ACTA ZOOL. MEX. (ns) , 10.

1985

NICHE CONVERGENCE IN THE

DESERT RODENTS OF TWO GEOGRAPHICALLy
ISOLATED COMMUNIl"IES

KA Rogovin·
A.V. Surov·
V. Serrano··

*The Severstov Institute of Animal

Evolutionary, Morphology and Ecology,
USSR Academy of Sciences,
Moscow 117071, Leninsky 33

** Instituto de Ecología, A.C.

Apartado Postal 18-845
C.P. 11800, México, D.F.

RESUMEN

Usando el método de análisis de conglomerados se realizó

una clasificación ecológica para dos comunidades de roedores geográficamente aisla-
dos, una en Asia Central (trans-Altai Gobi, República Popular de Mongolia) y la otra en
Norte América (Desierto Chihuahuense, Mapimí, México). El espacio ecológico de cada
una de las comunidades fue determinado por medio de la técnica de ordenación polar
(Bray y Curtis, 1957). Para la determinación de la dieta de los roedores se analizaron con-
tenidos estomacales tomando en cuenta, sólo grandes grupos de alimento. En la conduc-

This paper was realized wilhin Ihe ProjecI "Mammal, Ecologyand

Elology" (PCECNAL-800790), under Ihe Binalional Agreemenl for Scienlific and Technic Coopera-
lion Mexico - USSR Program, Nalional Council of Science and Technology (CONACyT) and Ihe
USSR Sciences Academy.
 Acta Zool. Mex. (na). 1Q. 1985

ta exploratorioa se utilizó la técnica de "Open field" (Eisenberg, 1963, 1967). Estos tipos
de conductas se relacionaron con el uso del espacio (microhábitat). Las especies geo-
gráficamente aisladas fueron en todos los casos agrupadas a un nivel de semejanza me-
nor que las relacionadas geográficamente. Sin embargo, las especies asiáticas fueron
más especializadas en el uso del microhábitat y en su alimentación. La distribución de es-
pecies en el espacio ecológico de Trans Altai Gobi fue generalmente más uniforme, to-
mando en cuenta que el espacio ocupado fue tan grande como en Mapim í. De los dos gru-
pos de especies que se encontraron en la comunidad del Desierto Chihuahuense (Mapi-
mí), uno estuvo integrado por Heterómidos y el otro por Cricétidos norteamericanos. La di-
ferencia que se encontró en el espacio ecológico de las comunidades de Trans Altai Gobi
y del Desierto Chihuahuense fue debido básicamente a un grado diferente de divergencia
ecológica entre los roedores saltadores bipedos y cuadrúpedos. La riqueza específica en
ambas comunidades fue similar. La comunidad de Trans Altai Gobi tiene una fuerte diver-
gencia de especies y estuvo compuesta por nuevé géneros y diez especies (Euchoreutes
naso, Sa/pingotus crassicauda, AI/actaga sibirica, AI/actaga taliae, Dipus sagitta, Mus
muscu/us, Cricetu/us migratorius, Rhodopus rovorovskii, Meriones meridianus, Rhom-
bomys opimus). En el Desierto Chihuahuense, estuvo compuesta por siete géneros y on-
ce especies (Spermophi/us spilosoma, Spermophi/us variegatus, Perognathus pencil/a-
tus, P. flavus, P. ne/soni, Dipodomys merriami, D. ne/soni, Neotoma a/bigu/a, Onichomys
torridus, Peromyscus eremicus y Sigmodon hispidus). Los trabajos más recientes realiza-
dos por los ecólogos rusos sobre convergencia ecológica en roedores del desierto y este-
pas se han basado en el concepto de "formas de vida" (Kaskarov 1938, 1944: Formozov
1950, 1956; Khodashova 1953: Bannikov 1954). Este concepto refleja el fenómeno de
evolución paralela (convergencia) de algunos caracteres en animales que se encuentran
en condiciones ambientales similares. Para cada una de las "formas de vida", las caracte-
rísticas morfológicas, que tienen una significancia adaptativa fueron consideradas pri-
mordiales. Diferentes autores han llegado a conclusiones basandose en diferente carac-
teres, y consecuentemente la clasificación ecológica final ha sido frecuentemente un tan-
to subjetiva. La razón de estas diferencias es que el significado del término "forma de vida"
ha sido tratado de manera diferente (Krivolutsky 1967). Con el desarrollo de la tecnología
de Taxonomía Numérica (Sokal y Sneath 1963; Smirnov 1969; Jardine y Sibson 1971),
las clasificaciones ecológicas pueden considerar al mismo tiempo un gran número de pa-
rámetros, más o menos relacionados ecológicamente, y una estimación numérica de ca-
da parámetro (Fuji 1969). Cada una de las clasificaciones usualmente están basadas en
la idea del nicho ecológico fundamental, tratado como una combinación de todos los re-
cursos usados por las especies (Hutchinson 1957), o todas las interacciones de las activi-
dades de las especies con esos recursos. El enfoque cuantitativo requiere una mayor eva-
luación objetiva, pero el problema es que aumenta la selección de parámetros por compa-
ración. Si usamos un gran número de características mofológicas, es difícil evitar confu-
siones en algunas características de valores adaptativos diferentes. La posibilidad de
reducir cada uno de los errores, es factible, siempre y cuando al clasificar, no se empleen
sólo los factores morfológicos, sino también los ecológicos. Entre estos factores ecológi-

2
 Rogovin, K.A., A.V. Surov and V. Serrano.
Nlche Convergence In the Desert Rodents of two Communities

cos se consideran primordiales: los hábitos alimentarios; patrones de utilización de espa-

cio, incluyendo conducta forrajera o de escape y el uso de refugio. Estas características
en ecología y conducta han sido extensamente utilizadas para determinar nichos tróficos
y espaciales (Pianka 1969; Brown 1975; Odum 1975). Este enfoque fue utilizado por
Shenbrot (1980, 1982) en la clasificación morfológica de los jerbos de Asia menor. Este sis-
tema se asemeja en gran parte a la clasificación morfológica de los jerbos adoptada en la
taxonomía actual. Esta semejanza es difícilmente posible cuando se hace la comparación
de caracteres ecológicos de formas filogenéticamente distantes e independientes y de co-
munidades geográficamente aisladas, debido a que la dibergencia ecológica dentro de un
grupo taxonómico puede exceder las diferencias entre representantes de diferentes taxa.
En tales casos es posible entontrar los equivalentes ecológicos entre formas no relaciona-
das filogenéticamente, los cuales han evolucionado en ambientes fí sicos y geográficos si-
milares. Tales trabajos pueden dar bases para la investigación de las peculiaridades es-
tructurales de comunidades geográficamente aisladas. El grado de semejanza en formas
ecológicas de diferentes comunidades nos conduce al uso del análisis de procesos de ta-
les evoluciones. Esto se basa en el estudio de características de la estructura, condicio-
nes ambientales y la historia de las comunidades geográficamente aisladas. En este tra-
bajo intentamos crear una clasificación ecológica de roedores del desierto, para identificar
formas ecológicas equivalentes, y analizar la estructura de las comunidades filogenética-
mente independientes de los desiertos de Asia Central y Norte América. Los datos sobre
la ecología y conducta de los roedores fueron colectados durante las expediciones de los
autores al área de Trans Altai Gobi como parte de la expedición biológica Soviético-Mon-
gola (verano 1980-1982), y durante la expedición al Desierto Chihuahuense (México) en
la Reserva de la Biosfera de Mapimí, Dgo. (Sept.-Nov. 1981-1983).

ABSTRACT

An ecological classification of rodents from two geographi-

cally isolated communities; Central Asia (Trans-Altai Gobi, MPR) and North America
(Central Chihuahuan Desert, Mexico) is proposed using the method of cluster analysis.
The ecological space of each community was determined through the use of the polar
ordination technique (Bray and Curtis, 1957). The criteria used were rodent diets (stomach
contents analized in terms of main food items) and exploratory behavior in "open field"
tests; this kind of behavior was related to space (microhabitat) use in nature. Geographi-
cally isolated species were in all cases clustered at a lower level of resemblance than
geographically related ones. Both in microhabitat use and feeding the Asian species were
most specialized. The distribution of species in the ecological space of Trans-altai Gobi
community was more uniform and the occupied volume was largerthan in Mapimi. There
were two groups of species in the Mapimi community one was composed of Heteromyid
rodents and another of North American Cricetidae. The diflerences in the volume of the

3
 Acta Zool. Mex. (ns), 10. 1985

ecological space in the Gobian and in the Mapimi communities were basically due to the
diflerent degree 01 ecological divergence between the bipedal and quadrapedal ricochet
rodents. The species richness in both communities are considered to be similar. The com-
munity 01 the Trans-Altai Gobi is composed 01 strongly divergent species 01 9 genera and
10 species. In Mapimi, there are 7 genera and 11 species.

INTRODUCTION

The desert, as an habitad for animals, has fo-

cused the attention and interest of ecologists many times, providing them
with a vast field of research on the convergence and parallel development
of characters (Kashkarov 1933, 1938, 1944; Kashkarov and Korovin
1936; Formosov 1956, 1976; Eisenberg 1967, 1975; Dlusski 1981; Mares
1976, 1980, 1983). Due to extreme temperatures, droughts, and great
weather fluctuations, life in the desert is shown in a rather limited way.
Natural selection is particularly powerful under such conditions, and its
trends coincide frequently in geographically insolated and faunistically
independent arid regions.
It seems likely that the structural differences in
rodent communities of the Trans-Altai Gobi and of Mapimi might have
resulted from the different environments in the deserts of Central Asia and
North America. The range of rodent macrohabitat variability in the Gobi
was greater than that in the Chihuahuan Desert. Differences in the propor-
tion of arid regions of the two continents could considerably affect the
rates of ecological and morphological divergence in desert rodents.

STUDYAREAS

The desert o Central Asia and southern North

America are located in different climatic belts. The former are in the
temperate and the laUer in the subtropical and tropical ones. They differ,
essentially, in terms of precipitation, diurnal and seasonal changes of tem-
perature and stability of climatic cycles. Despite these differences, there
are some common physiognomic features (similar geomorphological
parameters, aridity in climate, structure of plant communities) typical of
the desert in the broad sense (Meigs 1957; Petrov 1973; Walter 1975).

4
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communities

The Trans-Altai Gobi in the territory of Mongo-

lian Peoples Republic (between 43° - 45° N and 95° - 105° E) is charac-
terized by a great variety and extreme variability of its environmental
conditions. The climate is strongly continental. Precipitation occurs mainly
during summer and annual precipitation fluctuates between 20 to 100
mm. Air temperature varies seasonally from -34°C in winter to + 40°C in
summer (Gunin et al. 1980).
The Trans-Altai Gobi of the MPR was treated by
Mannikov (1954) as an idependent zoogeographic district of the Mongol-
Tibet province of the Palearctic Region.
The Mapimi Reserve (Durango, México) lies in
the northern part of the Mexican Plateau (26° 29' - 26° 52' N, 103° 32' -
103° 58' W). The climate is arid, subtropical, characterized by rather
steady fluctuations. Mean annual precipitation is 230 mm and is distribut-
ed mainly during the summer. Mean monthly air temperatures fluctuate
from 11°C in January-February to 28°C in June-August. Like in the Gobi
Desert, this region lies at an altitude aboye 1000 m.
Geomorphologically, the region is characterized by parallel ranges
of low mountains interrupted by alluvial valleys. The alluvial slope soil is
mostly covered with gravel and stones, and, like the Gobi Desert, sands
are rather rare and mainly represented by small fixed massifs. The alluvial
valleys are typical for their patchy vegetation chiefly composed of herbs,
while the flat interfluves are abundantly covered with low semi bushes,
shrubs and prickly pears of the genus Opuntia (Martínez and Morello
1977).
Biogeographically, the Mapimi region belongs to
the Chihuahuan province of the Sonoran subarea of the Neartic Region. It
can be treated as a special subprovince of the Chihuahuan Desert
(Morafka 1977).

METHODS

Rodents of Central Asia were collected in the

subzone of extra-arid deserts of the Trans-Altai Gobi (43° 10' - 43° 20' N
and 99° 00' E) ten species are widely distributed belonging to the following
families (Fig. 1):

5
 %

1°~3 .

~rm 1
m
t ¡ Eí
3b:J
j
3~1~Ib
.... ===k==~=d==~~~~
+ I + I I I ~
~
~
~
en
3 1 I ~
O
3 + + + i:Il
'"
3. 20 Ra.

I 1I

Figure 1
Rodenl species díslribulion Ihrough Ihe macrohabilals in Ihe extreme arid subzone
01 Trans-Allai GobL H measure 01 diversily 01 macrohabitales were utilized; "+ " - under
10% regislralion 01 Ihe species encounlered; Rodenl species: (see Table 1). I-VIII, see
"USA 01 Macro-Habilats", in melhods.
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Oesert Rodents 01 two Communities

Dipodidae (5 species): Euchoreutes naso

Sclater, Salpingotus crassicauda Vinogradov, Allactaga sibirica Foster,
A. natalie Sokolov, Dipus sagitta Pallas;
Muridae (1 species): Mus musculus Linnaeus;
Cricetidae (4 species): Cricetulus migratorius
Pallas, Phodopus roborovskii Satunin, Meriones meridianus Pallas,
Rhombomys opimus Lichtenstein (Kulikov and Rogovin 1980).

Rodents of North America were collected in the

Mapimi Reserve. In this zone there are 11 species of terrestrial rodents
(Grenot and Serrano 1981) of the following families (Fig. 2):
Sciuridae (2 species): Spermophilus variega-
tus Erxleben, S. spilosoma Bennet;
Heterornyidae (5 species): Perognathus flavus
Baird, P. penicillatus Woodhouse, P. nelsoni Merriam, Dipodomys merria-
mi Mearns, D. nelsoni Merriam:
Cricetidae (4 species): Neotoma albigula Hart-
ley, Onychomys torridus Coues, Peromyscus eremicus Baird and Sig
modon hispidus Say et Ord.

Of the two species of ground squirrels, Sper-

mophilus variegatus is extremely rareo Sigmodon hispidus is also
extremely rareo According to the new data Spermophilus mexicanus Erx-
leben and Peromyscus maniculatus Wagner are also found there,
however these species are too rare to be important to community
dynamics.
Rodents were captured with Sherman traps in areas of 1 ha in size.
There were 25 traps per plot with 20 m distances between them. Densities
were estimated through the Lincoln index during three days when most of
rodents were marked. In the Trans-Altai Gobi plots were followed and 20
in Mapimi.
Trapping areas were selected to cover the entire
range of environmental resources in the region.

Use of Macro-Habitats

In accordance with the floristic subdivisions of

7
 Acta lool. Mex. (na), 10. 1985

the extra-arid subzone of the Trans-Altai Gobi (Feodorova 1980) we dis-

tinguished the following macrohabitats (Fig. 1):

l. Large-stone slopes of the mountains with

no plant and animal species.
11. Medium-stone piedmonts with plant com-
munities composed of Ephedra przewals-
kii Stapf. + Zygophyllum xantoxilon Ma-
xim. + Iljinia regelii (Bunge) Korov.
111. Roch plains with vegetation composed of
l. regelii+ Haloxilon ammodendron (C.A.
Mey) Bunge.
IV. Rock plains without vegetation and de-
pressions covered by E. przewalskii, Z.
xantoxylon, Calligonum mongolicum
Turcz.
V. Proluvial valleys crossed by dry river-beds
with shrub communities - H. ammoden-
dron + C. mongolicum + E. przewalskii
and Reaumuria songorica (pall.) Maxim.
+ Artemisia scoparia W. et K.
VI. Sand dunes fixed by H. ammodendron.
VII. Dases with azonal vegetation.
VIII. Ecotones between oases and desert
areas.

In accordance with the floristic subdivisions of

Mapimí (Martínez and Morello 1977) the following macrohabitats were
distinguished (Fig. 2):

1. Large-stone slopes of the mountains with

sparse cactus-shrub vegetation and
yuccas.
11. Medium-stone slopes with flat surfaces
covered by Agave asperrima Jacobi + La-
rrea divaricata (D.C.) Coville + Euphorbia
antisyphilitica Zucc.
111. Small-rock and gravel piedmonts with cac-

8
 % H%
100
0 3 + + P. m.

3 + 80 S.s.

j + 70 P.f. z
g:
j '"
9
+ + 75 Pp
"<:xl
'" o
j 60 P.n. .31.0

--
'" !L
;:, o
'"
O"
.?<
j + + + 40 Dn. 5"-t"
'" »
(l) j 85 D.m. ~<
",.

'" a'"
:le:

j + + + 85 P.e.
:xl
o <
"-D>
""
¡a"-
j 35 O.t. "'<
o·

3I~
-'i(,'
.' i ~
. .
• '<.):
: .. ::.:
+ + 35 N.a.
o",
<:l"
~ o

I~ ) ~ . . .~"'. .. J .... \J ~ ~
. iof'::"~ .lo- ...
'" , ... . ..
3
e
"'"
. ·:e.'.,: ;>...... :"~•.:".' . ·,·c e .. :~'" . ". ';, ' .. " "', . X~63 ¡¡¡
I~
" ' " ,-.'.~

1 1I ¡ II . IV' "1 VII '"

Figure 2
Rodent species distribution through the macrohabitates in
the Mapimi deserto Rodent species (see Table 1). Types 01 macrohabitates: I - VII, see
"Use 01 Macro-Habitats" in methods. Other designations the same as in Fig. 1.
 Acta lool. Mex. (nB), 10. 1985

tus-shrub vegetation: Opuntia rastrera

Weber + K. divaricata + A. asperrima +
Fouquieria splendens Engelm.
IV. Gravel plains with sparse shrub vegeta-
tion of L. divaricata.
V. Sand dunes with shrub communities: L. di-
varicata + Jatropha dioica Cerv. + Yucca
sp. and herbs.
VI. Alluvial valleys covered by Prosopis glan-
dulosa and herbs (Hilaria mutica (Buck)
Benth).
VII. Alluvíal valleys wíth shrubs (Suaeda ni-
grescens Jonston) and herbs (H. mutica).

Within-Macrohabitat Distribution

In order to estímate the correlatíon between

rodent locomotíon actívítíes and wíthin-macrohabitat species distribution,
foot tracks were registered on squares of smoky foil glued to small boards,
30 x 30 cm in size. Sunflower seeds and oat flakes were used as baits. An
are a of 2 ha was selected at the nopalera of Mapími (Opuntia rastrera +
Larrea divaricata), and 50 boards with baits were placed in a grid system
of 20 m distances between them. We recorded 579 visits of 4 nocturnal
rodent species on these boards.

Exploratory Behavior

Individual exploratory behavior of an unknown

new territory was studied under laboratory conditions. Animals were kept
in individual cages to be tested later in plastic boxes (1 x1 x1 m)by using the
"open field" technique (Eisenberg 1963, 1967). Sifted sand was used as
the substrate. Each species was tested 10 times, during 10 min each time,
the number of males and females being equal.
. The following patterns of activity were recorded:
number of different gate transversals over the box, patterns of freezing
and sniffing postures, jumps, and substrate digging. It was proposed that

10
 Ragavin, KA, AV Surov and V. Serrano.
Niche Convergence in the Desert Aodents ol two Communities

all these patterns of activity relate to the use of 100d resources scattered in
the space, avoidance of predators, and shelter-building. Over 40 years
ago bird behavíor in food gathering was successfully used by Shulpin
(1939-1940), he identified the so-called "adaptative types" 01 birds. As
found in the experiments of the "open field" type, the level 01 activity in ro-
dents and some forms of their behavior were directly connected with their
mode of life (Wilson et al. 1976, Dewsbury et al. 1980, Vigorov 1980).
Types of locomotion were speci1ied in accord-
ance with the classification of gaits offered by Gambarian (1972): 1)
symmetrical (diagonal) gaits - a) diagonal quadrupedal run, b) symme-
trical bipedal run; 2) asymmetrical gaits - a) four-Iegged ricochet jumps
(quadrupedal ricochet, b) hind feet ricochetjumps (bipedal ricochet).

Food Habits

Food habits were determined by the ratio 01 vo-

lumes 01 basic food types, Le., seeds and fruits, green parts of plants, and
invertebrates, contained in the stomach (visual estimation with an error
of not more than 10%). We examined 469 stomach contents from Trans-
Altai Gobi and 122 stomach contents from Mapimí. Rodents were collect-
ed during the wet season of intensive vegetation, flowering and ripening of
plants.

Data Analysis

The species were compared in terms of their pe-

culiar ecology and behavior by using the coefficient of percentage resem-
blance:

¡
Yih I
1-% j Pij-Phj I
where Pij and Phj are the frequences of the j resource use by the species
or the frequence of their behavior j (Shoener 1968, 1970). The resem-
blance coefficients were employed to classify the species by their
ecological and/or behavioral features. Dendograms were based on the

11
 ActaZool. Mex. (ns), 10. 1985

data of the unweighed pair-group method of cluster analysis (Sokal and

Sneath 1963).
The species were ordinated in the system of
three axes of ecological space by using the Bray and Curtis method (1957)
(Beals 1960). The least similar species were at the ends of the X asis (the
first axis of coordinates) at the maximum distance from one another. The
position of all other species on the X axis was calculated by the formula of
Pythagorean distances:

where Xi is the position of the species "i" on the axis X, "a" is the distance
from the first end of the axis to the point Xi, "b" is the distance from the
second end of the axis to the point Xi, and "c" is the distance between the
ends of the axis X. The terminal points of the axis Y (the second axis) re-
presented a pair of species closely situated on the axis X but with low coef-
ficients of resemblance. Then the po sitio n of all species on the second
axis was found. After that, using the same principie, the third axis was
drawn. In this technique, the polar ordination axed do not show any actual
measurements of ecological space. The models of polar ordination
graphically reflect the interposition of species in the integral space of the
investigated ecological (ecologoethological characters),
The position of species in the structure of the
community in each desert was analyzed by the pattern of the regression
curves of the ecological resemblance coefficients. This method was used
by Inger and Colwell (1977) in their analysis of the community structure of
amphibians and reptiles, To draw these curves, the resemblance coeffi-
cients of each species were ranked on the principie of nearest neighbour.
Each rank yielded the mean value of the resemblance coefficients, The
number of ranks was equal to that of species in the community minus 1.
The curves were compared by their relative sizes, and their lack of para-
lIels (geumatríc íncongruence) (Plokhinsky 1970).
The mean values were compared by the stan-
dard criterion of Student.
The variety of rodents' diets and their use of ma-
crohabitats were assessed by using the information index of variety
offered by Shannon (Shannon and Weaver 1949):

12
 Rogovin, K.A, AV. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communmes

I
H = j Pij log Pij

where H is the variety index (entropy) Pij the frequency of use of the re-
source "j" by the species "i".

RESULTS AND DISCUSSION

Use of Macrohabitats

Rodent species distribution along an altitudinal

gradient of macrohabitats is shown in Fig. 1 and Fig. 2. The species of Ma-
pimí use a larger range of macroconditions. The average Shannons index
for macrohabitats occupied in Mapimí is 63%, in the Trans-Altai Gobi it is
41 %. The most specialized forms live in Mongolia, but the most plastic
(eurytopic) ones - Mexico. The limits of similarity of species in their use of
macrohabitats are equal in both communities: in the Trans-Altai Gobi from
9 to 67%, in Mapimi from O to 66%. But, because of the higher number of
eurytope species in Mapimi, the average overlaps in spatial rodent niches
differ: in the Trans-Altai Gobi - 17 ± 3% (the number of pairs compared (n)
is 45, 10 species), in Mapimí - 31 ± 3% (n = 36, 10 especies); P < 0,01).
We shall differ discussion of these differences to the last section .

.Exploratory Bhavior and Use of Microhabitats

Peculiar rodents' exploratory behavior reflect

their use of the local microconditions. The distribution of 4 species of noc-
turnal rodents living in the cactus-shrub (nopalera) plant community of
Mapimi iIIustrate this correlation.
It is characteristic of P. nelsoní, when exploring
an unfamiliar territory, to move by ricochet quadrupedal jumps, even in
slow locomotion. This species has typical horizontal orientation postures
and its fossorial activity is very intense. D. merriamí moves by quadrupe-
dal and also by bipedal ricochet jumps. Its frequency of orientation postu-
res and substrate-digging is lower than in P. nelsoní. The scheme of these
species behavior shows their ability to overcome places within vegetation

13
 Acta lool. Mex. (na), la. 1985

and to recover scattered food from the surface soil. In case of any danger,
these animals can change to quick bipedal and quadrupedal ricochet,
seeking the protection of shrubs or hiding in their intricate burrows. Under
natural conditions, as we found, these species gather their food from sites
covered with thick bushes and cacti and from those with thin vegetation.
Gambarian (1972) states that the primitive rico-
chet jump is typical for most rodents. However, this type of locomotion is
differently fixed in different species. A more progressive improvement of
the quadrupedal ricochet saltation reflects higher mobility and better
adaptation to life in deserts with patchy vegetation. The next step is a tran-
sition to bipedallocomotion which is best developed in Afro-Asian jerboas
(Fokin 1978). In American Dipodomys this mode of locomotion is still
primitive (simultaneous bipedal ricochet with the jerk effected si multa-
neously by both feet). When moving slowly, the animals often bear upon
their front feet. Their activity outside burrows is highly limited in time
(Reichman 1983).
There are numerous publications on the pro-
blem of spatial and trophic resources partitioning by representatives of
desert Heteromyd rodents (the latest reviews by Reichman 1983; Price
and Brown 1983). The hipothesis that bipedalism of Dipodomys was a
kind of morphological adaptation for more effective use of desert environ-
ment with patchy distributed resources of food (Reichman and Obershtein
1977; Price 1978; Reichman 1979) seems to be verified by visual obser-
vation of the feeding behavior of Dipodomys and Peronathus (Bowers
1982; Thompson 1982), as well as by experiments (Price 1983).
In case of N. albigula its exploration 01 an un-
familiar territory is characterized by numerous symmetric gaits, long
horizontal and vertical orientation postures with sniffing, freezing
postures, while its jumps, quick motions and soil-digging are rather infre-
quent . The behavior of N. albigula describes it as a dweller of well-protect-
ed sites with dense vegetation cover, as a good an careful scout-explorer
ready to escape to the thick bushes or to any other kind of shelter at any
time, in case of danger. The symmetric diagonal locomotion helps to
diminish the hight of the trajectory of the animal body during its movement;
this kind of locomotion is more efficient energetically in habitats with thick
grasses and shrubs (Fokin 1978). In Mapimí, N. algibula lives in thickets of
prickly pear and agaves, where it builds its vast nests and actively gathe-
red dry materials, prickles, fruits and stalks of near-by cactuses which to

14
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communitles

store. The mobility of this rather large animal (it weighs op to 250 g) is li-
mited to some hundred square meters. Wood rats move only along paths
and are extremely cautious.
As compared with N. albigula, Peremyscus
eremycus is more mobile and subtle (up to 20 9 in weight); its diagonal 10-
comotion is similar to the former one. The animals make frequent jumps,
their freezing are rather rare; like in wood rats, their fossorial activity is
relatively poor, the gather up their food on the soil surface. Under natural
conditions of cactus-shrub desert these animals were encountered on
sites abundant in dry materials, such as dry boughs, dry agave peduncles,
dead cacti. The animals can easily jump over stones and dried-up trees,
they can well climb up high branches or shrubs.
As our data show, species with similar forms of
behavior (Tabl. 1) were frequently found in similar places on our experi-
mental site (Fig. 3). Any direct competition between these species for
spatial resources seems to be low or absent, because they are rather
distant phylogenetically from one another (different genera) and also
differ in terms of size and appearance.

Exploratory Behavior
General Comparison

The peculiarities of behavior of 20 rodent spe-

cies during their exploration of an unfamiliar territory (Tabl. 1) were used to
make a special classification of these rodents in terms of their relation to
spatial resources (Fig. 4). Species with most similar forms of behavior in
some cases have similar morphology and are related phylogenetically (P.
eremicus - P. maniculatus, P. penicillatus - P. nelsoni, etc.). In other ca-
ses, judged by the features of maximum similarity in behavior, species of
more distant phylogeny were clustered together (S. spilosoma - o.
torri-
dus, S. crassicauda - E. naso).
There are 3 large groups of species:
The first group (from Mus to Onychomys) is
composed of slowly running, symmetrically locomoting forms which do
not build deep and intricate burrows (except S. spilosoma). They should
gather food mainly from the surface soil.
The second group (from Rhombomys to Dipo-

15
 Acta Zool. Mex. (na), 10. 1985

Tabla 1 - Part I

Frequences 01 the patterns 01 exploratory behavior 01 the

rodent species in the "open lield" tests.

• SPECIES
Patterns 01 behavior P.I P.p P.n D.m D.n P.m P.e

Diagonal quadrupedal 0.68± 0.61 ±

locomotion 0.04 0.2

Quadrupedal ricochet 0.63± 0.63± 0.60± 0.58± 0.46± 0.05± 0.06±

0.03 0.02 0.03 0.02 0.02 0.02 0.00

Bipedal ricochet 0.18± 0.16±

0.02 0.02

Diagonal bipedal
locomotion

Jumping 0.03± 0.02± 0.02± 0.02± 0.02± 0.06± 0.14±

0.01 0.00 0.00 0,01 0.01 0.02 0.01

Hind leet orientation 0.04± 0.06± 0.04± 0.06± 0,22± 0.04± 0.06±
(vertical) 0.01 0.01 0.01 0.01 0.02 0.02 0.01

Four limb orientation 0.07± 0.06± 0.10± 0.02± 0.03± 0.12± 0.09±
(horizontal) 0.02 0.01 0.02 0.01 0.01 0.03 0.01

Freezing postures 0.15± 0.01 ± 0.01 ± 0.03± 0.01 ± 0.01± 0.01±

0.02 0.01 0.01 0.01 0.00 0.01 0.00

Digging 0.09± 0.23± 0.22± 0.12± 0.10± 0.05± 0.04±

0.02 0.02 0.02 0.02 0.01 0.02 0.01

Sum 01 acts per 23.5 41.9 39.4 48.8 56.7 78.0 68.9
1 Test

*P.I - Perognathus flavus - P.p - P. penicillatus - P.n - P. nelsoni - D.m - Dipodomys

merriami - D.n - D. nelsoni - P.m - Peromyscus maniculatus - P.e - P. eremicus

16
 Rogovin, K.A., A.V. Surovand v. Serrano.
Niche Convergence in the Desert Rodents of two Communities

Table 1 - Part 11

Frequences 01 the patterns 01 exploratory behavior 01 the

rodent species in the "open lield" tests.

• SPECIES
Patterns 01 behavior N.a. 0.1. S.s. C.m. P.r. R.o. M.m.

Diagonal quadrupedal 0.59± 0.50± 0.54± 0.69± 0.72± 0.01± 0.01 ±

locomotion 0.02 0.03 0.02 0.02 0.02 0.00 0.00

Quadrupedal ricochet 0.05± 0.02± 0.48± 0.54±

0.00 0.01 0.02 0.02

Bipedal ricochet

Diagonal bipedal
locomotion

Jumping 0.01 ± 0.06± 0.01 ± 0.02± 0.01 ± 0.02±

0.01 0.01 0.00 0.01 0.00 0.01

Hind leet orientation 0.09± 0.12± 0.24± 0.02± 0.02± 0.10± 0.06±
(vertical) 0.01 0.02 0.02 0.00 0.00 0.01 0.01

Four limb orientation 0.12± 0.16± 0.08± 0.06± 0.14± 0.16± 0.15±
(horizontal) 0.02 0.02 0.01 0.01 0.01 0.01 0.01

Freezing postures 0.18± 0.02± 0.01± 0.03± 0.02± 0.01± 0.01±

0.02 0.01 0.00 0.01 0.01 0.00 0.00

Digging 0.02± 0.12± 0.13± 0.18± 0.10± 0.23± 0.21±

0.01 0.02 0.01 0.02 0.01 0.02 0.04

Sum 01 acts per 46.7 68.3 72.5 66.2 77.6 68.8 83.3
1 Test

'N.a. - Neotoma albigula - 0.1. - Onychomys torridus - S.s. - Spermophylus spilosoma-

C.m. - Cricetulus migratorius - P.r. - Phodopus roborovskii - R.o. - Rhombomys opimus-
M.m. - Meriones meriodianus

17
 Acta Zoo!. Mex. (na), 10. 1985

Table 1 - Part 111

Frequences 01 the patterns of exploratory behavior of the

roden! species in the "opan fieldo tests,

• SPECIES
Patterns of behavior E,n, S.c. A.n. A.s. 0.5. Mus,m.

Diagonal quadrupedal 0,S9±

lacomotion 0.02

Quadrupedal ricochet 0.04± 0,29± 0.28± 0.01 ± 0.03±

0.01 0,02 0.01 0,00 0.01

Bipedal ricochet 0.57± 0.69± 0,37± 0,38± 0.07±

0.02 0.01 0.02 0,02 0.01

Diagonal bipedal 0.Q1± OAS±

locomotion O,OO± 0.02±

Jumping 0.09± 0.12± 0,04± 0.03± 0.02±

0,01 0,01 0,01 0.01 0.01

Hind feet orientalion O.lS± 0.03± 0.17± 0.15± 0.31± 0,03±

(vertical) 0,01 0.01 0.01 0.01 0.02 0.01

Four limb orientation 0,01± 0,04±

(horizontal) 0.00 0,00

Freezing postures 0.03 0.05±

0,00 0,01

Digging 0.13± 0.13± 0,14:± 0,15± 0.13± 0.15±

0,01 0.01 0.01 0.01 0,01 0.01

Sum of acts per 79,S 122,0 9004 37,0 79.3 75,3

1 Test

"E.n. - Euchorautas naso - S,c, - Sa/pingotus crassicauda - A,n, - A//aclaga nata/iaa-

A.s, - A//aclaga sibirica D,s, - Dipus sagitta - Mus, m, • Mus musc%~~s

18
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communities

N.a. --------1------.
Pe.--------------------~
O.m.--------------~
p, n. _ _ _ _ _ _ r---------'
% 100 75 50 25 o

N.a. ----,..._ _ _ _ _ _ _ _ _----.

P.e. -----'
O.m. ---1------------.J
P. n. - - - - - - '

% 100 75 50 25 o
Figure 3
Classification of 4 nocturnal rodent species a) in their rela-
tions to microhabitates in the cactus-shrub (Nopal era) desert of Mapimi, b) in exploratory
behavior in the "open field" tests.

19
 Acta Zool. Mex. (n.), 10. 1985

,.,
al
:;
<:
C al
.t:
u.i :t
4 es
IIJ '><U
.t:
C .2l
O
.5
E .'!l
O <:
al
.,.;
Q;

e "ªt:
al
Q; '"
al
'C
Q, al
Q; -.t:;
E fe
~ =0
:lI <:

il:i
Q
a:
,g
'iJi
...
o
<U
(3
lIi ~
~
IIJ
ai ~
Q;
~
E
Q;
'E
~
<:
«i :>
Z <:
<U

'"
Q; ~
o
Q.
E )(
al
(,)
E
r¡j
:>
~I I I
o o o
...
o In

20
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communities

domys) is composed of active quadrupedal ricochet rodents which can

gather food from the soil surface and from its surface layer, and dig deep
burrows.
The third group (from Salpingotus to AI/actaga)
is composed of sprinter bipedal ricochet jerboas which dig deep burrows
and look for food as the species of the second group do. D. sagitta, which
represents tridactyl jerboas, is close to the last two groups. It is charac-
terized by symmetrical bipedal runo When this animal is quiet, it moves
only by pace, which seems to enable it to gather food at low energetical
cost and to hide behind and under shrubs or to climb over branches. At the
same time, when crossing sites covered with extremely thin vegetation, D.
sagitta can readily take to rapid ricochet saltation on its hind feed.
The symmetrical uniform rhymic form of locomo-
tion in D. sagitta and in sorne other species of jerboas seems to have
provided these species with an ability to widen the area of spatial re-
sources use.
AII jerboas use open areas and actively escape
those with thick vegetation. Quadrupedal ricochet rodents require well
protected environments which can be either their intricate burrows or thick
bushes. Species with predominant symmetrical type of locomotion corre-
late in their distribution with thickets of bushes and herbs.
No Gobian jerboas has similar exploratory be-
havior forms among Mapimi's rodents. American Perognathus and Dipo-
domys are grouped together with Asian Meriones and Rhombomys, but
not with the bipedal Dipodidae. The first genus is more similar to gerbils,
as it was also found by Mares (1980) when he analyzed the morphoecolo-
gical characters.

Food Habits

The classification of 20 species in terms of their

food habits, as presented in the dendrogram in Fig. 5, is bases on the com-
parison of their diets.
Among species which are mainly hebivorous
(green plant eaters) (group 1 in Fig. 5) the percentage of herbaceus plant
material in the stomach contents increases in the following order: D. sagit-
ta - D. nelsoni - N. algibula - R. opimus.

21
 Ro. N.a. 0.5. Dn Pe. Pm.CmMus.mMmAn D.m. Pp. Pn Pf Se As. 5.5 O.t. Pr En
100
.1

1\:)
75
l ti
¡;¡

~
~
1\:)
!
;:;
~
50

_1 2
25
Figure 5
The composilion 01 desert roden! die!s and classilieation 01
species in feeding. Designalions: 1 green parts 01 plants, 2 - seeds and Iruits, 3 - inverte-
brales. Digits wilhin the circular diagrams - indexes 01 Ihe diel diversity H.
 Rogovin, K.A., A.V. Surov and V. Serrano.
Nlche Convergence in the Desert Rodents ot two Communities

Peromyscus species to A. nataliae, used mixed

food, green parts of plants, seeds and insects (approximately in equal pro-
portions) are included in the second groups.
The next group embraces granivorous species.
The proportion of seeds in diets increases in the row: P. penicillatus - P.
nelsoni - P. flavus - S. crassicauda.
The kangaroo rat D. merriami is situated bet-
ween the second third groups.
At last, there are species which consume more
animal food (insects and other invertebrates), and these are included in
the fourth group. The rate of diet specialization increases in the row: A. si-
birica - o. torridus - P. roborovskii - S. spilosoma - E. naso.
In all comparisons, the most specialized are the
Asian species. Shannon's indexes are minimal in Asian R. opimus, E.
Naso and S. crassicauda (Fig. 5). However, there are exceptions. Several
species among Gobian forms which consume a wide spectrum of food.
On the whole, the average index of diet diversity in Mexican rodents is
70%, exceeding that in the Asian species (60%).
Four trophic groups are represented in each
community. The average overlap in diets of species in the Mapimí com-
munity is 60 ± 3% (n=45, 10 species), in the Trans-Altai Gobi it is 51 ± 4%
(n = 45, 10 species). The differences are not significant, (p>0.05).

Ecological Classification and Structure of

Communities

The generalized ecological classification of

desert rodents, represented in Fig. 6, shows graphically the results of mul-
tiplication of pair-wise similarity indexes in terms of the animals' exploratory
behavior and feeding. We found such an approach possible, beca use the
positive correlation between throphical and behavioral indexes of simila-
rity is rather low (r = 0,16, with n = 180 P< 0,05). Based on the aboye
data, we believe that the food type and the way in which the animals use the
space (in our case it is their exploratory behavior) are independent varia-
bles (Gody 1974).
In the dendrogram, the species, wich are ge-
ographically related, are clustered together by the principie of maximum

23
 Na Pr Ss Ot MusmCm Pe Pm Pp Pn Pf On OmMm Ro An As Os Se En
100
%

-,
-¡-

1 '--,-

-- '-- a
~
~
50 J
I !¡:;
~

I -

I
I
o
Figuree
General classiflcation of desert rodents in eXploratory
behaviof and feeding.
 Rogovin, KA., A.V. Surov and V. Serrano.
Niche Convergence in lhe Desert Rodents of two Communities

similarity (Fig. 6). Geographicany separated 10rms are in all cases group-
ed at a lower level of resemblance. Ittestifies to low level relationship bet-
ween ecological forms representing different communities. It is not easy
to specify ecological groups of distinct characteristics. Nevertheless,
sorne correlations are interesting. The American kangaroo rats (genus Oi-
podomys) are closer to the Asian gerbils (genus Meriones and g. Rhom-
bomys). The American deer mice (g. Peromyscus) are grouped together
with the Asian C. mígratorius and Mus musculus. The Gobían P. roborovs-
kii is closest in its ecology and behavior to o. torridus and S. spílosoma
from Mapimí.
In the system we have accepted for measuring
the ecological space, sorne Gobian jerboas (genera Euchoreutes, Salpín-
gotus, Oipus) have no close ecological 10rms in the communíty of Mapimí.
A reciprical situation is found in Mapimí's Perognathus which are trophi-
cally more specialized for seed-eating than the Gobian ricochet rodents.
The position 01 species in the system of two
parameters 01 ecological space (the trophics and the pattern of territory
use) was analyzed by using the technique of polar ordination of species on
the basis of reverse coefficients of similarity betwenn each species and
other members of the community (Fig. 7, a, b). The species' ordination in
the three dimensional space was established for the Trans- Altai Gobian
and Mapimí communities separately.
The distribution of species in the community 01
the Trans-Altao Gobi is more uniform, and the occupied hypervolume of
the ecological space there is largerthan in Mapimí. The species with spe-
cialized locomotion or/and diet (E. naso, S. crassicauda, O. sagitta and R.
Opimus) lie in the periphery of the model. In the Mapimí community there
are two groups of species, one of which is composed of seed-eating and
seed-herbaceous plant-eating ricochet rodents 01 the heteromyidae
family. The second group is made 01 North American Cricetidae. The
squirrel S. spilosoma bolongs to the same group.
The mean value of generalized coefficients 01
rodent similarity in the community of the Trans-Altai Gobi is 19.9 ± 2.7%,
with extreme values from Oto 79% (n = 45). The generalized coefficients
of rodent similarity in Mapimí vary from 8 to 69% (n=45), with their mean
value of 31.7 ± 3.7%. The differences are reliable in terms of Student's
test criterion (P<O.05).
In case of the Mapim í community, the dispersion

25
 Acta Zool. Me•. (na), 10. 1985

s.c.
E.n.

A.s.
A.n.

Figure 7
Ordinatlon 01 desert rodent species in three exes 01 ecolo-
glcal space: a) community 01 Trans-Altai Gobi, b) community 01 Mapimí.

26
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desart Rodanl. 01 two Cornmunilie.

D.m.
P.p. P.f.
Pn. D.n.

27
 Acta Zool. Mex. (na), 10. 1985

of the rodent similarity coefficients here is somewhat higher (6 Mapimí =

25,6 Gobi = 18).
The same conclusion can be drawn when analy-
zing the regression curves of the ecological similarity coefficients (Fig. 8).
The curves were drawn by ranking the similarity coefficients of each spe-
cies; in both communities the curves were descending. The curve for the
Mapimí community is above that for the Trans-Altai Gobi. The differences
are reliable in terms of one row excess over the other (P< 0.01). However,
the differences mainly involve forms of similar ecology.
On their terminal ends the curves converge clo-
sely. It can be attributed to the fact that in both communities the maximal
distance between species in the ecological space in close. The existence
of two groups of species in the Mapimí community is reflected in the step-
like pattern of the curve. At the same time the curve for the Trans-Altai
Gobi community is descending more smoothly, reflecting a more uniform
distribution of species in the ecological space there. 80th curves are
significantly, nonunparallel (P<0.01).
Discussing the sctructural difference of the com-
munities under consideration, one should pay attention to the differences
in the taxonomic distances between species in each of these communi-
tieso The community of the Trans-Altai Gobi is composed of strongly di-
verged, representatives of 10 species relating to 9 genera. In Mapimí,
there are rodents only of 10 species representing 7 genera; there are also
3 pairs of close species of one genus. There is only one pair of close spe-
cies of one genus in the Gobian community.
The differences in the volume of ecological
space in the Gobian and in the Mapimí communities (Fig. 7 a,b) are basi-
cally due to the different degree of ecological divergence of the bipedal
and quadrupedal ricochet rodents. As noted by Mares (1980,1983), such
ecological groups are characteristic of all typical desert faunas of long
evolutionary history. 80th communities reveal al! trophic groups of spe-
cies, including those specialized for seed eating, and nence, for living on
waterless diet.
The desert rodent faunas of Nort America and
Asia were developing independently from each other since the later Mio-
cene; their phylogenetic similarity is relatively low (Simpson 1945; Hall and
Kelson 1959; Walker 1964).
Finds of ricochet rodent fossils are related to

28
 Rogovin, KA, A.V. SUfOV and V. Serrano.
Niche Convergence in \t1e Desert Aodents 01 IWO Communities

80

70

60

50

40

30

20

10 a
b

pahru
Figurea
Regression CUNes of generalizad coefficients of the acolo-
glcal slmilarity of the rodents a) in Trans-Altai Gobi, b) in Mapiml.

29
 ActaZool. Mex. (ns), 10. 1985

Miocene. Jerboa of recent habit are known from the Early Miocene depo-
sits of Central Asia, though it seems most likely that they beca me sepa-
rated from the general Dipodoid stalk as far back as in lower Oligocene.
Gerbils are of African origin, they appear in Central Asia during Late Mio-
cene (Shevyreva 1983). The Heteromyidae differentiated in North Ameri-
ca during Miocene-Pliocene adapting the increasing landscape aridiza-
tion (Lindsay 1972; Hafner 1978; Patton et al 1981; Hafner and Hafner
1983). The family was became distinct much earlier, as Wood claims
(1935), in the Middle Oligocene. The recent subfamilies Perognathinae
and Dipodomyinae were already in existence in North America during
Late Oligocene (Hafner 1978), During Middle Miocene, Heteromyidae
were so widely spread in North America that some of their forms penetrat-
ed even into Central Asia (Shevyreva 1983),
In this way, paleontological data do support a
relatively modern development of the ricochet rodents of North America.
In terms of their abundance in species, the com-
munities under consideration can representatively characterize the fau-
nas of the large desert of both continents. The deserts in Magnolia are
most severe climatically. The are situated much northward than the de-
serts in Mexico which are more humid and warmer. On the whole, rodent
habitat variability (variation) in the Gobi is granter than that in the Chihua-
huan Desert.
The latter's phytocoenoses are more productive
and include more plant species. Association of abundantly fruiting bushes
(brushes) and cacti occupy a largerspace there. Vegetation-free areas are
rareo It is more difficult to identify floristic units geobotanic subdivisions
here than in the Gobi where communities are not so rich in species and the
boundaries between them are strictly correlated with the geomorpholo-
gical local characters. The greater ecological plasticity of the American
rodents in their macrohabitats use shown in our example (Fig. 1, Fig. 2)
may be due to the peculiarities of environments in Mapimí.
North American deserts seem never to have
been developed as much as those of Asia did. During glaciations, the So-
nora desert, for example, was covered by individual patchy arid refuges
alternated with humid areas (Martin andMehringer, 1965). This structure of
environment could have accelerated the rate of speciation but it did not
influence the scale of ecological divergence of the species. The diversity
of environment in different desert refuges was similar and restricted. It

30
 Rogovin, K. A. , A.V. Surov and V. Serrano
Nlche Convergence in the Desert Aodents of two Communities

seems likely that the structural differences observed by the authors in the
coomunities of the Trans-Altai Gobi and in Mapimí might have resulted
from the different environment in the deserts of Central Asia and North
America. Involving large arid regions of the two continents, these differen-
ces could considerably affect the rates of ecological and morphological
divergence in desert rodents.

ACKNOWLEDGMENTS

We acknowledge the cooperation of M.S. Jorge

Nocedal and Dr. Gary Adest for corrections and suggestions to the text in
the english translation. Nonetheless the mistakes or errors are all ours.
The third author would like to extend her gratitu-
de to Don Adalberto Herrera, for his great help during the field work in Mé-
xico.

LlTERATURE CITED

Axelrod, 0.1. 1956. Mio-Pliocene lloras lrom west central Nevada. Univ. Calif. Pub/o
Geol. Sci. 33: 1-321 .

Bannikov, A.G. 1954. Mammals 01 the Mongolian Peoples Republic. Trudy Mongo/skoi
Komissiiv. 53, Moskva·Leningrad, AN SSSR publ. (In Russian).

Beal., E.W. 1960. Forest bird communities in the Apostle Islands 01 Wisconsin. Wi/son
Bull.72:156-181.

Brav, J.R. and Curti., J.T. 1957. An ordination 01 the upland lorest communities 01
Southern Wisconsin. Eco/. Monogr. 27:3?5-349.

Brown, J.H. 1975. Geographical ecology 01 desert rodents. In: Cody, M.L. and Diamond,
J.M. (Eds.). Ec%gy and evo/ution ofcommunities. Harvard Univ. Press, Cambridge. 315·
341.

31
 Acta Zool. Mex. (n.), 10. 1985

Bowers, M.A. 1982, Foraging behavior 01 heteromyd rodents: lield evidence 01 resource
partitioning, J. Mamma/, 63:361-367,

Cody, M.L_ 1974, Competition and the structure of bird communities, Princeton Univ,
Press.

Oewsbury, O.A., Lanier, O.L. and Miglietta A. 1980, A laboratory study 01 climbing
behavior in 11 species 01 muroid rodents. Amer, Mid/and Natur, 103:66-72,

Olussky, G.M. 1981, Ants of deserts, Moskva, Nauka publ. (In Russian),

Eisenberg, J.F. 1963. The behavior 01 Heteromyd rodents, Univ, California Pub/, Zoo/,
69:1-100.

-1967, A comparative study in rodent ethology with emphasis on evolution 01 social be-
haviour. Proceedings of the United States Nationa/ Museum, 122:1-51,

-1975, The behavior patterns 01 desert rodents, In: Prakash, 1. and Ghosh, P,K, (Eds,),
Rodents in desert environments. Dr. W. Junk b, v, publishers, Hague, 189-224,

Feodorova, I.T. 1980, Vegetation 01 the eastern portion 01 the Trans Altai Gobi. Prob-
/ems of desert deve/opment 2: 46-57 (In Russian),

Fokin, 1, M. 1978. Locomotion and morpho/ogy of locomotor system of jerboas, Lenin-

grad, Nauka publ. (In Russian),

Formosov, 1950. World 01 animals, In: Grigoriev, A. A. (Ed.). Kazakhstan. Obshaya fizi-
ko-geograficheskaya kharacteristika/, Moskva-Leningrad, AN SSSR publ. 346-461, (In
Russian),

- 1956, Biological lorms 01 animals in arid and semi arid regions 01 Middle and Central
Asia, In: Gerasimov 1. p, (Ed,), Voprosy geografii. Sbornik statey d/ya 18 Mezhdunarodno-
go geograficheskogo kongressa, Moskva-Leningrad, AN SSSR publ. 238-248, (In Rus-
sian).

Formozov, A. N. 1976, Adaptations and convergent phenomena between animals in

similar landscapes, In: Nasimovich, AA (Ed,), Zveri, ptitsi i ikh vzaimootnosheniya so
sredoi obitaniya, (Sbornik izbrannykh trudov A,N, Formozov), Moskva, Nauka publ. 267-
294, (In Russian),

Fuji, K_ 1969, Numerical taxonomy 01 ecological characteristics and the niche concept.
Syst, Zoo/, 18: 151-153,

32
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents of two Communities

Gambarian, P. P.1972. Locomotion 01 nammals. Leningrad, Nauka publ. In Russian).

Goodall, D. W. 1953. Objective methods lor the classilication 01 vegetation. 1. the use 01
positive interspecilic correlation. Austral. J. 80t. 1 :39:63.

Goodall, D. W. 1954. Objetive methods lor the classilication 01 vegetation. 111. Essey in
the use 01 lactor analysis. Austral. J. Bot. 2:304-324.

Grenot, C. and Serrano, V. Ecological organization 01 small mammal communities at

the Bolson 01 Mapimi (México). In: Barbault, R. and Halffter, G. (Eds.), Ecologyofthe ehi-
huahuan Desert. Publ. 8. Inst. de Ecología, México. 89-100.

Gunin, P.D. Dedkov, V.P. and Dedkova, N.A. 1980. Radiation·heat balance and
main leatures 01 microclimate 01 Trans-Altai Gobí. Problems of desert development 2: 30-
46. (In Russian).

Hutchinson, G. E.1957. Concluding remarks. Cold Spring Harbor. Symp. Quant. 8iol.
22: 415-427.

Inger, R.F. and Colwell, R.K. 1977. Organization 01 contiguous communities 01

amphibians and reptiles in Thailand. Ecol. Monogr. 47.

Jardine, N. and Sibson, R. 1971. Mathematical taxonomy. N.Y. Wiley.

Kashkarov, D. N. 1933. Environment and community. (Fundamentals 01 sinecology).

Moskva, Medgiz publ. (In Russian).

Kashkarov, D. N. 1938. Fundamentals of animal ecology. Moskva-Leningrand, Medgiz

publ. (In Russian).

Kashkarov, D. N. 1944. Fundamentals of animal ecology. Leningrand, Uchpedgiz publ.

(In Russian).

- - and Korovin, E.P. 1936. Life of Desert. Moskva-Leningrad, Biomedgiz publ. (In
Russian).

Khodashova, K. S. 1953. Lile lorms 01 rodents in the Kazakhstan Plain and some princi-
pies 01 their geographical distribution. Proceedings of the Institute of Geography AS USSR
54: 33-194. (In Russian).

Krivolutsky, D. A. 1967. To the notion 01 "Lile lorm" in animal ecology. J. General

8iology 28: 153-162. (In Russian).

33
 AclaZool Mex. (ns).10. 1985

Kuilikov, V.F. and Rogovin, K.A. 1980. Small mammal species composition, areal
distribution and their numbers in Trans·Altai Gobi. Problems of desert development 2: 58-
70. (In Russian).

Lindaay, E.N. 1972. Small mammals lossils Irom the Bartsow Formation, Calilornia.
Univ. California Publ. Geol. Sci. 93:1-104.

Marea, M. A. 1976. Convergent evolution 01 desert rodents: multivariate analysis and

zoogeoraphic implications. Paleobiology 2: 39·63.

- 1980. Convergent evolution among desert rodents: a global perspective. Bull. of Caro
negue Museum of Natural History 16: 1·57.

- 1983. Desert rodent adaptation and community structure. Great Basin Natur. Mem. 7:
30-43.

Martin, P.S. and Mehringer, P.J. 1965. Pleistocene pollen analysis and biogeo-
graphy 01 the Southwest. In: H.E. Wright and D.G. Frey (Eds.). The Quaternary of the
United States 433·451. Princeton Univ. Press. Princeton, New Jersey.

Martínez, O.E. and Morello, J. 1977. El medio físico y las unidades fisonómico-florís-
ticas del Bolsón de Mápimí. Pub!. 3. Instituto de Ecología, México.

Meiga, P. Arid and semiarid climatic types 01 the world. In: International Geophysical
Union, 17th Congress Washington, D. C. 8th General Assembly Proceedings. 135·138.

Morafka, O.J. 1977. A biogeographical analysis of the Chihuahuan Desert through its
Herpetofauna. W. Junk bY publishers, Hague.

Naumov, N.P. 1948. Essays on the comparative ecology of micelike rodents. Moskva,
AN SSSR pub!. (In Russian).
Odum, Eu. P. 1975. Fundamentals of ecology. Moskva, Mir pub!. (In Russian).

Patton, J. L., Sherwood, S. W. and Yang, S. Y. 1981. Biochemical systematics 01

chaetodipine pocket mice, genus Perognathus. J. Mammal. 62: 477·492 ..

Petrov, M. P. 1973. Deserts of the world. Leningrad, Nauka pub!. (In Russian).

Pianka, E. R. 1969. Sympatry 01 desert lizards (Ctenotus) in Western Australia Ecology.

50: 1014.. 1030.

34
 Rogovin, K.A., A.V. Surov and V. Serrano.
Niche Convergence in the Desert Rodents 01 two Communi1ies

Plokhinsky. N. A. 1970. Biometry. Mosk. Univ. Press. Moscow. (In Russian).

Price. M. V. 1978. Seed dispersion prelerences 01 coexisting desert rodent species. J.

Mamma/. 59: 624-626.

Price. M. V. 1983: Laboratory studies 01 seed size and seed species selection by hete-
romyid rodents. Oec%gía. 60: 259-263.

- and Brown. J. H. 1983. Patterns 01 morphology and resource use in North American
desert rodent communities. Great Basin Natur. Mem. 7: 117-134.

Reichman. O. J. 1979. Subtly suited to a seedy existence. New Sci. 81: 658-660.

Reichman. O. J. 1983. Behavior 01 desert Heteromyid. Great Basin Natur. Mem. 7: 77-
90.

Reichman. O.J. and Oberstein. D. 1977. Selection 01 seed distribution types by Dipo·
domys merriami and Perognathus amp/us. Ec%gy 58: 636-643.

Shenbrot. G. l. 1980. Comparative ecology 01 jerboas in Turan deserts. Ref. Disserl.

Kand. Bio/. Nauk. Moskva, MGU publ. (In Russian).

-1982. Organization 01 jerboa (Rodentia, Dipodidae) communities in Turan deserts. J.

Genera/ Bi%gy 43 (4): 529-540. (In Russian).

Shevyreva. N. S. 1983. Neogen rodents (Rodentia, Mammalia) 01 Eurasia and Northten

Alrica. The evolutionary backgroundol Pleistocen and Modern launa 01 Palearctic rodents.
In: "History and evo/ution of modern rodent fauna. Vaprosi terriologii". Nauka publ. Mosk-
va. 7-146. (In Russian).

Shoener. T. W. 1968. Tlle anolis lizards 01 Bimini: resource partitioning in a complex lau-
na. Ec%gy. 49: 704-726.

- - 1970. Non-synchronous spatial overlap 01 lizard in patchy habita!. Eco/ogy. 51 :408-

418.

Shannon. e.E. and Weawer. W. 1949. The mathematica/ theoryof communication. Ur-
bana. Univ. lIIinois Press.

Shulpin. L. M. 1939. Ecological essay 01 the birds. Proceedings of A/maatinski State

Reserve. 1: 1·148. (In Russian).

35
 Acta Zool. Mex. (na), 10. 1985

- 1940. Omitho/ogy. Leningrad, LGU pub!. (In Russian).

Sinitsin, V. M. 1967. /ntroduction to Pa/eoc/imato/ogy. Leningrad, "Nedra". (In Russian).

Simpson, G. G. 1945. The principies 01 classilication and a classilication 01 mammals.

Bull. Amer. Mus. Nat. Hist. 85: 1-350.

Smirnov, E.S. 1969. Taxonomica/ ana/ysis. Moskva, MGU pub!. (In Russian).

Sokal, R.R. and Sneath, R.H. 1963. The principIes of numerica/ taxonomy. San Fran-
cisco-London, Freeman.

Vigorov, Yu, L. 1980. On species and ecological differences in exploratory behavior in

series offorms 01 litlle rodents. In: Riabitsev, V. K. (Ed.). Eco/ogicheskie aspecty povedeni-
ya zhivotnikh. Sverdlovsk, Uralskiy nauchniy tsentr AN SSSR pub!. (In Russian).

Walker, E.P. 1964. Mamma/s of the wor/d. John Hopkins Press, Baltimore. 1: 1-646; 11:
647-1500.

-1975, Vegetation of the wor/d. Moskva, Mir pub!. 3. (In Russian).

Wilson, R.e., Vacek, T., Lanier, O.L., and Oewsburv, O.A. 1976. Open-field
behavior in muroid rodents. Behaviora/ Bio/ogy 17: 495-506.

Wood, A.E. 1935. Evolution and relationships 01 the heteromyid rodents with new forms
from the Tertiary 01 western North America. Ann Camegie Mus. 24:73-262.

36