Ecohydrology & Hydrobiology 15 (2015) 160–170

Contents lists available at ScienceDirect

Ecohydrology & Hydrobiology

journal homepage: www.elsevier.com/locate/ecohyd

Original Research Article

Effects of riparian forest management on Chilean mountain

in-stream characteristics
Giovany Guevara a,b,c,*, Roberto Godoy a, Pascal Boeckx d, Carlos Jara e,
Carlos Oyarzún a
a
Instituto de Ciencias Ambientales y Evolutivas, Facultad de Ciencias, Universidad Austral de Chile, Campus Isla Teja, Valdivia, Chile
b
Departamento de Desarrollo Rural y Recursos Naturales, Facultad de Ciencias Agropecuarias, Universidad de Caldas, Manizales, Colombia
c
Present address: Departamento de Biologı´a, Facultad de Ciencias, Universidad del Tolima, Ibagué, Colombia
d
Laboratory of Applied Physical Chemistry – ISOFYS, Faculty of Bioscience Engineering, Ghent University, Gent, Belgium
e
Instituto de Ciencias Marinas y Limnológicas, Facultad de Ciencias, Universidad Austral de Chile, Campus Isla Teja, Valdivia, Chile

A R T I C L E I N F O A B S T R A C T

Article history: We assessed the effect of forest management practices on temporal differences in riparian
Received 18 December 2014 litterfall input and autumnal leaf litter processing of 3 evergreen plant species, using
Accepted 23 July 2015 independently 2 mesh sizes (for two of the three species), between two contrasting
Available online 1 August 2015 headwater streams draining (a) near-pristine and (b) harvested evergreen old-growth
forests in Chile (398 S). Additionally, we compared rainfall, streamflow, suspended
Keywords: sediment concentration, and the in-stream invertebrate community, collected two years
Coarse particulate organic matter after logging in the harvested microcatchment. The harvested forested stream had higher
Litter decomposition
runoff, suspended sediment concentration and input of litter (3377 kg ha1 yr1) than the
Headwaters
pristine stream (3166 kg ha1 yr1). However, leaf-litter decay rates were faster for the
Andes
BMPs
pristine (ranging from k = 0.0120 to 0.0024 day1) than for the harvested forest stream
(k = 0.0049 to 0.0023 day1) depending on leaf species and litterbag type. Likewise, the
pristine stream revealed higher shredder taxa richness than harvested stream, suggesting
a higher correlation with the decomposition rate and leaf mass losses. Therefore, the
results of our study suggest that despite the implementation of best management
practices on riparian forests, there are detectable shifts in the structure and functioning of
aquatic invertebrate community in the Andean streams of south-central Chile.
ß 2015 European Regional Centre for Ecohydrology of the Polish Academy of
Sciences. Published by Elsevier Sp. z o.o. All rights reserved.

1. Introduction of which 60–80% have been replaced with exotic tree

plantations (Altieri and Rojas, 1999). These activities have
The forest land cover in south-central Chile (39–558 S) altered the natural riparian vegetation of most Chile’s
supports a large timber industry (Echeverrı́a et al., 2006). It aquatic systems, but their effects have not been totally
has been estimated that Chilean old-growth native forests addressed by scientific studies.
are being logged at a rate of 120 000–200 000 ha per year, Forest management activities in headwater areas can
potentially degrade water quality and the physical habitats
within streams (Carroll et al., 2004), affecting woody
debris, leaf litter and sediment supplies (Webster et al.,
* Corresponding author. Present address: Departamento de Biologı́a,
Facultad de Ciencias, Universidad del Tolima, Barrio Santa Elena Parte
1990; Gurnell et al., 1995; Giller and O’Halloran, 2004),
Alta, A.A. 546, Ibagué, Colombia. Tel.: +57 8 2771212. modifying physicochemical conditions and invertebrate
E-mail address: ggcolombia@gmail.com (G. Guevara). diversity (Danger and Robson, 2004; Davies et al., 2005),

http://dx.doi.org/10.1016/j.ecohyd.2015.07.003
1642-3593/ß 2015 European Regional Centre for Ecohydrology of the Polish Academy of Sciences. Published by Elsevier Sp. z o.o. All rights reserved.
 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170 161

and altering hydrology, nutrient fluxes and productivity streams and the slowest in disturbed ones. We also
(Campbell and Doeg, 1989; Lockaby et al., 2004; Lin and expected benthic invertebrate assemblages to change as a
Wei, 2008). Therefore, proper forestry best management consequence of forest harvesting, with less shredder taxa
practices (BMPs, sensu Aust and Blinn (2004)) should be in managed forest streams than in pristine forested
implemented to minimize effects on streams. Timber streams. The outcome of this study will be helpful for
harvesting has the potential to affect the quantity, quality, the management of riparian forest ecosystems in Chile,
and timing of allochthonous inputs from riparian vegeta- where such studies are scarce. The objectives of the present
tion (Campbell et al., 1992; Delong and Brusven, 1994; investigation were to (a) assess the temporal pattern of
Xiong and Nilsson, 1997), with concomitant effects on leaf riparian litterfall in two evergreen forested streams
litter processing and aquatic invertebrate communities located in the Andean mountains, and (b) compare leaf
(Lockaby et al., 2004; Davies et al., 2005). litter decomposition in streams draining natural and
Riparian litterfall is widely recognized as an extremely managed forests in south-central Chile, linking forest
important structural and functional component of headwa- management to stream consumers and organic matter
ter streams (Benfield, 1997; Webster et al., 1999; Abelho, processing.
2001; Graça and Canhoto, 2006). Within undisturbed forest
ecosystems, litter from the riparian areas constitutes a key
2. Methods
energy source for stream biota (Wallace et al., 1997; Gessner
and Chauvet, 2002; Benfield, 2006), affecting food chains, 2.1. Study area
nutrient cycling and productivity (Cummins et al., 1989;
Boulton and Boon, 1991; Wantzen et al., 2008). Aquatic The study was conducted in two streams located at San
fungi and leaf-eating invertebrates (shredders) are the main Pablo de Tregua Experimental Forest, in the Andes
organisms assimilating the energy from litter in low-order mountains, South-Central Chile (398380 S, 728050 W, 600–
streams (Abelho and Graça, 1998; Graça, 2001; Graça et al., 1600 m a.s.l.; Fig. 1). The area consists of typical old-
2001). growth mixed evergreen (e.g., Laureliopsis philippiana
Although litter decomposition has been widely studied Looser, Dassyphyllum diacanthoides (Less.) Cabr., Saxe-
in both aquatic and terrestrial environments around the gothaea conspicua Lindl), and deciduous forests (e.g.,
world, few studies have examined leaf litter decomposi- Nothofagus alpina (P. et E.) Oerst; Schlegel and Donoso,
tion in headwater streams in a riparian silvicultural 2008). Soils are deep to moderately Andisols derived from
management context within South American temperate recent coarse volcanic ash (Veblen et al., 1996) and
old-growth rainforests (Valdovinos, 2001; Albariño and pumicitic material, overlying andesitic and basaltic lavas
Balseiro, 2002). However, this information is important to and fluvio-glacial sediments (Newton et al., 2007). The soil
understand how forest management practices influence show high carbon storage capacities in relation with soil
stream trophic dynamics (Guevara et al., 2009). age relative to the evolution of the mineral phase
In Chile there are important natural private reserves (Neculman et al., 2013). The region has a humid, cool
(Holmes, 2015; Rivera and Vallejos-Romero, 2015) such as temperate climate, with short and dry summers with an
San Pablo de Tregua (398 S, Panguipulli Municipality) with average maximum temperature of 20 8C in February and a
relevant and unique biodiversity and native forest. In this minimal mean temperature of 5 8C in August, even with
private protected area, researches and data collection snow events (Lara et al., 2002). Mean annual precipitation
started in 2002 to establish the effects of timber harvesting is ca. 5000 mm (Neira, 2005), concentrated in autumn and
on water quality and quantity and on other biogeochemi- winter. Monthly average discharge of the three streams
cal factors (e.g., forestry, soil dynamics, terrestrial and and precipitation for the years 2003–2007 are shown in
aquatic invertebrates, biodiversity; see Redel et al., 2008; Fig. 2.
Schlegel and Donoso, 2008; Lara et al., 2009; Oyarzún et al., One of the streams runs through undisturbed natural
2011) as a representative (montane) area of the Chilean evergreen forest (hereafter pristine) with predominance
Andes. The hydrobiological research proposals were of Myrceugenia planipes (H. et A.) Berg, Laureliopsis
established following the watershed-ecosystem approach philippiana and Saxegothaea conspicua. Another stream
of the Hubbard Brook Experimental Forest in the White runs through a harvested evergreen forest (hereafter
Mountains of New Hampshire (United States) under a LTER harvested), originally with the same tree composition as
perspective (Likens, 2013). To determine the effect of the ‘‘pristine’’ stream, but subjected to selective logging
forestry activities on structural and functional role of 60 years ago. The forest has undergone natural post-
benthic invertebrates, previous experimental trials were harvest succession, but there was a change in its
carried out on selected microcatchments. Seasonal litter- composition and structure due to the incorporation of
fall input, mixed (summer 2007, January to March) and the companion old-growth tree, Dassyphyllum dia-
single leaflitter decomposition, and colonization by canthoides (Redel et al., 2008). In November 2006, this
invertebrates during autumn 2007 (March to June) were catchment was thinned extracting 40% of the total basal
registered (sensu Guevara et al., 2009). In this paper we area through a BMPs program during forest harvesting
show additional results obtained from that experiment. (i.e., selective cutting, skidding was done by ox dragging
We hypothesized that riparian management practices and residues were left on site and a 15-m riparian buffer
would affect the litterfall supply to the stream channel and, zone was preserved). Some characteristics of the two
consequently, in-stream litter decomposition, with the forests and streams studied are given in Table 1 (see more
fastest decomposition rates in pristine forest headwater details in Oyarzún et al., 2011).
162 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170

Fig. 1. Location of selected experimental streams in the San Pablo de Tregua Reserve, Región de Los Rı́os, south-central Chile. Adapted from several sources.

Fig. 2. Rainfall and runoff monthly mean values measured from April 2003 to November 2007 at two contrasting catchments in temperate rainforest
ecosystems of south-central Chile.
 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170 163

Table 1
Characteristics of two contrasting forest stands located near first order streams in south-central Chile and mean litterfall dry mass (SD) recorded during
four years (October 2002–September 2006), prior to thinning experimental trial.

Forest stand Altitude Slope Tree Density Height Age Basal area Litterfall
(m) (%) comp. (tree ha1) (m) (yr) (m2 ha1) (kg ha1 yr1)

Pristine 805 12 Sc/Lp 267 45 >200 91.3 4811 (850)

Harvested 734 26 Lp/Sc/Mp 873 43 >200 119 4535 (411)

Sc: Saxegothaea conspicua, Lp: Laureliopsis philippiana, Mp: Myrceugenia planipes.

2.2. Runoff and suspended sediment concentration assigned to a functional feeding group following Anderson
and Rosemond (2007). Due to our knowledge, there are no
Runoff was recorded in V-notch 908-section gauges other similar studies in Chile, we compare our results with
following standard procedures during four hydrological those registered by Valdovinos (2001) in a study related to
years (April 2003 to November 2007). In the same notches, riparian leaf litter processing in a woodland stream. The
water levels were measured with an hourly resolution by ratio between k-values for harvested and pristine (kh:kp)
DIVER pressure sensors and recorded by data loggers. and for the coarse and fine mesh bags (kc:kf), which has
Precipitation was monthly recorded during April 2003 to been suggested as a predictor of functional stream
March 2007 with a HOBO tipping bucket rain gauge in a integrity (Gessner and Chauvet, 2002), was also calculated
nearby meadow. Runoff water samples were monthly for the streams. All those results compliment the
collected in the streams adjacent to the water level information previously reported by Guevara et al. (2009).
recorders. Two water samples were collected at each
stream in clean polyethylene 1000 ml bottles that were 2.4. Data analyses
washed with water from the stream before sampling. The
samples were returned to the laboratory, filtered and the Differences in litterfall and sediment yield between
filters dried at 105 8C for 24 h for to determine suspended sites were assessed by Kruskal–Wallis non-parametric test
sediment concentration (g L1). since data were not normally distributed and attempts to
normalize the data by transformations were unsuccessful.
2.3. Litterfall, leaf decomposition and benthic invertebrates Decomposition rates (k, day1) per treatment (plant
species, stream and mesh size) were calculated as the
We compare the litterfall input per trap obtained from slope of a linear regression of the natural log of the
two sets of six traps on each side of the stream, along a 120- percentage of remaining mass against time in days. The
m stretch during December 2006 to November 2007 at an time (days) required for 50% and 95% mass loss was
attempt to detect the effect of harvest on the organic calculated as t50 = 0.6931/k and t95 = 3/k, respectively
matter input in each stream. Data were later pooled to (Olson, 1963). Two-way analysis of covariance (ANCOVA)
determine coarse differences between pristine and har- of the percentage of dry weight remaining after each
vested microcatchment. period of time, with time as a covariate, was used to detect
Single leaf decomposition assay and leaf mass loss of differences in decomposition rates between treatments
selected plant species (M. planipes, L. philippiana and S. (factors: leaf species; litter bag type). One-way ANCOVA
conspicua) followed the litterbag standard procedures was used to compare leaf mass loss from S. conspicua
(Bärlocher, 2005). Air dried leaves (10 g) were enclosed (factor: stream; covariate: time). Overall analyses were
in 20 cm  20 cm coarse- (5 mm mesh) and fine-mesh done following confirmation of statistical assumptions.
nylon bags (0.5 mm mesh, to exclude invertebrates) and Normality was confirmed using Shapiro–Wilk test, where-
incubated in both streams during autumn (21 March to 20 as residuals were plotted to assess inequality of variance.
June 2007). Twenty-four bags per plant species and per Data were log or arcsine-transformed when they did not
mesh size were randomly placed in each stream and 4 meet parametric statistical assumptions (Zar, 1999). All
random replicates per treatment were retrieved after 3, 7, analyses were performed using Statistica 7 (StatSoft, 2004)
15, 30, 45, and 60 days. Litter was removed from each with the level of significance set at 0.05.
sampled bag, rinsed with water in order to eliminate
inorganic particles and oven dried at 70  5 8C during 48 h.
3. Results
Because S. conspicua tree has small leaf size (<4 mm width),
only fine bags were used for that species. 3.1. Runoff and suspended sediment concentration
In each stream, the benthic invertebrate community
was evaluated on December 2008 along a 120-m reach There were not significant differences between streams
with a 30 cm  30 cm Surber sampler (250-mm mesh size; when considering both variables precipitation and runoff.
five replicate samples per stream). In addition, on both However, in the harvested stream a significant difference
streams a hand net qualitative sampler (250-mm mesh was detected during the calibration period (i.e., first year
size) was used to capture invertebrates with low densities. after harvesting). It also was observed that there was an
Invertebrates were sorted from the samples and stored in average increase in total annual streamflow between 4.4
70% alcohol. They were identified to the lowest practical and 7.8% during the first and second year after thinning,
taxonomic level (Fernández and Domı́nguez, 2001) and respectively for the harvested microcatchment compared
164 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170

Fig. 3. Suspended sediment concentration at contrasting catchments with evergreen old-growth riparian forest in the Andean mountains, south-central
Chile. Mean values (g L1  SE) were higher for the harvested than pristine stream (0.79  0.11 vs. 0.45  0.06) with significant differences between them
(Kruskal–Wallis test, H(1,86) = 7.11, P < 0.01).

to pristine one. Higher rainfall and runoff mean values mass in the fine mesh bags. On day 60, in most leaf species,
occur between May and June during the evaluated period 40  5.3% of the leaf mass remained in both coarse and fine
(Fig. 2). The harvested catchment revealed significantly bags with the exception of L. philippiana (coarse bag). The
higher concentration of suspended sediments than pristine largest mass loss in fine mesh bags was recorded for M.
forest stream (Fig. 3). Air temperature and precipitation planipes in the harvested forest stream. However, the leaf
values were conversely related (Fig. 4a). Precipitation mass loss remained >40% in both the coarse and fine bags
during the study period was similar to 2003–2006 until 60 days had passed. The estimated mean time to achieve
variations registered for the area (see Fig. 2). 50 and 95% mass loss for all plant species was higher in the
harvested than in the pristine forest streams and, in fine than
3.2. Riparian litterfall dynamics in coarse mesh bags, with the exception of M. planipes
incubated in fine mesh bags (Table 2).
Litterfall consisted primarily of leaves (78%) in both Leaf breakdown rates from both L. philippiana and M.
streams. Overall, the highest litterfall input occurred from planipes plant species differed (P < 0.0001) among coarse
April to August 2007 (excluding June and July because an and fine bags but only decomposition rate of L. philippiana
extreme snowfall event collapsed the litter traps; Fig. 4b). was significantly higher in the pristine than in the
Litterfall peaked in August in the pristine and harvested harvested forest stream (Table 3). Nevertheless, decompo-
forests (Fig. 4b). However, no significant differences among sition rate of M. planipes incubated in coarse bags was
the monthly organic matter production of these riparian higher in the pristine than harvested streams (Table 2). The
forests were observed (Kruskal–Wallis test, H(2,30) = 0.245, decomposition rate of S. conspicua leaves did not differ in
P = 0.885). Estimations of annual riparian litterfall inputs relation to streams (Table 3). The functional scores, derived
were 3166 and 3377 kg ha1 yr1 for pristine and har- from the k-ratios (kh/kp and kc/kf), indicated that functional
vested forests, respectively (Fig. 4c). integrity ranged from ‘‘not compromised’’ to ‘‘severely
compromised’’ in relation to pristine and harvested forest
3.3. Leaf decomposition streams (Table 2).

Leaf decomposition curves followed a similar pattern 3.4. Invertebrate composition

after seven days of incubation timing mainly between
broad-leaved plants into fine litter bags (Fig. 5). Most Invertebrate taxa richness was higher in the pristine
curves were significant (P < 0.05) and had r2 values 0.63 than the harvested forest stream. Shredders in the
(Table 2). Leaf mass loss rates were as follow: L. pristine forest stream corresponded to 50% of all
philippiana > M. planipes > S. conspicua. In all cases, a taxa, whereas in the harvested forest stream these
higher mass loss from the coarse than from the fine invertebrates represented only 36% of the benthic
meshed bags was observed, and loss was higher in the community. The pristine forest stream invertebrate
pristine than in the harvested forest stream (Table 2, community comparatively included a higher diversity
Fig. 5). With the exception of S. conspicua, <45 days were of functional feeding groups (FFGs) and a higher richness
needed for the leaf mass to fall below 80% of the initial leaf of Ephemeroptera, Plecoptera, Trichoptera taxa. These
 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170 165

Fig. 4. (a) Temperature and precipitation per month, (b) variation of litterfall production per riparian traps, (c) monthly litterfall mass production per ha in
each forested stream. Bars indicate mean  SE.

results contrast with those registered in forested streams inputs among these streams is likely related with the
of central Chile (Table 4). proper best management practices applied during the
harvesting of the forest i.e., trees near to stream banks
remained uncut. Consequently, riparian trees can continue
4. Discussion
to provide litterfall directly to the stream channel (Lockaby
4.1. Litterfall patterns et al., 2004). That situation contrasts with the fact that leaf
litter inputs to streams in heavily logged lands are usually
Total riparian litter inputs can be considered as an much lower than inputs to streams in undisturbed or
estimation of the potential stock of organic matter pristine forests (Webster et al., 1990). In addition,
reaching the stream from terrestrial sources (allochtho- differences in litter inputs to streams between disturbed
nous supply). This input was lower than the average input and undisturbed catchment areas seem to be related to the
measured during four years in adjacent upland areas (see replacement of mature vegetation by successional species
Table 1). The total annual litterfall mass was not (Delong and Brusven, 1994). Our prediction that silvicul-
significantly different when pristine and harvested forest tural management would affect the riparian litter input to
streams were compared. The lack of a difference in litterfall streams was partially supported, because monthly litterfall
166 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170

did not significantly differ among streams, although a

slightly higher input to the harvested forest stream was
registered (see Fig. 4b). However, those values should be
regarded with caution, because the inter-annual average
litterfall in the pristine forest stand was higher than in the
harvested forest stand, despite tree density and basal area
differences between these microcatchments (see Table 1
and Staelens et al., 2011). Tree species with large basal area
are expected to contribute more to litterfall biomass than
trees with small basal area (Benfield, 1997).
The pattern of riparian litterfall observed in this study
is broadly comparable with that observed in other
southernmost Chilean temperate areas with evergreen
forests, having a continuous litter input throughout the
year but with pulses of higher litterfall in early mid-
autumn (April to May), and a marked decrease during
spring (September to December) (Perez et al., 2003).
Accuracy of estimated annual litter production from this
study is limited because there is only one year of monthly
collections and due to the missing of two-month litterfall
collections (Fig. 4b). The annual litterfall estimate of
3166 kg ha1 yr1 from our study for the pristine forest is
lower than that recorded for the evergreen Nothofagus
betuloides (Mirb.) Oerst (7412 kg ha1 yr1) in Antillanca
valley, Chilean Andes (408470 S, 800 m a.s.l.) (Leiva and
Godoy, 2001). These differences could indicate a higher
litterfall input to the forest floor than to the stream
channel in headwater Andean catchments, contrasting
with that suggested from studies around the world, where
higher litterfall inputs would be expected to occur from
riparian zones (Xiong and Nilsson, 1997). Nevertheless,
values were closer to those from a New Zealand catchment
with evergreen Nothofagus fusca (Hook F) Oerst and
Nothofagus menziesii (Hook F) Oerst (mean annual input
4927 kg ha1 yr1) (Sweetapple and Fraser, 1992). Total
Fig. 5. Leaf mass loss (mean  SE, n = 4) of selected evergreen riparian trees litterfall is also similar to that reported, in comparable
in small streams of south-central Chile during autumn 2007. Filled symbols
studies, in temperate forests of other countries (Abelho,
(* !) indicate coarse litter bags, open symbols fine litter bags. In the lower
panel open symbols indicate forest stream. 2001; O’Keefe and Naiman, 2006).

Table 2
Exponential breakdown rates (k, mean  1SE, n = 4) of three leaf litter species and litter bags with 2 different mesh sizes as estimated by exponential decay
model after 60 days incubation in two headwater streams of south-central Chile during the study period.

Leaf species Stream Bag Exponential regression k ratiosa

2 1
R F(1,4) P k (days  SE) 50% 95% kh/kp kc/kf Criterion
remaining remaining
(days) (days)

L. philippiana Pristine C 0.9693 126.24 0.0004 0.0120  0.0040 58 250 0.4083 3.1579 Bad
Harvested C 0.8043 16.44 0.0154 0.0049  0.0034 141 612

M. planipes Pristine C 0.6327 6.89 0.0585 0.0072  0.0084 96 417 0.5972 1.3611 Good
Harvested C 0.3758 2.40 0.1956 0.0043  0.0081 161 698

L. philippiana Pristine F 0.8496 22.59 0.0089 0.0038  0.0030 182 790 0.9474 Good
Harvested F 0.7475 11.84 0.0263 0.0036  0.0025 193 833 2 Moderate

M. planipes Pristine F 0.4820 3.72 0.1259 0.0036  0.0060 193 833 1.1944 Good
Harvested F 0.3924 2.58 0.1833 0.0043  0.0067 161 698 1 Moderate

S. conspicua Pristine F 0.6558 7.62 0.0051 0.0024  0.0014 289 1250 0.9583 Good
Harvested F 0.8854 30.91 0.0508 0.0023  0.0008 301 1304
a
Based on the work by Gessner and Chauvet (2002) where good criterion indicates ‘‘not compromised functional integrity’’, moderate criterion indicates
‘‘compromised functional integrity’’ and bad criterion indicates ‘‘severely compromised functional integrity’’. C = coarse, F = fine litterbag. Decomposition
rate for leaf species in harvested (kh) and pristine (kp) sites and, coarse-to-fine ratio (kc/kf) with values ordered as first pristine and then harvested stream.
 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170 167

Table 3
Results of the analysis of covariance tested on leaf mass losses from three evergreen plant species incubated in coarse and fine bags (Litter Bags= LB) in two
contrasting forested streams of south-central Chile during autumn 2007.

Plant species Factor SS DF MS F P

Laureliopsis philippiana Time 78.190 1 78.190 239.678 <0.0001

Stream 5.510 1 5.510 16.891 <0.0001
LB 23.404 1 23.404 71.740 <0.0001
Stream*LB 3.227 1 3.227 9.891 0.0022
Error 29.687 91 0.326

Myrceugenia planipes Time 48.226 1 48.226 59.886 <0.0001

Stream 0.023 1 0.023 0.029 0.8649
LB 15.925 1 15.925 19.775 <0.0001
Stream*LB 3.488 1 3.488 4.332 0.0402
Error 73.282 91 0.805

Saxegothaea conspicua Time 8.217 1 8.217 103.95 <0.0001

Stream 0.005 1 0.005 0.07 0.7986
Error 3.557 45 0.079

Table 4
Comparison list of taxa registered inside the experimental leaf packs exposed to colonization by benthic invertebrates in the Rucue stream (central Chile)
and in a rapid survey in two forested streams (south-central Chile), during December 2008.

Taxa FFG Rucue streama (central Chile) San Pablo de Treguab

Nothofagus pumilio Pinus radiata Pristine Harvested

(deciduous) (coniferous)

Annelida Hirudinea par-pred X

Oligochaeta gath, pred X X

Arthropoda Hydracarina pred X X

Amphipoda Hyalella simplex gath

Coleoptera Psephenidae sc X
Elmis sp. sc X X

Diptera Chironomini gath X

Orthocladius spp. gath X X X
Tanypodinae pred X
Hemerodromia sp. pred X
Simulium sp. fil X
Hexatoma sp. gath, pred
Tipula sp. gath, sh X

Ephemeroptera Meridialaris spp. sc X X

Meridialaris diguilina gath X
Baetis sp. gath X X

Plecoptera Antarctoperla michaelseni gath, sh X X X X

Limnoperla jaffueli gath X X X
Diamphipnoa helgae sh X X
Diamphipnopsis samali sh X
Perlidae pred

Trichoptera Smicridea chilensis fil X X X

Phylloicus sp. sh X X
Sericostoma sp. sh X X
Triplectides sp. sh X

Megaloptera Protochauliodes sp. pred X X

Sialis sp. pred X

Odonata Libellulidae pred X

Aeshna sp. pred X

Gastropoda sc

Nematomorpha Gordius? gath X

FFG refers to functional feeding group: par = parasite, pred = predator, gath = collector-gather, fil = collector-filterer, sc = scraper and sh = shredder; adapted
from Anderson and Rosemond (2007).
Sources:
a
Valdovinos (2001).
b
This study.
168 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170

Despite the similarity in total litterfall production resulting from timber logging (Lockaby et al., 2004; Young
between pristine and harvested streams, the composition et al., 2008), would in part explain our findings.
of the litterfall (i.e., leaves, sticks, seeds, etc.) may vary Sedimentation is expected to reduce leaf litter breakdown
annually in the streams (Campbell et al., 1992; Webster by burying the detritus and also to negatively affect
et al., 1999; Staelens et al., 2011). In addition, a sensitive invertebrate taxa, such as shredders. Evaluation
combination of environmental factors, such as wind- of the abundance of leaf-eating invertebrates over
storms, seasonal droughts during summer and snow decomposition time should be a priority for further
events during winter can alter the flux of litterfall to studies. Previous observations of invertebrate colonization
streams throughout the year in the study area. Riparian (Guevara, pers. obs.) showed that larger invertebrates were
evergreen forested streams draining the Andean Cordillera almost never found in fine-mesh bags, indicating that fine-
are susceptible to heavy snow events mainly during late mesh bags effectively excluded the most important
autumn and winter when branch and tree falling shredders. Only smaller chironomids were observed in
commonly supply pulses of coarse organic matter to these fine bags in the harvested stream. Attributing the
streams. However, these kinds of inputs are not adequately leaf-eating invertebrate presence in the streams to
captured by the methods used in this study. differential decomposition patterns in fine vs. coarse bags
would imply that high shredder abundances relative to
4.2. Leaf decomposition patterns fastest decomposition rates in the pristine stream are a
consequence of these patterns.
The harvested forest stream had a taxonomic richness The time to achieve 50% and 95% mass loss was also
similar to that reported in a previous leaf breakdown study variable among streams and was leaf species-dependent.
developed in central Chile (Valdovinos, 2001). The For example, considering only the coarse mesh bag
hypothesis that leaf processing would be faster in the treatment, the species that came to lose between 40 and
pristine than in the harvested forest stream was only 80% of initial mass during the 60 days experiment. This
supported by the differences in the decomposition rates for indicates that for a 95% or even 50% threshold to be used as
L. philippiana and M. planipes within coarse bags (see a benchmark in southern Andean streams, pristine and
Table 3). Decomposition rates depend on several factors, harvested sites would have to be monitored over longer
such as plant species (i.e., their physical and chemical periods, which might not be practical for biomonitoring
characteristics) (Sabetta et al., 2006) and the presence of (Bergfur et al., 2007) or for seasonal variation studies, as
leaf-eating invertebrates (Albariño and Balseiro, 2002). In evergreen plant species evaluated here would need more
our study, decomposition rate varied with leaf species and than a season (>90 days) to lose 95% of their initial mass
stream type (pristine vs. harvested). This variability could (Fig. 5).
be the result of the in-stream heterogeneous decomposer In our study, we explored the utility of the measures of
(microorganisms) and detritivores distribution. Consistent functional integrity on basis of silvicultural intervention
with the fastest decomposition rate in the pristine stream, effects. However, we found that the criterion interpreta-
we found a diverse invertebrate shredder assemblage tions from the k ratios were inconsistent among species
inhabiting this stream, which should be related to highest and ratio (kh/kp and kc/kf) and therefore, these might
leaf mass losses into coarse bags. These results were become an inaccurate indicators of stream integrity for our
similar to that observed in coarse litterbags during the study of Andean streams. In fact, k-ratios varied markedly
study period (Guevara, pers. obs.), where a great abun- with leaf species and indicated exactly contrary to
dance of shredders, mainly belonging to the EPT orders, expected i.e., criterion for pristine stream indicated
was registered. The number of other invertebrates severely compromised its functional integrity and vice
(collector–gatherers, collector–filterers, scrapers, and pre- versa (see Table 2). Hence, not only bag mesh size but also
dators) was consistently low across streams; however, the leaf species used in the analysis seemed to affect the
Odonates were more common in the harvested stream results. Further analysis is required to determine the
suggesting a higher ‘‘predation effect’’ on other local sensitivity of this metric in assessing the effect of
invertebrates. silvicultural and land use changes on ecosystem attributes
The higher decomposition rate recorded for M. planipes in relatively little impacted streams in the Chilean Andes as
in fine bags in the harvested than in the pristine stream can suggested by several authors (Pascoal et al., 2003; Castela
be due to higher physical abrasion associated with higher et al., 2008; Young et al., 2008). Likewise, it is possible that
flow velocities (discharge) in this stream (see Fig. 2). A other ecosystem-level processes should be explored to
possible higher density of microbes also would have determine those that can serve as good indicators of
increased decomposition rates in this stream but further functional integrity (Gessner and Chauvet, 2002).
analyses are required to test the role of bacteria and fungi
on litter processing differences in our studied streams. 4.3. Effects of silviculture on stream hydrobiology
A comprehensive study of the macro- and micro-
organisms associated with the decaying leaves was beyond Headwater streams transport small volumes of water
the scope of this paper, but such study is necessary to and consequently are more susceptible to environmental
explain in more detail the differences for the breakdown changes (Hagan et al., 2006). In the study area, the
rates between the contrasting evergreen streams. It is harvested watershed yielded higher stream discharge
possible that the significant higher sediment inputs to the values compared to the pristine catchment during two
harvested forest stream, which is the most likely stressor years observation period (December 2005 to January 2007;
 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170 169

Fig. 2). These differences between watershed treatments project. G. Guevara thanks to Universidad Austral de Chile
coincided with the differences observed in the study by (MECESUP UCO0214 – AUS0703) and CONICYT (TT-
Oyarzún et al. (2008), during a four year study period. It has 23090247) for the doctoral scholarships.
been reported that following forestry activities, biota are
affected by the high erosion and sedimentation, due in part
to changes in hydrological regimes, e.g. increased stream References
flow peaks (Hutchens and Wallace, 2002; Hutchens et al.,
Abelho, M., 2001. From litterfall to breakdown in streams: a review. Sci.
2004). Considering the significant differences in the
World 1, 656–680.
suspended sediment concentration between harvested Abelho, M., Graça, M., 1998. Litter in a first-order stream of a temperate
and pristine streams, we expected that litter breakdown deciduous forest (Margaraça Forest, central Portugal). Hydrobiologia
rates would be negatively affected by increasing sediment 386, 147–152.
Albariño, R.J., Balseiro, E.G., 2002. Leaf litter breakdown in Patagonian
supply (Young et al., 2008), in the harvested stream. streams: native versus exotic trees and the effect of invertebrate size.
Similarly, shredder assemblage also would suffer altera- Aquat. Conserv.: Mar. Freshw. Ecosyst. 12, 181–192.
tions in richness, density, biomass or replacement by more Altieri, M.A., Rojas, A., 1999. Ecological impacts of Chile’s neoliberal
policies, with special emphasis on agroecosystems. Environ. Dev.
sediment-tolerant species. Compensatory effects within Sustain. 1, 55–72.
species assemblages can occur when a stress factor affects Anderson, C.B., Rosemond, A., 2007. Ecosystem engineering by invasive
some types of organisms involved in the processes but not exotic beavers reduces in-stream diversity and enhances ecosystem
function in Cape Horn, Chile. Oecologia 154, 141–153.
others (Gessner and Chauvet, 2002). Forestry operations Aust, W.M., Blinn, C.R., 2004. Forestry best management practices for
near the channel in the harvested stream should cause high timber harvesting and site preparation in the eastern United States:
rainfall or snowmelt runoff, erosion, sedimentation an overview of water quality and productivity research during the
past 20 years (1982–2002). Water Air Soil Pollut.: Focus 4, 5–36.
increase, and flooding, and destroy vital habitat for Bärlocher, F., 2005. Leaf mass loss estimated by litterbag technique. In:
invertebrates and functional attributes within aquatic Graça, M.A.S., Bärlocher, F., Gessner, M.O. (Eds.), Methods to Study
system (Lockaby et al., 2004). As a result, local substratum Litter Decomposition: A Practical Guide. Springer, Berlin, pp. 37–42.
Benfield, E.F., 1997. Comparison of litterfall input to streams. Stream
characteristics (Veblen et al., 1996) and invertebrate taxa
organic matter budgets. J. N. Am. Benthol. Soc. 16, 104–108.
richness and/or density could be changed. This suggests Benfield, E.F., 2006. Decomposition of leaf material. In: Hauer, F.R.,
that higher supply of fine sediments to headwater streams Lamberti, G.A. (Eds.), Methods in Stream Ecology. Academic Press,
may explain in part the slowest leaf litter decomposition San Diego, CA, pp. 711–720.
Bergfur, J., Johnson, R.K., Sandin, L., Goedkoop, W., Nygren, K., 2007.
under silvicultural treatment. Effects of nutrient enrichment on boreal streams: invertebrates, fungi
and leaf-litter breakdown. Freshw. Biol. 52, 1618–1633.
Conflict of interest Boulton, A.J., Boon, P.I., 1991. A review of the methodology used to
measure leaf litter decomposition in lotic environments: time to turn
over an old leaf? Aust. J. Mar. Freshw. Res. 42, 1–43.
Part of our data, corresponding to results of the PhD Campbell, I.C., Doeg, T.J., 1989. Impact of timber harvesting and produc-
dissertation from the first author (Giovany Guevara). This tion on streams: a review. Aust. J. Mar. Freshw. Res. 40, 519–539.
Campbell, I.C., James, K.R., Hart, B.T., Devereaux, A., 1992. Allochthonous
paper in its early version was presented at the Workshop coarse particulate organic material in forest and pasture reaches of
on Bilateral (International) and Technological Coopera- two south-eastern Australian streams. I. Litter accession. Freshw. Biol.
tion between Flanders and Chile held in Valdivia (Chile), 27, 353–365.
Carroll, G.D., Schoenholtz, S.H., Young, B.W., Dibble, E.D., 2004. Effec-
16–18th December 2008. The special issue of the tiveness of forestry streamside management zones in the sand-clay
workshop papers was published on 2009 in a proceedings hills of Mississippi: early indications. Water Air Soil Pollut.: Focus 4,
book (https://biblio.ugent.be/input/download?func= 275–296.
Castela, J., Ferreira, V., Graça, M.A.S., 2008. Evaluation of stream ecological
downloadFile&recordOId=1154616&fileOId=1154625#
integrity using litter decomposition and benthic invertebrates. Envi-
page=42). Otherwise, none declared. ron. Pollut. 153, 440–449.
Cummins, K.W., Wilzbach, M.A., Gates, D.M., Perry, J.B., Taliaferro, W.B.,
1989. Shredders and riparian vegetation. Bioscience 39, 24–30.
Ethical statement
Danger, A.R., Robson, B.J., 2004. The effects of land use on leaf-litter
processing by macroinvertebrates in an Australian temperate coastal
None declared. stream. Aquat. Sci. 66, 296–304.
Davies, P.E., Cook, L.S.J., McIntosh, P.D., Munks, S.A., 2005. Changes in
stream biota along a gradient of logging disturbance, 15 years after
Acknowledgements logging at Ben Nevis, Tasmania. Forest Ecol. Manage. 219, 132–148.
Delong, M., Brusven, M., 1994. Allochthonous input of organic matter
from different riparian habitats of an agriculturally impacted stream.
We are grateful to C. Aguilera who assisted during Environ. Manage. 18, 59–71.
fieldwork. An opening MS draft was highly enriched for the Echeverrı́a, C., Coomes, D., Salas, J., Rey-Benayas, J.M., Lara, A., Newton,
comments of M.A.S. Graça, R. Albariño and the Stream and A.C., 2006. Rapid deforestation and fragmentation of Chilean Tem-
perate Forests. Biol. Conserv. 130, 481–494.
Riparian Research Laboratory team (University of British Fernández, H.R., Domı́nguez, E. (Eds.), 2001. Guı́a para la determinación
Columbia). We also deeply thank Managing Editor and two de los artrópodos bentónicos sudamericanos. Universidad Nacional
anonymous reviewers for your helpful suggestions on an de Tucuman, Tucuman – Argentina.
Gessner, M.O., Chauvet, E., 2002. A case for using litter breakdown to
earlier draft of the manuscript. Persisting mistakes are assess functional stream integrity. Ecol. Appl. 12, 498–510.
responsibility of authors. Giller, P.S., O’Halloran, J., 2004. Forestry and the aquatic environment:
studies in an Irish context. Hydrol. Earth Syst. Sci. 8, 314–326.
Funding body Graça, M.A.S., 2001. II. Leaf litter processing and invertebrates. The role of
This research was partially funded by grants from the invertebrates on leaf litter decomposition in streams – a review. Int.
Rev. Hydrobiol. 86, 383–393.
DIDUACh (D-2007-05), RLB (RLB08-AT04), and CONICYT Graça, M.A.S., Canhoto, C., 2006. Leaf litter processing in low order
(AT-24080035) by means support for doctoral thesis streams. Limnetica 25, 1–10.
170 G. Guevara et al. / Ecohydrology & Hydrobiology 15 (2015) 160–170

Graça, M.A.S., Ferreira, R.C.F., Coimbra, C.N., 2001. Litter processing along Oyarzún, C.E., Godoy, R., Staelens, J., Donoso, P.J., Verhoest, N.E.C., 2011.
a stream gradient: the role of invertebrates and decomposers. J. N. Seasonal and annual throughfall and stemflow in Andean temperate
Am. Benthol. Soc. 20, 408–420. rainforests. Hydrol. Processes 25, 623–633.
Guevara, G., Godoy, R., Boeckx, P., Jara, C., Oyarzún, C., 2009. Leaf litter Oyarzún, C.E., Godoy, R., Verhoest, N.E.C., 2008. Interactions between
dynamics in headwater streams of the Chilean Andes: influence of hydrology and biogeochemical cycles of native vegetated mountain
shredders and silvicultural activities. In: Verhoest, N.E.C., Boeckx, P., watersheds under different management southern Chile. In:
Oyarzún, C., Godoy, R. (Eds.), Ecological Advances on Chilean Tem- Bruthans, J., Kovar, K., Hrkal, Z. (Eds.), Proceedings of the International
perate Rainforests. Academia Press, Valdivia, Chile, pp. 41–54. Interdisciplinary Conference on Predictions for Hydrology, Ecology,
Gurnell, A.M., Gregory, K.J., Petts, G.E., 1995. The role of coarse woody and Water Resources Management: Using Data and Models to Benefit
debris in forest aquatic habitats: implications for management. Society – HydroPredict’2008. Czech Association of Hydrogeologists,
Aquat. Conserv.: Mar. Freshw. Ecosyst. 5, 143–166. Prague, Czech Republic, pp. 49–52.
Hagan, J.M., Pealer, S., Whitman, A.A., 2006. Do small headwater streams Pascoal, C., Pinho, M., Cássio, F., Gomes, P., 2003. Assessing structural and
have riparian zone defined by plant communities? Can. J. Forest Res. functional ecosystem condition using leaf breakdown: studies on a
(Rev. Can. Rech. For.). 36, 2131–2140. polluted river. Freshw. Biol. 48, 2033–2044.
Holmes, G., 2015. Markets, nature, neoliberalism, and conservation Perez, C.A., Armesto, J.J., Torrealba, C., Carmona, M.R., 2003. Litterfall
through private protected areas in southern Chile. Environ. Plan. A dynamics and nitrogen use efficiency in two evergreen temperate
47, 850–866. rainforests of southern Chile. Austral Ecol. 28, 591–600.
Hutchens, J.J., Batzer, D.P., Reese, E., 2004. Bioassessment of silvicultural Redel, Y., Rubio, R., Godoy, R., Borie, F., 2008. Phosphorus fractions and
impacts in streams and wetlands of the Eastern United States. Water phosphatase activity in an Andisol under different forest ecosystems.
Air Soil Pollut.: Focus 4, 37–53. Geoderma 145, 216–221.
Hutchens, J.J., Wallace, J.B., 2002. Ecosystem linkages between Southern Rivera, C., Vallejos-Romero, A., 2015. La privatización de la conservación
Appalachian headwater streams and their banks: leaf litter break- en Chile: repensando la gobernanza ambiental. Bosque (Valdivia) 36,
down and invertebrate assemblages. Ecosystems 5, 80–91. 15–25.
Lara, A., Altamirano, A., Thiers, O., Tacon, A., 2002. Plan de manejo CIPMA- Sabetta, L., Zaccarelli, N., Mancinelli, G., Madrone, S., Salvatori, R., Con-
FNAM, unidad demostrativa piloto predio San Pablo de Tregua. Uni- stantini, M.L., Rossi, L., 2006. Mapping litter decomposition by re-
versidad Austral de Chile, Facultad de Ciencias Forestales, Valdivia, mote-detected indicators. Ann. Geophys. 49, 219–226.
Chile. Schlegel, B.C., Donoso, P.J., 2008. Effects of forest type and stand
Lara, A., Little, C., Urrutia, R., McPhee, J., Álvarez-Garretón, C., Oyarzún, C., structure on coarse woody debris in old-growth rainforests in
Soto, D., Donoso, P., Nahuelhual, L., Pino, M., Arismendi, I., 2009. the Valdivian Andes, south-central Chile. Forest Ecol. Manage.
Assessment of ecosystem services as an opportunity for the conser- 255, 1906–1914.
vation and management of native forests in Chile. Forest Ecol. Manag. Staelens, J., Ameloot, N., Almonacid, L., Padilla, E., Boeckx, P., Huygens, D.,
258, 415–424. Verheyen, K., Oyarzún, C., Godoy, R., 2011. Litterfall, litter decompo-
Leiva, J., Godoy, R., 2001. Production and decomposition of litterfall in sition and nitrogen mineralization in old-growth evergreen and
Nothofagus forest in southern Chile. In: De Schrijver, A., Kint, V., Lust, secondary deciduous Nothofagus forests in south-central Chile. Rev.
N. (Eds.), Proceedings of the Workshop Comparison of Ecosystems Chil. Hist. Nat. 84, 125–141.
Functioning and Biogeochemical Cycles in Temperate Forest in South- StatSoft, 2004. STATISTICA (Data Analysis Software System).
ern Chile and Flanders. Academia Press, Gent, Belgium, pp. 103–106. Sweetapple, P.J., Fraser, K.W., 1992. Literfall from a mixed red beech
Likens, G.E., 2013. Biogeochemistry of a Forested Ecosystem. Springer- (Nothofagus fusca)-silver beech (Nothofagus menziesii) forest, central
Verlag, New York. North Island, New Zealand. N. Z. J. Bot. 30, 263–269.
Lin, Y., Wei, X., 2008. The impact of large-scale forest harvesting on Valdovinos, C., 2001. Riparian leaf litter processing by benthic macro-
hydrology in the Willow watershed of Central British Columbia. J. invertebrates in a woodland stream of central Chile. Rev. Chil. Hist.
Hydrol. 359, 141–149. Nat. 74, 445–453.
Lockaby, B.G., Governo, R., Rummer, B., Colson, C., 2004. Silvicultural Veblen, T.T., Donoso, C., Kitzberger, T., Robertus, A., 1996. Ecology of
management within streamside management zones of intermittent southern Chilean and Argentinean Nothofagus forest. In: Veblen, T.T.,
streams: effects on decomposition, productivity, nutrient cycling, and Hill, R., Read, J. (Eds.), The Ecology and Biogeography of Nothofagus
channel vegetation. South. J. Appl. For. 28, 211–224. Forest. Yale University Press, Yale, pp. 293–353.
Neculman, R., Rumpel, C., Matus, F., Godoy, R., Steffens, M., de la Luz Mora, Wallace, J.B., Eggert, S.L., Meyer, J.L., Webster, J.R., 1997. Multiple trophic
M., 2013. Organic matter stabilization in two Andisols of contrasting levels of a forest stream linked to terrestrial litter inputs. Science 277,
age under temperate rain forest. Biol. Fertil. Soils 49, 681–689. 102–104.
Neira, E., 2005. Producción de agua en dos microcuencas con diferente Wantzen, K.M., Yule, C.M., Mathooko, J.M., Pringle, C.M., 2008. Organic
cobertura de renovales de Nothofagus en la cordillera de Los Andes. matter processing in tropical streams. In: Dudgeon, D. (Ed.), Tropical
(San Pablo de Tregua)Facultad de Ciencias Forestales, Universidad Stream Ecology. Academic Press, London, pp. 43–64.
Austral de Chile, Valdivia, Chile. Webster, J.R., Benfield, E.F., D’Angelo, D.J., Peters, G.T., 1990. Effects of
Newton, A.C., Echeverria, C., Gonzalez-Espinosa, M., Williams-Linera, B., forest disturbance on particulate organic matter budgets of small
Ramirez-Marcial, N., Thiers, O., Armesto, J.J., Aravena, J.C., Lara, A., streams. J. N. Am. Benthol. Soc. 9, 120–140.
2007. Testing forest biodiversity indicators by assessing anthropo- Webster, J.R., Benfield, E.F., Ehrman, T.P., Schaeffer, M.A., Tank, J.L., Hutch-
genic impacts along disturbance gradients. In: Newton, A.C. (Ed.), ens, J.J., D’Angelo, D.J., 1999. What happens to allochthonous material
Biodiversity Loss and Conservation in Fragmented Forest Landscapes: that falls into streams? A synthesis of new and published information
The Forest of Montane Mexico and Temperate South America. CAB from Coweeta. Freshw. Biol. 41, 687–705.
International, Wallingford, United Kingdom, p. 432. Xiong, S., Nilsson, C., 1997. Dynamics of leaf litter accumulation and its
O’Keefe, T.C., Naiman, R.J., 2006. The influence of forest structure on effects on riparian vegetation: a review. Botan. Rev. 63, 240–264.
riparian litterfall in a Pacific Coastal rain forest. Can. J. Forest Res. Young, R.G., Matthaei, C.D., Townsend, C.R., 2008. Organic matter break-
(Rev. Can. Rech. For.). 36, 2852–2863. down and ecosystem metabolism: functional indicator for assessing
Olson, J.S., 1963. Energy storage and the balance of producers and river ecosystem health. J. N. Am. Benthol. Soc. 27, 605–625.
decomposers in ecological systems. Ecology 44, 322–331. Zar, J.H., 1999. Biostatistical Analysis. Prentice Hall, New York.