of the former latte housing footprint, a paving alignment had been constructed of smooth-surfaced cobbles and boulders (see Figure 118). Within the footprint of the Feature B latte remains, pottery and other artifacts occur in generally low density of less than 2 items per sq m. Considerably denser occurrences of artifacts and shells are found in a midden concentration of blackened soil at the southeast side of Feature B, covering an area of approximately 25 sq m. Another dense concentration occurs with the stone rubble at the northeast corner of Feature B. Within the footprint of the former latte structure, controlled surface collections targeted the visible stone and shell artifacts (collections 04, 05, 29, and 30). Two volcanic stone tools included an abrader and a flaked piece (Figures 121 and 122). The shell specimens included a cut piece of Tridacna sp. shell possibly from adze-making and a piece of Isognomon sp. shell possibly from discard of fishhook manufacture (Figures 123 and 124). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 115 Figure 118. Latte structural remains of Feature B, Site Cluster 05. Portion of paving alignment is visible in the foreground on the downslope side of the latte remains. View is to the southsoutheast, with scale bar in 20-cm increments. Figure 119. Latte structural remains of Feature B, Site Cluster 05. View is to the east-southeast, with three scale bars in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 116 Figure 120. Haligi, recycled from a lusong at Feature B, Site Cluster 05. View is to the south, with scale bar in 20-cm increments. Figure 121. Volcanic stone abrader, 122.8 g, collection 04 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 117 Figure 122. Volcanic stone flake, 16.5 g, collection 05 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. Figure 123. Cut piece of Tridacana sp. shell, 37.2 g, collection 29 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 118 Figure 124. Isognomon sp. shell artifact, 3.8 g, collection 30 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. In the area of black clay silt (Munsell color 10 YR 2/1), controlled surface collections targeted instances of stone and shell tools (collections 07, 08, 11, and 33), as well as two representative spots of 50 cm by 50 cm of broken pottery (collections 06 and 10). The stone tools included a polished stone ball possibly used as a burnishing stone, a piece of flaked pumice, and a broken fragment of a pumice abrader (Figures 125, 126, and 127). One shell artifact resembled a v-shaped fishing gorge, made most likely of Isognomon sp. shell (Figure 128). The recovered pottery fragments included diagnostic traits of thickened-rim profiles and exterior combing, typical of the latte period generally (Figures 129 and 130). Within a patch of stone rubble at the northeast corner of Feature B, four stone tool fragments were collected (collections 09, 12, 31, and 32). The recovered items included one fragment of a stone pounder or pestle (lummok) and three pieces of pumice abraders (Figures 131, 132, 133, and 134). Given that the rubble likely had been displaced from a slightly different position at Feature B, the contexts of the stone artifacts are unclear except that they must have been used as part of the latte habitation. Another patch of rubble at Feature B did not contain any visible artifacts, but it did contain at least one disc-shaped tasa exceeding 1 m in diameter and with a roughened socket on its bottom side (see Figure 113). This rubble evidently had tumbled down the slope from an original position slightly father landward (south) and perhaps at Feature A. Other than the one apparent tasa, the other pieces of rubble did not show clear forms of specifically identifiable latte elements, although at least a few of the pieces had been broken into small chunks from larger original elements with quarrying or cutting marks on their outer surfaces. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 119 Figure 125. Polished stone ball, 117.7 g, collection 07 from Feature B of Site Cluster 05. Scale bar is in 1-cm increments. Figure 126. Flaked pumice, 41 g, collection 08 from Feature B of Site Cluster 05. Scale bar is in 1-cm increments. Figure 127. Fragment of pumice abrader, 34.3 g, collection 11 from Feature B of Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 120 Figure 128. V-shaped fishing gorge, made of Isognomon sp. shell, 0.9 g, collection 33 from Feature B of Site Cluster 05. Scale bar is in 1-cm increments. Figure 129. Examples of broken pottery, collection 06 from Feature B, Site Cluster 05. Photograph at top, with scale bar is in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 121 Figure 130. Examples of broken pottery, collection 10 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. Figure 131. Fragment of pounder, made of volcanic stone, 176.8 g, collection 09 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 122 Figure 132. Fragment of pumice abrader, 60.3 g, collection 12 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. Figure 133. Fragment of pumice abrader, 43 g, collection 31 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 123 Figure 134. Fragment of pumice abrader, 39.6 g, collection 32 from Feature B, Site Cluster 05. Scale bar is in 1-cm increments. Yet another patch of rubble is situated to the west of Feature B, covering the area between Features B and E. In this spot, cobble-sized rubble had filled into a natural depression in the rough limestone ground surface. Very sparse pottery fragments are visible here, occurring in densities far less than one piece per sq m. The observable pieces are redundant of the fragments as described for Site Cluster 05. Controlled excavations in two part of Feature B found a single cultural layer from the surface down to the limestone bedrock, varying down to 30 cm depth over the uneven limestone (Table 15). The deposits contained variable amounts of broken pottery and shells, generally decreasing in depth. Excavations entailed 1.5 sq m within the latte structural footprint (TU-8 and TU-8-ext) and another 1.5 sq m within the midden concentration of blackened soil (TU-7 and TU-9). A cache of Isognomon sp. shells was documented inside the black soil midden, within the excavation of TU-7 and TU-9 (Figure 135). Many of the shells had been partially cut, most likely in preparation for fishhook manufacture. Isognomon sp. shells appear as the most popular raw material of fishhooks generally in the latte period of the Mariana Islands and specifically at the Ritidian area. Evidence of a human burial was found within the latte structural footprint, specifically inside the excavation of TU-8 (Figure 136). A portion of a human scapula (shoulder blade) was observed, documented, and preserved in place. In accordance with the pre-established investigative procedures (outlined elsewhere in this report), the location was documented as thoroughly as possibly without any further disturbance, USFWS staff were notified, and the location has been securely preserved in place. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 124 Table 15. Summary of materials recovered from TU-8, TU-8-ext, and TU-9 at Feature B, Site Cluster 05. Provenience Pottery Shell Worked shell Human bone (preserved in place) TU-7 0–10 cm 270 (825 g) 2225 g 10 g 10–20 cm 145 (450 g) 775 g 378 g 20–30 cm 32 (120 g) 250 g 13.2 g 197 (50 g) 250 g 13.2 g 95 (154 g) 425 g 19.6 g 0–10 cm 186 (575 g) 525 g 15.1 g 10–20 cm 58 (210 g) 276 g 295 g 20–30 cm 29 (250 g) TU-8 0–10 cm Partially exposed human scapula TU-8-ext 0–10 cm TU-9 150 g Figure 135. Examples of Isognomon sp, shells from a cache in TU-7 and TU-9 at Feature B, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 125 Figure 136. Portion of human scapula (shoulder blade), partially exposed in TU-8 at Feature B, Site Cluster 05. North arrow is indicated in the image, with scale bars in 1-cm increments. Figure 137. Disturbed latte remains and broken pottery at Feature C, Site Cluster 05. View is to the northwest, with three scale bars in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 126 Feature 05-C Feature C includes the disturbed remains of a latte set and a midden concentration (Figure 137). The midden concentration contains an impressively dense occurrence of broken pottery, along with lesser amounts of shells and stone tools. The disturbed latte elements apparently had been broken from their original forms, no longer discernable as specific types of haligi or tasa, yet they retained at least portions of their original surfaces with quarrying or cutting marks. Figure 138. Adze fragment, made of volcanic stone, 76.2 g, collection 13 from Feature C, Site Cluster 05. Scale bar is in 1-cm increments. Figure 139. Flaked pumice, 138.6 g, collection 14 from Feature C, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 127 Figure 140. Examples of broken pottery, collection 15 from Feature C, Site Cluster 05. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. Controlled surface collections targeted two occurrences of stone tools (collections 13 and 14) and two spots of 50 cm by 50 cm of the densest pottery fragments (collections 15 and 16). The stone tools here included a volcanic stone adze fragment and a flaked piece of pumice (Figures 138 and 139). The pottery fragments are redundant of other findings in the vicinity, including pieces with thickenedrim profiles and combed exteriors, except that the combing here is noticed in two forms of broad-width and narrow-width combing. Pottery in collection 15 (Figure 140) included four thickened rims with ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 128 exterior broad-width combing (353.7 g), 19 plain thickened rims (1040.4 g), 41 body pieces with broadwidth combing (1118.8 g), three body pieces with narrow-width combing (311.5 g), and 53 plain body pieces (1233.4 g). Pottery in collection 16 (Figure 141) included two thickened rims with exterior broad-width combing (179.1 g), three thickened rims with exterior narrow-width combing (407.2 g), five plain thickened rims (388.5 g), two body pieces with broad-width combing (25.8 g), one body piece with narrow-width combing (50.5 g), and 17 plain body pieces (243.5 g). Figure 141. Examples of broken pottery, collection 16 from Feature C, Site Cluster 05. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 129 Among the pottery fragments, the co-occurrence of narrow-width combing and broad-width combing may deserve further study. Broad-width combing overall tends to be more popular during the later end of the latte period, whereas narrow-width combing tends to decline in popularity over time. The co-occurrence as at Feature C is a rarity for a surface-visible midden concentration, and it suggests that people in this area continued a tradition of narrow-width combing more visibly than was the case in most other places. Feature 05-D Feature D includes four patches of stone rubble, of which at least a few stones display diagnostic indications of quarried or cut surfaces as expected of latte elements. The broken fragments however do not retain enough of their original shapes to verify any specific forms of haligi or tasa. The spatial distribution suggests possible damage by any especially high storm-surge episodes, situated between 6.5 and 7.5 m elevation. Feature 05-E Feature E refers to the remains of a latte structure in good condition (Figure 142). Three haligi are visible in upright position, and other elements have fallen close to their original locations, reflecting an original construction aligned roughly with the 9.5 m elevation contour. The haligi stood at 85–95 cm height, but the accompanying tasa are no longer discernable among the nearby broken rubble. A paving alignment of cobbles and boulders had been constructed on the downslope (northeast) side of the latte housing footprint. Figure 142. Latte structural remains at Feature E of Site Cluster 05. View is to the southeast, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 130 Figure 143. Fallen haligi, revealing notched base and bracing stones, at Feature E of Site Cluster 05. View is to the southeast, with scale bar in 20-cm increments. Among the haligi elements, two observations reveal aspects of how they had been erected for achieving structural stability (Figure 143). First, the bases around the haligi had been braced by tightly packed cobbles, thus creating a firmer basal foundation. Second, the bottom ends of at least a few of the haligi had been shaped with a slight notch, so that the two-pointed bottom could provide more stability than a single-pointed or flat bottom. Within the area of the Feature E latte structural remains, controlled surface collections targeted the instances of stone and shell artifacts (collections 17, 18 19, and 20) and one large iron nail (collection 34). The stone tools included two pumice abrader fragments (Figures 144 and 145) and an adze made of dense volcanic stone (Figure 146). A single shell tool was identified as a Tridacna sp. shell adze or chisel Figure 147). The iron nail (Figure 148) potentially could have been taken from a Spanish ship or other foreign source during the late 1600s. Within and around the Feature E latte remains, extremely sparse pottery fragments were recorded in densities of far less than 1 item per sq m. Among these few pottery fragments, none of the pieces showed the diagnostic traits of thickened-rim profile or combed exterior surface. Nonetheless, the observed items were redundant of others as documented elsewhere at Site Cluster 05, including 8–11 mm thick bodies with coarse-sized temper inclusions. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 131 Figure 144. Pumice abrader, 61.4 g, collection 17 from Feature E, Site Cluster 05. Scale bar is in 1-cm increments. Figure 145. Pumice abrader fragment, 29.4 g, collection 18 from Feature E, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 132 Figure 146. Adze, made of volcanic stone, 147.6 g, collection 20 from Feature E, Site Cluster 05. Scale bar is in 1-cm increments. Figure 147. Tridacna sp. shell adze or chisel, 42 g, collection 19 from Feature E, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 133 Figure 148. Iron nail, 33.8 g, collection 34 from Feature E, Site Cluster 05. Scale bar is in 1-cm increments. Figure 149. Latte ruins at Feature F, Site Cluster 05. View is to the southwest, with three scale bars in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 134 Feature 05-F Feature F includes the remains of a latte structure and an associated midden concentration (Figure 149). In this case, the latte footprint partially overlaps with the dark-stained midden area, unlike most other occurrences in the 2015 survey area. In most other cases, the latte footprints and midden concentrations are spatially separated. The latte elements have been mostly disturbed and at least slightly displaced, yet one haligi of 80 cm height remains in its original upright position. The other elements have fallen or have been displaced and broken, in some cases found among piles of stone rubble. Nevertheless, enough of the elements are found close to their original positions to discern the original footprint of the latte between the 9 m and 9.5 m elevation contours. The midden concentration is indicated by an area of dark grayish brown clay silt (Munsell color 10 YR 4/2), measuring approximately 9.5 m northwest-southeast by 4.5 m northeast-southwest. Within the midden concentration, pottery fragments are visible in densities of 2–5 pieces per sq m. Scattered shells include Turbo sp opercula and shell fragments, as well as Tridacna sp. shells. Additionally, several stone and shell tools are found here. Controlled surface collections targeted the visible stone and shell artifacts (collections 21, 22, 24, 25, 26, 28, 35, 36, 37, and 38). Recovered stone tools included three chipped volcanic stone flakes, two pumice abrader fragments, one chert flaking core, and one broken limestone slingstone (Figures 150 through 159). Shell objects included two Tridanca sp. adze or chisel fragments and a large Cypraea sp, shell that potentially may have functioned as a part of an octopus-luring device (Figures 160, 161, and 162). Figure 150. Chipped volcanic stone flake, 99.2 g, collection 21 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 135 Figure 151. Flaked pumice, 23.5 g, collection 22 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 152. Pumice abrader, 21.3 g, collection 24 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 153. Slingstone fragment, made of shaped and polished limestone, 43.1 g, collection 25 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 136 Figure 154. Volcanic stone fragment, 123.6 g, collection 26 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 155. Chert flaking core, 33 g, collection 37 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 156. Volcanic stone flake, 2.7 g, collection 38 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 137 Figure 157. Tridacna sp. shell adze fragment, 31.6 g, collection 28 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 158. Large Cypraea sp. shell with broken hole, 73.4 g, collection 35 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 138 Figure 159. Tridacna sp. shell adze or chisel fragment, 24.9 g, collection 36 from Feature F, Site Cluster 05. Scale bar is in 1-cm increments. Figure 160. Examples of broken pottery, collection 23 from Feature F, Site Cluster 05. Photograph at top, with scale bar is in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 139 Figure 161. Examples of broken pottery, collection 27 from Feature F, Site Cluster 05. Photograph at top, with scale bar is in 1-cm increments. Rim profiles at bottom. Additional controlled surface collections targeted two 1 by 1 m areas of pottery fragments (collections 23 and 27). Collection 23 included one thickened rim with combing (49.1 g), two plain thickened rims (115.7 g), and one body fragment with a broad incision (58.9 g) (Figure 160). Collection 27 included one thickened rim with combing (60.4 g) and another plain thickened rim (57.2 g) (Figure 161). Site Cluster 05 Summary The findings at Site Cluster 05 very convincingly reflect the remains of a latte residential habitation, last occupied during the late 1600s. Test excavations confirmed a single undifferentiated cultural deposit associated with the surface-visible ruins, dated by radiocarbon most likely in a rage spanning the 1500s through 1600s. Moreover, the plentiful surface-visible latte structural ruins and other components of diagnostic latte period materials refer to the last time when the residential area had been occupied, presumably during the Spanish-Chamorro encounters of the late 1600s as attested in historical documents. Site Cluster 05 offers probably the most efficient view of a latte residential habitation within the 2015 survey area, where an appreciable number of features can be seen in good condition within a small and easily accessible space. It contains the remains of three easily discernable latte sets in close proximity to one another, among additional latte-related ruins and other features. Furthermore, the habitation can be understood in relation to the nearby occurrences of a possible latte quarry and a freshwater seep at the adjacent shoreline (Areas of Interest B and D). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 140 14. SITE CLUSTER 06 Site Cluster 06 contains three features, designated A through C, that represent habitation ruins and low-density midden concentrations (Figure 162). All of these features are spatially distinct, but they are closer to each other than they are to any other features of the next nearest Site Clusters. Additionally, these three features are distributed roughly along an apparently shared east-west axis, between the 14 m and 15.5 m elevation contours. Feature 06-A Feature A includes a sparse midden scatter, bordered by a pile of stone rubble. The midden scatter is identified by a dark grayish brown clay silt (Munsell color 10 YR 4.2), with very sparse pottery fragments measured in densities of less than one piece per sq m. One of the pottery fragments displays diagnostic traits of a thickened-trim profile and combed exterior, typical of the latte period (Figure 163). Trowel probing in three locations encountered limestone bedrock at depths of 4–6 cm, thus indicating very low potential for archaeological excavation. Figure 162. Map of Site Cluster 06. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 141 Figure 163. Fragment of combed thickened-rim pottery at Feature A, Site Cluster 06. Figure 164. Fragment of thickened-rim pottery at Feature B, Site Cluster 06. The adjacent pile of rubble occurs on the downslope (northeast) side of the midden scatter, likely removed and displaced from an original position atop the dark stained soil area. Visual inspection did reveal any clear signs of displaced latte elements as may be seen in similar features of the 2015 survey area. The possibility of a burned and dismantled latte stricture therefore is uncertain at Feature A, and other interpretations can be considered. Feature 06-B Feature B includes a very small area of 1.5 sq m with five pieces of broken pottery on a surface of rough limestone bedrock. The pottery fragments are typical of the latte period generally. One fragment displays a thickened-rim profile (Figure 164). All pieces have body thickness of 8–11 mm and coarse-sized temper inclusions. Feature 06-C Feature C consists of broken limestone rubble over an area of approximately 6 sq m on rough limestone bedrock terrain. The rather small area of only 6 sq m of rubble appears too small to account for the debris of a latte structure, yet two of the pieces of rubble exhibit surfaces with quarrying or cutting marks, suggestive of broken latte elements. The broken pieces of rubble are too fragmented to reveal clear signs of any specific forms of haligi or tasa. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 142 Site Cluster 06 Summary Site Cluster 06 reflects a zone of low-intensity cultural activity within the rough limestone terrain. Of the identifiable materials, a general latte period association can be concluded. These features most likely supported cultural activities during the time when people inhabited the surrounding latte village complex in the late 1600s or perhaps earlier. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 143 15. SITE CLUSTER 07 Site Cluster 07 includes three features, designated Features A through C, over an area of approximately 2320 sq m (Figures 165 and 166). Feature A consists of a subsurface cultural deposit that is not visible from the surface of a sandy coastal plain landform. Features B and C are surface-visible in the elevated limestone terrain, including a midden concentration (Feature B) and the remains of a latte structure (Feature C). All three features contain material with diagnostic traits of the latte period generally and consistent with findings of the 2015 survey area overall. This residential habitation evidently made use of the elevated limestone terrain with immediate access to the adjacent cove and coastal zone. The associated shoreline includes a coral limestone formation that may have been used as latte quarry, recorded as Area of Interest C (see Figure 16). A freshwater seep may have been accessible in the sandy landform of the cove, but today any such possibility has been obscured thick layers of storm-surge sand, such as the 60-cm-thick layer documented at Feature A. Feature 07-A Feature A refers to a subsurface cultural deposit, found at 60–80 cm depth near the base of a limestone cliff at the landward (south) side of a small cove (Figure 167). No artifacts, food remains, structural ruins, or other archaeological materials are visible on the surface. The setting suggests the possibility of a buried cultural layer, beneath sediments now accumulated to an elevation of 4.5 m to 5.5 m above current sea level. A 1 m by 1 m test unit (TU-1) found a single cultural layer of rocky silty sand at 60–80 cm depth. This layer contained broken pieces of pottery and shells (Table 16). The pottery was diagnostic of the latte period generally, including five thickened-rim pieces, one combed body piece, and six plain body pieces (Figure 168). The marine shells included fragments of Strombus sp., Turbo sp., and Cypraea sp. specimens. The full spatial extent of Feature A is not yet known, pending further test excavations to explore for the spatial boundary. The subsurface cultural layer may be expected to have formed over a temporarily stable natural backbeach surface during the latte period. The identified portion at TU-1 was found at 60–80 cm depth, but other portions may occur at slightly different depth according to the undulating surface of the ancient backbeach. The spatial extent most likely covered the back area of the sandy plain landform nearest the base of the limestone cliff, in a zone most safe from periodic stormsurge events. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 144 Figure 165. Overall map of Site Cluster 07. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 145 Figure 166. Detail map 1 of Site Cluster 07, showing Features B and C. Figure 167. Northwest profile of TU-1 at Feature A, Site Cluster 07. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 146 Figure 168. Broken pottery rims with thickened profiles, from 60–80 cm depth in TU-1 at Feature A, Site Cluster 07. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 147 Table 16. Summary of materials recovered from TU-1 at Feature A, Site Cluster 07. TU-1 depth (cm) Sediment Marine shell Pottery fragments 0–15 Rocky silty sand, dark brown (Munsell color 10 YR 3/3) none None 15–60 Pulverized sand, white (Munsell color 10 YR 8/2) none None 60–80 Rocky silty sand, dark gray (Munsell color 10 YR 4/1) 2 Turbo sp. opercula (10.3 g) 5 thickened rims (188.5 g) 6 Turbo sp. shell fragments (54.5 g) 1 combed body piece (6.2 g) 2 Trochus sp. shell fragments (8.1 g) 6 plain body pieces (47.3 g) none None 80–150 + Pulverized sand, white (Munsell color 10 YR 8/1) Figure 169. Pumice abrader, 36.5 g, collection 02 from Feature B, Site Cluster 07. Scale bar is in 1-cm increments. Feature 07-B Feature B is a large midden concentration, covering an area of approximately 185 sq m of black (Munsell color 10 YR 2/1) clay silt. The spatial extent is mostly between the 12.5 m and 13.5 m elevation contours, although it extends slightly downslope (northeast) and upslope (southwest). Artifacts and shells occur in variable densities across the blackened soil surface, bordered by limestone rubble on the west, and north-northeast edges. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 148 The blackened soil surface revealed artifacts and shells typical of the latte period, documented in four controlled surface collections (collections 02, 03, and 04). Collection 02 included a pumice abrader of 36.5 g (Figure 169). Collections 03 and 04 recovered pottery fragments from two different 1 m by 1 m areas with the visibly densest pottery concentrations. Collection 03 included pottery fragments of two thickened rims with combed exteriors (91.5 g), three plain thickened rims (201.2 g), one combed body piece (21 g), and one plain body piece (41.3 g) (Figure 170). Collection 04 included pottery fragments of two thickened rims (181.5 g), one combed body piece (35.4 g), one simply incised body piece (74.6 g), and one plain body piece (6.4 g) (Figure 171). Figure 170. Examples of broken pottery, collection 03 from Feature B, Site Cluster 07. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 149 Figure 171. Examples of broken pottery, collection 04 from Feature B, Site Cluster 07. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. Figure 172. Excavation of TU-2 at Feature B Site Cluster 07, showing rocky clay silt over limestone bedrock. View is to the east-southeast, with scale bar in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 150 Around the perimeter of the midden concentration, piles of rubble presumably had been cleared from the area of blackened soil. The rubble did not reveal any pieces of discernable latte elements as seen in other similar features of the 2015 survey area. Instead, the limestone rubble consisted of mixed cobbles and small boulders without any diagnostic indications of quarried or cut surfaces that otherwise could signify latte elements. Near the downslope (north) side of the blackened soil area, a 1 m by 1 m test unit (TU-2) documented a cultural deposit within the upper 10 cm of the ground surface (Figure 172). As summarized in Table 17, recovered materials included one Turbo sp. shell fragment, one Trochus sp shell fragment, eight thickened pottery rims, and 63 plain pottery body pieces (Figure 173). Beneath the 10-cm-thick zone of cultural material, the rocky clay silt continued to depths of 20–30 cm until contacting the underlying limestone bedrock yet without any cultural materials. Figure 173. Examples of broken pottery from TU-2 at Feature B, Site Cluster 07. Photograph at top, with scale bar in 1-cm increments. Rim profiles at bottom. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 151 Table 17. Summary of materials excavated from TU-2 at Feature B, Site Cluster 07. TU-2 depth (cm) 0–10 10–20/30 Marine shell Pottery fragments 1 Turbo sp, shell fragment (20.2 g) 8 plain thickened rims (339.6 g) 1 Trochus sp. shell fragment (5.6 g) 63 plain body pieces (670.1 g) None None Feature 07-C Feature C consists of the remains of a latte structure with a footprint slightly larger than most others in the 2015 survey area, at approximately 80 sq m (Figures 174 and 175). The fallen haligi and tasa suggest that the latte had been designed as a five-paired or possibly six-paired set of haligi each with an accompanying tasa. A paving alignment of cobbles had been constructed along the seaward (northwest) edge of the latte structural footprint, where the western corner stone apparently was recycled from a haligi with a notched bottom. Within the latte footprint, only extremely sparse cultural materials are visible on a surface of dark brown rocky clay silt (Munsell color 10 YR 3/3). One volcanic stone flake (collection 01) was collected near the recycled haligi in the paving alignment (Figure 176). Very few pieces of plain pottery body fragments were recorded in densities of less than 1 piece per sq m, but they exhibited the same range of 8–11 mm thick bodies and coarse-sized temper inclusions as seen in other features throughout the 2015 survey area. Scattered marine shells included Turbo sp. opercula and shells, Tridacna sp. shells, and Cypraea sp. shells. At the corner of the paving alignment, a small test unit (TU-3) confirmed very sparse cultural material in just 5 cm of rocky clay silt (Figures 177 and 178). Recovered materials included Turbo sp. opercula, Cypraea sp. shell fragments, and broken pottery (Table 18). The few pottery body fragments were consistent with findings in the surrounding area and with expectations of a latte period occupation, with traits of exterior combing, 8–11 mm thick bodies, and coarse-sized temper inclusions. The excavation found slightly more material at the upslope (southeast) side as compared to the downslope (northwest) side of the paving stones. At the middle of the latte structural footprint, a formal 1 m by 1 m test unit (TU-4) documented more of the apparently sparse cultural deposit in 8–10 cm of rocky clay silt over the limestone bedrock (Figures 179 and 180). The recovered materials were consistent with the surface-visible observations (Table 19), including Turbo sp. opercula and shell fragments, a Cypraea sp. shell fragment, one piece of a thickened pottery rim with external combing, and two pieces of plain broken pottery (Figure 181). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 152 Figure 174. Latte structural remains at Feature C, Site Cluster 07. View is to the west, with three scale bars in 20-cm increments. Figure 175. Latte structural remains at Feature C, Site Cluster 07. View is to the north, with scale bars in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 153 Figure 176. Volcanic stone flake, 59 g, collection 01 from Feature C, Site Cluster 07. Scale bar is in 1-cm increments. Figure 177. Excavation of TU-3 at Feature C, Site Cluster 07, showing the paving alignment based in shallow rocky clay silt over the limestone bedrock. Recycled haligi with notched bottom is visible in the corner of the paving alignment. View is to the southeast, with scale bar in 20-cm increments. Table 18. Summary of materials recovered from TU-3 at Feature C, Site Cluster 07. TU-3 provenience Marine shell Pottery fragments Front, 0–5 cm 1 Cypraea sp. shell fragment (1.8 g) 1 plain body piece (4 g) Back, 0–5 cm 2 Turbo sp. shell opercula (14.5 g) 1 combed body piece (12.1 g) 1 Cypraea sp. shell fragment (3.2 g) 1 plain body piece (3 g) ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 154 Figure 178. West profile of TU-3 at Feature C, Site Cluster 07. Figure 179. Excavation of TU-4 at Feature C, Site Cluster 07, showing shallow rocky clay silt over the limestone bedrock. View is to the southeast, with scale bar in 20-cm increments. Figure 180. West profile of TU-4 at Feature C, Site Cluster 07. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 155 Figure 181. Pottery fragment of thickened rim with exterior combing, recovered from TU-4 at Feature C, Site Cluster 07. Scale bar is in 1-cm increments. Table 19. Summary of materials recovered from TU-4 at Feature C, Site Cluster 07. TU-4 depth (cm) 0–10 Marine shell 3 Turbo sp. opercula (9.9 g) 1 Cypraea sp. shell fragment (2.9 g) Pottery fragments 1 thickened rim with exterior combing (61.5 g) 2 plain body pieces (7.5 g) 1 Turbo sp. shell fragment (2.9 g) Site Cluster 07 Summary Site Cluster 07 represents the remains of a latte residential habitation, situated primarily on an area of elevated limestone terrain but also involving cultural use of a patch of sandy landform in an adjacent cove. The identified components include the remains of one large-sized latte structure (Feature C), an area of blackened soil and rubble related to either a dismantled latte or another intensive cultural activity zone (Feature B), and a subsurface deposit of broken artifacts and food remains reflecting the use of a stable backbeach in the adjacent cove (Feature A). Additionally, a coral limestone formation along the nearby shoreline may have been used as a latte quarry (Area of Interest C). These feature likely functioned together as part of a cohesive habitation, evidently during the latte period as generally defined. Considering the context of the Ritidian area as a whole, the features at Site Cluster 07 reflect one of several areas of residential habitation that operated probably with a degree of autonomy or perhaps with a different nuance of their functionality as compared to others. Compared to other clustering of residential features in the 2015 survey area, the features at Site Cluster 07 are few in number and thus suggest that fewer people lived here than at other parts of Ritidian overall. Appreciably more numerous latte remains are situated more closely together at other places, such as at Site Clusters 01, 02, and 05. On the other hand, the features at Site Cluster 07 are measurably larger than those at other places in the 2015 survey area, to they may have been functioned with slightly different connotations or associations than was evident elsewhere. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 156 16. SITE CLUSTER 08 Site Cluster 08 includes four features, designated Features A through D, over an area of approximately 150 sq m (Figure 182). Feature A consists of a midden concentration of dark soil and pottery fragments. Feature B includes the ruins of a latte structure. Feature C is comprised of rubble from a badly disturbed or perhaps intentionally destroyed latte structure. Feature D refers to a midden concentration of dark soil and pottery fragments. All of these features contain material as typical of the latte period generally and consistent with findings of the 2015 survey area overall. Feature 08-A Feature A refers to a surface-visible midden concentration, covering an area of about 10 sq m. The observed materials included broken pottery in an area of dark brown rocky clay silt (Munsell color 10 YR 3/3). Probing by trowel showed that the layer of clay silt was less than 10 cm over the underlying limestone bedrock. Figure 182. Map of Site Cluster 08. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 157 Figure 183. Examples of pottery rims with corresponding profile illustrations, collection 05 from Feature A of Site Cluster 08. Scale bar is in 1-cm increments. Pottery fragments were collected from an area of 1 by 1 m, recorded as Collection 05 at this site cluster. Recovered items included five rim fragments weighing 38.7 g, including two that could be reconnected (Figure 183), plus 14 plain body pieces weighing a total 71.3 g. All of these fragments were made of coarse earthenware, showing no external combing or other decoration. The rim profiles indicated large incurving bowls, with thickened rims, typical of the latte period generally. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 158 Figure 184. Latte ruins of Feature B of Site Cluster 08, view to east-southeast. Scale bar is in 20cm increments. Figure 185. Pottery rim, collection 01 from Feature B of Site Cluster 08. Scale bar is in 1-cm increments. Feature 08-B Feature A contains the ruins of a latte set, positioned immediately landward (south) of the 14.5 m elevation contour. Although badly disturbed by tree roots and other factors, the observed remains suggested an original construction of a four-paired pillar set (Figure 184). Inspection of the ground surface and probing with a trowel in five locations showed very thin clay silt in a layer less than 5 cm over the underlying limestone bedrock. The only observed portable artifacts included a single thickenedrim potsherd (Collection 01) and a Tridacana sp. shell adze fragment (Collection 02) (Figures 185 and 186). ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 159 Figure 186. Tridacna sp. shell adze fragment, 130.1 g, collection 02 from Feature B of Site Cluster 08. Scale bar is in 1-cm increments. Figure 187. Rubble of Feature C of Site Cluster 08, including apparent displaced haligi in the foreground, view to southwest. Scale bar is in 20-cm increments. Feature 08-C Feature A is comprised of loose limestone cobble rubble, including at least two elements that resemble haligi that were displaced from latte sets (Figure 187). The rubble is distributed near the 15 m elevation contour, and the position suggests a past relationship with the latte structure of Feature B. The surrounding ground surface does not reveal any indication of burning or destroyed latte structural remains, but rather the rubble pile itself contains the only material of this feature. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 160 Figure 188. Examples of pottery rims with corresponding profile illustrations, collection 03 from Feature D of Site Cluster 08. Scale bar is in 1-cm increments. Feature 08-D Feature D refers to a midden concentration of broken pottery in an area of dark brown rocky clay silt (Munsell color 10 YR 3/3). Probing by trowel showed that the layer of clay silt was less than 5 cm over the underlying limestone bedrock. Collection 03, in an area of 1 by 1 m, produced five rim fragments of a total 45.6 g (Figure 188) and another 32 plain body pieces of a total 148.4 g. Collection 04, in another area of 1 by 1 m, produced five rim fragments of a total 208.5 g (Figure 189) with 51 plain body pieces of a total 198.9 g. All of the pottery fragments appear typical of the latte period generally, made of coarse paste and broken from large bowls with incurving thickened-rim profiles. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 161 Figure 189. Examples of pottery rims with corresponding profile illustrations, collection 04 from Feature D of Site Cluster 08. Scale bar is in 1-cm increments. Site Cluster 08 Summary Site Cluster 08 represents the remains of a latte residential habitation, with individual components situated in pockets of shallow rocky clay silt within a surrounding zone of rough limestone terrain. The features appear to be in poor condition overall, but observations were sufficient to document the existence of at least on latte structure (Feature B), one other likely destroyed or disturbed latte structure now in the form of a pile of rubble (Feature C), and two areas with dense concentrations of broken pottery (Features A and D). Given the very shallow sedimentary unit over the limestone bedrock, the surface-collected artifacts may be considered satisfactory to characterize the artifacts and confirm an association with the latte period consistent with the other findings in the 2015 survey area. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 162 17. SITE CLUSTER 09 Site Cluster 09 refers to a set of three features, designated Features A through C, over an area of approximately 80 sq m, situated between 1.5 m and 3.5 m elevation (Figure 190). All three features include limestone elements of partially cut and shaped haligi (Figures 191 through 194), most likely intended for making a latte structure that was not finished for unknown reasons. The stone quarrying may have occurred close to the shoreline in one of the “Areas of Interest” of possible quarries as noted in this report. The individual latte elements appear larger and thicker than most others that were observed in the 2015 survey area. The larger and thicker dimensions in this case may be due to the unfinished character, not yet reduced to the final size as intended. The tool marks, especially as seen in Feature A (see Figures191 and 192), suggest that the stones were in the process of being reduced into smaller size. Figure 190. Map of Site Cluster 09. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 163 Figure 191. Probable latte elements in process of working, at Feature A of Site Cluster 09, view to west. Scale bar is in 20-cm increments. Figure 192. Detail of tool marks in probable latte element in process of working, at Feature A of Site Cluster 09, view down to south. Scale bar is in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 164 Figure 193. Probable latte element in process of working, at Feature B of Site Cluster 09, view to south-southeast. Scale bar is in 20-cm increments. Figure 194. Probable latte elements in process of working, at Feature C of Site Cluster 09, view to northeast. Scale bar is in 20-cm increments. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 165 18. CONCLUSIONS AND RECOMMENDATIONS The research activities generated at least four major benefits: 1) The results contribute substantially toward an inventory and evaluation of historic resources within the jurisdiction of USFWS, in accordance with Section 110 of the National Historic Preservation Act (NHPA). 2) The results increase general archaeological knowledge within the Ritidian Unit of GNWR. 3) The resource inventory enables potential nominations in the U.S. National Register of Historic Places (NRHP) and the Guam Historic Property Inventory (GHPI). 4) The findings provide a substantive basis for ongoing development of programs for education, appreciation, and interpretation. The 2015 survey area contains the remains of several spatially distinct yet functionally interrelated features of a latte village complex, here proposed as the contributing elements of a Historic District. The full scope of the proposed Historic District likely extends beyond the eastern boundary of the 2015 survey area, where further survey can be recommended. Within the limits of the existing documentation, however, the major characteristics of the Historic District can be presented. The identified archaeological features all belong to a broadly defined latte period of A.D. 1000 through 1700, although the predominantly surface-visible and surface-related materials mostly relate to the last few generations of habitation leading up to site abandonment in the late 1600s. Limited test excavations revealed overall a single undifferentiated cultural deposit associated with the surface-visible materials. Subsurface findings were not significantly different from surface observations of artifacts and food remains, so that the archaeological materials overall offer a unified picture of the last historically known use of the latte village at Ritidian during the late 1600s. Two radiocarbon dating samples confirmed date ranges of the 1500s or 1600s. The representation of a latte village of the late 1600s can be understood in relation to the full chronological scope of habitation at Ritidian. People had lived in different parts of the Ritdian area at least since 1500 B.C. and began producing artifacts typical of the latte period as early as A.D. 1000, yet the surviving materials specifically within the 2015 survey area relate almost entirely to the late 1600s. Materials of other times periods are found in different places of the Ritidian Unit of GNWR, but the 2015 survey area offers a holistic view of a latte village and landscape that was last occupied during the late 1600s. The context of the late 1600s at Ritidian is recognized today as a key turning point in local and regional history. Spanish-Chamorro encounters of the late 1600s were complex and varied, ultimately resulting in the termination of a traditional latte village society and the beginning of a Spanishinfluenced colonial lifestyle. The Ritidian area supported one of the first Jesuit missionary outposts of the region, attested in historical documents of the 1670s. Within the next decades, after repeated episodes of warfare and probably foreign infectious disease, the local Chamorro populations throughout the Mariana Islands had suffered a very high death rate, and surviving groups were re-located by Spanish authorities into a few easily controlled villages of Guam and Saipan. By the year 1700, the latte villages such as at Ritidian had been abandoned, and the associated social systems of these habitations no longer could be sustained in a context of Spanish colonial lifestyle. The archaeological materials of a late 1600s latte village at Ritidian cohesively represent the context of the last time when such a traditional village system operated in the region. The contributing elements can be observed in their individual positions, as well as in relation to one another holistically. These contributing elements include several structural remains of latte houses, ruins of latte that apparently were burned and dismantled, and associated midden concentrations. Additional contributing elements clearly were parts of the village complex yet served functions that are not clearly specifiable ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 166 today, including a set of upright boulders, an apparently unique stone-filled platform, and a small cave that all bear evidence of cultural use contemporary with the surrounding features of the latte village complex. In addition to the clearly identifiable archaeological features, certain aspects of the natural environment may be considered as parts of the culturally inhabited landscape of the latte village context. At least three coral limestone formations along the shoreline may have been used as quarries for making the haligi and tasa of the nearby latte sets. At least one freshwater seep very likely provided an essential supply for people living at the nearby latte habitations. Other features of the natural landscape may yet be considered as culturally significant components of the Historic District of the latte village complex. For instance, traditions of a “child’s leap” are linked to the cliff overlooking Ritidian, associated with a child who leaped from there across the water channel to the next island of Rota. The limestone cliff, water sources, shoreline, and other natural features very likely were culturally significant, although further studies would be recommended to clarify how they may have related with the proposed Historic District. The contributing elements of the proposed Historic District retain integrity as archaeological ruins in terms of their design, association, and representation of the late 1600s latte village and landscape. The identified features, both individually and as a group, embody the distinctive characteristics of a type, period, and method of construction. These qualities in total convey a sense of the traditional village settlement as a significant and distinguishable entity of cultural history. The features each have contributed substantive knowledge about archaeology and history, and further research certainly can generate additional information of significance. Based on the 2015 survey findings, several research questions can be pursued, and at least three themes may be recommended with the existing materials and datasets. First, the collected artifacts and sediment samples can be examined for traces of starchy plant foods and other residues. Second, the spatial patterns within and between the archaeological features can be analyzed for comprehending the organization of cultural activity areas. Third, the architectural designs of latte structures can be examined more closely for learning about their associated engineering characteristics and artistic expressions. The 2015 survey results have generated sufficient new information to support programs of public appreciation and education. The information content, as reported here, supplies significant knowledge about the latte village complex, yet more work can be recommended to develop the archaeological information in a manner most suitable for public appreciation and outreach. Special rules apply for any activities within the Ritidian Unit of GNWR and particularly at archaeological sites, so issues of public access will need to be considered in accordance with all of these rules and relevant laws. Meanwhile, the information content can be incorporated into the existing educational programs of GNWR overall. Toward an eventual nomination of the proposed Historic District, the survey coverage can be recommended extend beyond the boundary of the 2015 survey. Meanwhile, the major characteristics of the proposed Historic District have been outlined as in this report. Sufficient information already exists for developing programs of research, resource management, and public appreciation. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 167 19. REFERENCES April, Victoriano N. 2004 Latte Quarries of the Mariana Islands. Latte: Occasional Papers in Anthropology and Historic Preservation, Number 2. Guam Historic Resources Division, Department of Parks and Recreation, Agana Heights. Bayman, James, Hiro Kurashina, Mike T. Carson, John A. Peterson, David Doig, and Jane Drengson 2012a Latte household organization at Ritidian, Guam National Wildlife Refuge, Mariana Islands. Micronesica 42: 258–273. 2012b Household economy and gendered labor in the 17th Century A.D. on Guam. Journal of Field Archaeology 37: 259–269. Carson, Mike T. 2005 Salvage of Disturbed Human Skeletal Remains at Guam National Wildlife Refuge, Ritidian, Guam. Letter report prepared for Guam National Wildlife Refuge. International Archaeological Research Institute, Inc., Honolulu. 2010 Radiocarbon chronology with marine reservoir correction for the Ritidian Archaeological Site, northern Guam. Radiocarbon 52: 1627–1638. 2011 Palaeohabitat of first settlement sites 1500–1000 B.C. in Guam, Mariana Islands, western Pacific. Journal of Archaeological Science 38: 2207–2221. 2012a An overview of latte period archaeology. Micronesica 42: 1–79. 2012b Evolution of an Austronesian landscape: The Ritidian Site in Guam. Journal of Austronesian Studies 3: 55–86. 2013 Austronesian migrations and developments in Micronesia. Journal of Austronesian Studies 4: 25–52. 2014a Learning about latte at Ritidian. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 96–104. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. 2014b Archaeological resource survey at entrance road and parking lots in Ritidian Unit of Guam National Wildlife Refuge. Report prepared for U.S. Fish and Wildlife Service. Richard F. Taitano Micronesian Area Research Center, University of Guam, Mangilao. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 168 2014c Contexts of natural-cultural history: A 3500-year record at Ritidian in Guam. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 1–43. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. 2014d Natural-cultural landscape heritage at Ritidian, Guam. Journal of Cultural Heritage Management and Sustainable Development 4: 35–56. 2014e Paleo-terrain research: Finding the first settlement sites of Remote Oceania. Geoarchaeology 29: 268–275. 2014f Micronesian archaeology. In The Global Prehistory of Human Migration, edited by Peter Bellwood, pp. 314–319. Wiley Blackwell, West Sussex. 2014g First Settlement of Remote Oceania: Earliest Sites of the Mariana Islands. Springer, New York. Carson, Mike T., and Hsiao-chun Hung 2015 On the Beach in Remote Oceania. In Field Archaeology from around the World: Ideas and Approaches, edited by Martin Carver, Bisserka Gaydarska, and Sandra Monton-Subias, pp. 13 – 136. Springer, New York. Carson, Mike T., and Hiro Kurashina 2012 Re-envisioning long-distance Oceanic migration: Early dates in the Mariana Islands. World Archaeology 44: 409–435. Carson, Mike T., and John A. Peterson 2011 Calcrete formation and implications for buried archaeological deposits in the Mariana Islands, western Pacific. Geoarchaeology 26: 501–513. 2012 Radiocarbon dating of algal bioclasts in beach sites of Guam. Journal of Island and Coastal Archaeology 7: 64–75. Carson, Mike T., Hsiao-chun Hung, Glenn Summerhayes, and Peter Bellwood 2013 The pottery trail from Southeast Asia to Remote Oceania. Journal of Island and Coastal Archaeology 8: 17–36. Carson, Mike T., Boyd Dixon, John A. Peterson, and Andrea Jalandoni ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 169 2014 Site definitions in a complex archaeological landscape: An example at Ritidian, Guam. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 105–112. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. de Freycinet, Louis Claude des Sausles 2003 An Account of the Corvette l’Uranie’s Sojourn at the Mariana Islands, 1819. Translated and edited by Glynn Barratt. Originally published as Volume II of Voyage Autour du Monde, Executé Corvettes de S.M. l’Uranie et la Physicienne, pendant les Années 1817 à 1820 and Journal de Voyage autour du Monde à bord de l’Uranie 1817 à 1820. Occasional Historical Papers No. 13. Commonwealth of the Northern Mariana Islands Division of Historic Preservation, Saipan. Dixon, Boyd 2000 Archaeological survey and testing at proposed brown tree snake pen, Guam National Wildlife Refuge, Ritidian Point, Territory of Guam, Mariana Islands. Report prepared for U.S. Geological Survey. International Archaeological Research Institute, Inc., Honolulu. Dixon, Boyd, Richard Schaefer, and Todd McCurdy 2011 Level 1 archaeological reconnaissance of 800 acres within selected properties on Andersen Air Force Base, Guam: Volume 1 narrative. Report prepared for SEARCH Inc. Cardno-TEC Inc., Honolulu. Dixon, Boyd, Andrea Jalandoni, Erik Lash, Richard Schaefer, and Trina Meisner 2013 Proposed Guam and CNMI military relocation 2012 roadmap adjustments, SEIS, live-fire training range complex footprint, access routes, and main cantonment alternatives, in-fill survey and testing. Report prepared for U.S. Naval Facilities Engineering Command. Cardo-TEC Inc., Honolulu. Fritz, Georg 1904 Die Chamorro. Ethnologisches Notizblatt 3 (2): 25–110. 1984 Die Chamorros: A History and Ethnography of the Marianas. Translated by Elfriede Craddock. Working Papers 45. Rchard F. Taitano Micronesia Area Research Center, University of Guam, Mangilao. Hornbostel, Hans 1925 Unpublished notes and catalogs. Library Department, Bernice P. Bishop Museum, Honolulu. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 170 Hung, Hsiao- chun, Mike T. Carson, Peter Bellwood, Frediliza Campos, Philip J. Piper, Eusebio Dizon, Mary Jane L. A. Bolunia, Marc Oxenham, and Zhang Chi 2011 The first settlement of Remote Oceania: The Philippines to the Marianas. Antiquity 85: 909–926. le Gobien, Charles 1700 Histoire des Îles Marianes, Nouvellement Covnertés à la Réligion Chréstienne, et du Martyre des Premiers Apôtres qui y on Presché la Foy. Nicholas Pepie, Paris. Jalandoni, Andrea 2011 Casa real or not real? A Jesuit mission house in Guam. M.A. thesis. University of the Philippines, Diliman. 2012 The casa real site in Ritidian, northern Guam: An historical context. Philippine Quarterly of Culture and Society 39: 27–53. 2014 Conflict at contact: Late 17th Century Spanish missions and la reduccion in northern Guam. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 53–63. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. Kurashina, Hiro (editor) 1990 Archaeological investigations at the Naval Facility (NAVFAC) Ritidian Point, Guam, Mariana Islands. Report prepared for U.S. Department of the Navy. Richard F. Taitano Micronesian Area Research Center, Mangilao, and Applied Research Group, Bernice P. Bishop Museum, Honolulu. Laguana, Andrew, Hiro Kurashina, Mike T. Carson, John A. Peterson, James M. Bayman, Todd Ames, Rebecca A. Stephenson, John Aguon, and Harya Putra 2012 Estorian I latte: A story of latte. Micronesica 42: 80–120. Lévesque, Rodrigue (editor) 1995 History of Micronesia: A Collection of Source Documents, Volume 6: Revolts in the Marianas, 1673–1678. Les Éditions Lévesque, Gatineau, Québec. Liston, Jolie 1996 The legacy of Tarague Embayment and its inhabitants, Andersen AFB, Guam, Volume I: Archaeology. Report prepared for 36 CES/CEV, Unit 14007, Environmental Flight, Andersen Air Force Base. International Archaeological Research Institute, Inc., Honolulu. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 171 Madrid, Carlos 2014 Ritidian and Songsong across the Rota Channel: Historic significance and culture of historic preservation. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 44–52. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. Osborne, Douglas 1947a Chamorro archaeology. Unpublished manuscript on file. Richard F. Micronesian Area Research Center, University of Guam, Mangilao. 1947b Archaeology on Guam: A progress report. American Anthropologist 49: 518–524. Peterson, John A. 2014 Community archaeology at Ritidian: University of Guam archaeological field schools and contributions to community history and cultural resource management. In Guam’s Hidden Gem: Archaeological and Historical Studies at Ritidian, edited by Mike T. Carson, pp. 84–95. British Archaeological Reports, International Series 2663. Archaeopress, Oxford. Randall, Richard H. 1990 General description of environment. In “Archaeological investigations at the Naval Facility (NAVFAC) Ritidian Point, Guam, Mariana Islands,” edited by Hiro Kurashina, pp. 40–46. Report prepared for Department of the Navy. Richard F. Taitano Micronesian Area Research Center, Mangilao, and Applied Research Group, Bernice P. Bishop Museum, Honolulu. Reed, Erik K. 1952 General report on archaeology and history of Guam. Report prepared for Governor of Guam. U.S. National Park Service, Washington, D.C. Reinman, Fred M. n.d. Preliminary report, National Science Foundation, Grant G5-662: Notes on an archaeological survey of Guam, Mariana Islands, 1976–66. Unpublished document on file at Micronesian Area Research Center, University of Guam, Mangilao. 1968 Guam prehistory. In Prehistoric Culture in Oceania: A Symposium, edited by Ichiro Yawata and Yosihiko H. Sinoto, pp. 41–50. Bishop Museum Press, Honolulu. 1974 Archaeological site survey records: Guam Register of Historic Places inventory data forms. Unpublished documents on file. Guam Historic Preservation Office, Agana. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 172 1977 An Archaeological Survey and Preliminary Test Excavations on the Island of Guam, Mariana Islands, 1965–1966. Richard F. Taitano Micronesian Area Research Center, University of Guam, Mangilao. Rogers, Robert F. 1995 Destiny’s Landfall: A History of Guam. University of Hawaii Press, Honolulu. Spoehr, Alexander 1957 Marianas Prehistory: Archaeological Survey and Excavations on Saipan, Tinian and Rota. Fieldiana: Anthropology 48. Chicago Natural History Museum, Chicago. Stephenson, Rebecca A. 1990 Ethnohistorical background. In “Archaeological investigations at the Naval Facility (NAVFAC) Ritidian Point, Guam, Mariana Islands,” edited by Hiro Kurashina, pp. 18–35. Report prepared for Department of the Navy. Richard F. Taitano Micronesian Area Research Center, Mangilao, and Applied Research Group, Bernice P. Bishop Museum, Honolulu. Thompson, Laura 1932 Archaeology of the Mariana Islands. Bernice P. Bishop Museum Bulletin 100. Bishop Museum Press, Honolulu. 1940 The function of latte in the Marianas. Journal of the Polynesian Society 49: 447–465. U.S. Fish and Wildlife Service 2009 Guam National Wildlife Refuge Comprehensive Conservation Plan. Report prepared for government management and public reference. Region 1 of U.S. Fish and Wildlife Service, Portland. Walth, Cherie K. (editor) 2014 Archaeological monitoring and data recovery, Guam National Wildlife Refuge, septic system rehabilitation project, Yigo, Guam. Report prepared for Techni-con Inc. SWCA Environmental Consultants, Barrigada, Guam. ARCHAEOLOGICAL RESOURCE SURVEY, EASTERN PORTION OF RITIDIAN UNIT, GNWR, 2015 173