Commensalism: Was it Truly a Natufian Phenomenon? Recent Contributions from Ethnoarcheology and Ecology

Full text

699
Introduction
The commensalism model in archaeology, which
was first proposed by Bar-Yosef and Tchernov (1966)
links ecological consequences to social and cultural
processes in human settlements. According to the
model, important changes in the duration and inten-
sity of human settlement occupation are expected
to result in patterned and predictable changes in
the populations and structure of communities of
non-domesticated species of small animals (e.g.
small rodents and birds) that have long coexisted
with humans. Development of the model was tied
to research on the Natufian culture of southwest
Asia (ca. 16,000/15,000-11,200 cal BP). This is be-
cause the Natufian period is widely believed to have
involved a shift to sedentary lifeways and a signif-
icant increase in the level of human occupation of
settlements (e.g. Bar-Yosef and Belfer-Cohen 1989;
Belfer-Cohen and Bar-Yosef 2000; Garrod 1957). Use
of biological indicators from small animal remains
to gauge changes in the level of human settlement
occupation was also seen as a less biased method in
comparison to other commonly employed indicators
that are based on artifact and site characteristics
(see Bar-Yosef and Belfer-Cohen 1989; Tchernov
1984).
Early evidence for commensalism in the Natu-
fian was based on abundant remains of house mice
(Mus musculus domesticus) and house sparrows
(Passer domesticus) in deposits of Hayonim Cave,
northern Israel. These findings suggested that
Natufian occupation of the site was long-term and
intense based on the known association of such
species with modern highly sedentary settlement
environments (Bar-Yosef and Tchernov 1966). This
inference was further supported by the finding that
house mouse remains were rare in earlier deposits
from the Final Pleistocene. In later publications,
Tchernov elaborated on the theoretical underpin-
Commensalism: Was it Truly a Natufian Phenomenon? Recent
Contributions from Ethnoarcheology and Ecology
Lior Weissbrod, Daniel Kaufman, Dani Nadel, Reuven Yeshurun
and Mina Weinstein-Evron
nings of the commensalism model and together
with colleagues provided additional supporting
evidence for linking commensalism and Natufian
sedentism (Auffray et al. 1988, 1990; Tchernov 1984,
1991a, b). This included evidence from both zooar-
chaeological material and from ecological studies
of commensalism in contemporary settings. Other
authors have undermined the model, however, due
to lack of empirical data on the range of settlement
contexts wherein commensalism can be expected to
develop, including sedentary or more mobile ones
(Edwards 1989; Tangri and Wyncoll 1989).
Here we assess the validity of the commen-
salism-sedentism linkage and its relevance to
the Natufian culture. We examine variability in
Natufian micromammalian assemblages in light
of empirical data from ethnoarcheological and
ecological studies in contemporary settlement
settings and other human modified environments.
In this comparison we employ ecological data
from present day settlements of seasonally mobile
pastoralists (Weissbrod 2010a, b) and sedentary
agriculturalists (Misonne 1963). We use data on
living communities of small rodents and shrews
(micromammals) associated with the settlements
and consider patterns in the organization of these
communities including the composition and relative
frequencies of species and biological diversity, which
may be represented by the number of species and
heterogeneity in distribution of frequencies among
different species. Data is also presented on assem-
blages of micromammalian remains from a number
of important Natufian sites and different phases of
the Natufian (Fig. 1; Table 1). These samples were
retrieved from different ecological and environmen-
tal areas of the Mediterranean climate region and
may represent different climatic fluctuations. We
examine the variability among these samples and
indications for commensalism by considering the
recent information from ecology and ethnoarcheol-

700
Lior Weissbrod et al.
ogy and information on the archaeological context
and taphonomy of the assemblages.
Factors affecting variability
among Natufian micromammalian
assemblages
The question of sedentarization in settlement
occupation in the Natufian and in particular its
pre-agricultural origins have been central to dis-
cussions on the significance of the Natufian phe-
nomenon in the cultural history of southwest Asia
(Bar-Yosef 2002; Bar-Yosef and Belfer-Cohen 1989;
Belfer-Cohen and Bar-Yosef 2000; Binford 1968;
Flannery 1972; Henry 1985; Kaufman 1992; Perrot
1968). A wide range of drivers including climatic
and social, limiting or enabling factors have been
proposed to account for the appearance of sites
with seemingly abrupt increase in size, density of
artifacts, scale of construction, storage activities,
burial, occurrence of heavy stone implements, and
frequency and variety of material expressions of
art and decoration.
The majority of such Natufian sites with
evidence for significant sedentism and all those
contributing samples to the present study are con-
fined to the Mediterranean climate region of Mount
Carmel and the Galilee that are considered the
core area of the Natufian culture (Fig. 1) (Bar-Yosef
1983). Evidence has also been considered, however,
for significant variability in mode and intensity of
settlement occupation both within and between
different phases of the Natufian (Bar-Yosef 1983;
Byrd 1989; Valla 1998). Thus, whereas some sites
that are located in prime resource locations, typ-
ically in ecotonal zones, contain all the landmark
characteristics of the Natufian culture and evoke
significant levels of occupation (e.g. Bar-Yosef 1991;
Valla et al. 2002; Weinstein-Evron 2009) others
are of much smaller scale and suggest use on an
ephemeral basis possibly for special purpose and/
or seasonally scheduled activities such as use of
seasonal resources, burial, and ritual/ceremonial
practices (e.g. Bar-Yosef 1983; Grosman 2003;
Marder et al. herein; Nadel et al. 2008; Valla 1998;).
Such differences in the function of sites and
scale of occupation can also be identified within
sites in relation to the different temporal phases
of the Natufian and may be tied to either social
or climatic developments or both. At the site of
el-Wad Terrace on the western coastal aspect of
Mount Carmel such a shift is recognized in the
transition from the Early Natufian (EN) to the
Late Natufian (LN) (Garrod 1957; Weinstein-Evron
2009; Weinstein-Evron et al. herein). Evidence for
an extensive settlement in the EN phase includes
a complex array of architectural components such
as stone walls, slab pavements, and rock-cut ba-
sins indicating residential activities in addition to
organized burial in the form of concentrations of
group and individual burials (Garrod and Bate 1937;
Weinstein-Evron 1998, 2009; Weinstein-Evron et al.
2007). This can be contrasted with the LN phase of
el-Wad Terrace when a part of the terrace area was
utilized mainly as a burial ground and the spatial
extent of occupation appears to have become more
restricted (Weinstein-Evron et al. 2007). At the
open-air Natufian site of Eynan in the Hula basin
of the Jordan Valley (Valla and Khalaily 1997;
Valla et al. 2002) there is some indication that the
construction of structures became progressively
less substantial from the EN phase through the
LN and a final short phase of the Natufian (Final
Natufian, FN).
Other sites such as the caves of Raqefet in in-
land Mount Carmel (Lengyel et al. 2005; Nadel et
al. 2008, 2009) and Hilazon Tachtit in the Galilee
Mountains (Grosman 2003; Grosman and Munro
2007), both are belonging to the LN phase, present a
very different picture of mode of use and occupation.
Both cave sites are situated on steep slopes and at
relatively high elevations of 50 m and more above
the valley bottom. The archaeological deposits of the
two sites mainly occur within the confined space of
the caves, which do not include extensive outside
terraces as is the case at the larger-scale site of
el-Wad, for example. At Raqefet Cave evidence was
Fig. 1. Map of northern Israel showing the location
of the three Natufian sites supplying micro-
mammalian assemblages to this study.

701
Commensalism: Was it Truly a Natufian Phenomenon?…
unearthed mainly of burial activities in addition
to the hewing and utilization of a wide variety of
bedrock mortars and cupmarks (Lengyel and Boc-
quentin 2005; Nadel and Lengyel 2009). Similarly,
a burial context was also uncovered at the cave of
Hilazon Tachtit in conjunction with evidence for
specialized ritual activities in the form of elaborate
grave preparation and distinctive association of
grave offerings (Grosman 2003; Grosman et al.
2008).
Variability in environmental context and climat-
ic fluctuations are additional factors to take into
account in comparing micromammalian assemblag-
es from different sites and temporal phases of the
Natufian. The site of el-Wad Terrace is situated in
what may be considered the quintessential ecotonal
zone at the contact of Mount Carmel and the coastal
plain and at the opening of one of the Carmel’s sea-
sonal stream valleys (Weinstein-Evron 1998). Data
on the fauna of the site have shown that species fa-
voring open and mainly grass-covered environments
thriving on the coastal plain such as gazelle (Gazella
gazella) and the social vole (Microtus guentheri)
dominate the LN assemblage (Bar-Oz et al. 2004;
Weinstein-Evron et al. 2007). A similar situation
appears to have pertained throughout the Natufian
sequence of el-Wad Terrace (Valla et al. 1986; see
also Garrod and Bate 1937; Weinstein-Evron 1998).
Species representing more forested environments
occur as well but in relatively low frequencies (Valla
et al. 1986; Weinstein-Evron et al. 2007; Yeshurun
et al. herein).
Table 1. Description of samples of micromammalian assemblages from Natufian sites included in this
study
Site/
assem-
blage
Temporal
phase
Approxi-
mate area of
excavation
and sampling
(m2)
No.
samples Context description
el-Wad
Terrace
Early
Natufian
(EN)
58/12 3
Samples collected from top of EN layer in associa-
tion with a wall of a partially preserved structure
(Str.II): 1) a living floor inside the structure (I/S
Str.II) adjacent to the concave side of the wall,
2) above the structure (Abv. Str.II) in deposits
overlying the wall, 3) outside the structure (O/S
Str.II) adjacent to the convex side of the wall and
situated between the wall and the bluff overhang-
ing the terrace (Yeshurun et al. this volume).
el-Wad
Terrace
Late
Natufian
(LN)
58/1a8
Seven of the samples (numbered I-VII) collected
from a sediment column (50×50 cm2) in spatially
undifferentiated deposits of LN layer, each sample
representing a c.5 cm spit (Weissbrod et al. 2005).
Additional sample collected from a LN burial de-
posit (Bur.; Weinstein-Evron et al. 2007).
Raqefet
Cave
Late
Natufian
(LN)
70/11 3
Two of the samples collected from loci with LN
deposits containing multiple burials (Loc. I and
Loc. III) and an additional sample from inside
two human-made bedrock holes within the cave
floor (HBH CI-II) (Nadel et al. 2008).
Eynan
Final
Natufian
(FN)
160/16 4
Samples collected from deposits inside well-de-
fined structures associated with features in-
cluding living floors, hearths, and burials: 235,
200/222/240b, 225, 228/242 (Valla et al. 2007).
aThe area of sampling in this case may be misleading in terms of the volume sampled because unlike all other
samples in the study it represents a column with a substantial depth (c. 0.25 m).
bSamples from several of the features were lumped for the purpose of this study due to their spatial association
and relatively small sample sizes analyzed so far.

702
Lior Weissbrod et al.
The location of Raqefet Cave is further inland
within an inner valley of southern Mount Carmel
and at an elevation of 230 m above msl (Nadel et
al. 2008, 2009). The cave is situated within a more
mountainous topography but is also adjacent to
the Menashe plateau and both gazelle and voles
are highly abundant. Typical woodland species
such as roe deer (Capreolus capreolus) and Persian
squirrels (Sciurus anomalus) are represented at
Raqefet Cave at somewhat greater frequencies
than at the contemporaneous LN deposits of el-Wad
Terrace on the western coastal side of the mountain
(Nadel et al. 2008, 2009). The open-air site of Ey-
nan lies at an elevation of 80 m above msl within
the Jordan Valley and at the piedmont zone of the
Galilee Mountains and the inundated basin of the
Hula (Valla et al. 2007). Quantitative data on the
micromammalian assemblage of the site have not
yet been published in detail (Bouchud 1987; Valla
et al. 1998). The Natufian inhabitants of Eynan
would have had access to both open grassland and
marshland environments of the valley floor and
higher elevation woodlands of the Galilee Moun-
tains (e.g. Leroi-Gourhan 1984).
Recent high-resolution climate data from
isotope records in cave speleothems in the region
indicate that the Natufian represents a period of
marked increase in rainfall and temperatures in
comparison to the preceding relatively dry and cold
period of the last glacial maximum (LGM) that
peaked ca. 19,000 bp (Bar-Matthews et al. 2003).
The LN phase is generally considered to coincide
with a short-term retreat in the trend of warming
and increased rainfall known as the Younger Dryas
although the consequence of this oscillation for the
environment of southwest Asia and for temporal
developments within the Natufian culture has
been widely debated (summarized in Byrd 2005;
Weinstein-Evron 2009). Data on the geology and
sedimentology of the Dead Sea in the Jordan Val-
ley also reveal an increase rather than decrease
in rainfall providing conflicting indications on the
nature of Younger Dryas climate change (Stein et
al. 2009). Cultural developments from the early to
the later phase of the Natufian whether driven by
external climatic events or by internal processes
appear to have involved major changes and possibly
increasing variability in the mode and intensity of
use of sites within the Natufian core area (Bar-Yosef
2002; Weinstein-Evron 2009).
A final set of considerations for inter-site/phase
comparisons of micromammalian assemblages
is related to the issue of taphonomy and factors
which could have affected the manner in which
micromammalian remains accumulated at the
sites. Taphonomic considerations that can further
complicate such comparisons include differences
in the mode of accumulation and preservation
among sites in cave/terrace and open-air settings,
the potential for direct contribution of humans to
accumulation through small-game hunting, and
within-site variability in assemblage composition
that may be related to the spatial organization of
human activities or to other processes affecting
site formation such as site topography and modes
of sedimentation.
Previous study of the taphonomy of the LN
assemblage of el-Wad Terrace revealed a signifi-
cant contribution to accumulation at the site by
nocturnal raptors such as owls (order Strigiforms)
that could have been nesting or roosting within
the cliff overhanging the site (Weissbrod et al.
2005). Certain species of owls that typically uti-
lize cave or cliff settings where they accumulate
the remains of their micromammalian prey are
believed to be important taphonomic agents in
such sites but less important in open-air sites
(Tchernov 1984). Additional evidence from el-Wad
Terrace showed that secondary contributions to
the LN micromammal ian assem blage included
human consumption of mole rats (Spalax ehren-
bergi) and some in situ accumulation of mice of the
genus Mus that may have lived at the settlement
as commensals (Weissbrod et al. 2005). In situ or
autochthonous accumulation may be an especially
important process in the formation of assemblages
of commensal animal remains but has received little
attention in research (Evans 1978; Reitz and Wing
1999; Tchernov 1984; Weissbrod 2010a). Within-site
spatial variability in taxonomic composition has
been detected to some extent at el-Wad Terrace
LN (Weinstein-Evron et al. 2007) and in a more
pronounced manner at Raqefet Cave (Nadel et al.
2008), in both cases partly in relation to distinct
activity contexts such as burials and bed-rock
mortars versus less differentiated deposits.
Conditions of preservation can also vary great-
ly among cave, terrace, and open-air depositional
environments. Previous analyses have shown that
the el-Wad Terrace LN assemblage exhibits con-
siderably greater rates of fragmentation than the
LN Raqefet Cave assemblage in spite of the fact
that both assemblages had high representation of
elements from all parts of the skeleton (Weissbrod et
al. 2005; Nadel et al. 2008). The high fragmentation
at el-Wad Terrace was related to processes of tram-
pling, exposure to aerial weathering, and transport
from the location of initial deposition by raptors

703
Commensalism: Was it Truly a Natufian Phenomenon?…
in the cliff to the terrace (Weissbrod et al. 2005). A
final important influence on observed variability in
taxonomic composition of micromammalian assem-
blages is the method of retrieval during excavation.
Retrieval of the assemblages described here was
conducted through fine wet-screening in mesh of
1-2 mm. It has been shown that some of the molar
teeth of especially small species such as mice of
the genus Mus may become lost even through 1
mm mesh screens (Weissbrod et al. 2005). It can
therefore be expected that differences in the level
of precision of retrieval will to some extent affect
assemblage composition and comparability.
The taphonomy of the assemblages included
in our paper has so far been studied only in part.
Nonetheless, in our preliminary consideration of
variability among micromammalian assemblages
from different sites and temporal phases of the
Natufian we attempt to take into account a number
of compounding factors and examine the potential
for preservation of information on commensalism.
A brief review of the commensalism
model and available evidence
The commensalism model posits that the dura-
tion and intensity of human settlement occupation
are major factors affecting the taxonomic composi-
tion and organization of associated communities of
micromammals. Tchernov invoked commensalism
to account for the appearance or marked increase
in abundance in the archaeological record of species
that are well known commensals in present day
settlement environments (Bar-Yosef and Tchernov
1966; Tchernov 1984, 1991a, b; see also Hesse 1979).
He focused on the Natufian but also traced earlier
Late Pleistocene fluctuations in the frequencies of
commensal species including three rodents, house
mouse, black rat (Rattus rattus), and spiny mouse
(Acomys cahirinus), and one bird, house sparrow
(Tchernov 1984). Tchernov linked early and rel-
atively small-scale fluctuations in frequencies of
commensal species to variability in duration and
intensity of settlement occupation through time and
emphasized a marked increase in frequencies in the
Natufian and its relation to significant sedentariza-
tion. In later publications (Tchernov 1991a, b) the
emphasis shifted from evidence that was based on
relative frequencies of species to establishing the
appearance of a single key commensal species – the
house mouse – in the Natufian. This approach was
based on morphometric analysis of mouse cranial
characteristics that revealed the occurrence in the
record of two closely related species of common
mice of the genus Mus – one which occurred in the
region at least since Middle Pleistocene times (Mus
macedonicus) and another which appeared for the
first time in Natufian deposits and is considered
to be the true commensal house mouse (Auffray et
al. 1988).
In order to link the seemingly abrupt appear-
ance of house mice in the Natufian to significant sed-
entism, Tchernov also drew on ecological evidence
on contemporary mouse species in Israel, which
suggests their strong dependence on human set-
tlement environments. A study on the distribution
of the two extant species of common mouse showed
that in drier regions of southern Israel where only
house mice occur they occupy both settlements and
non-settlement environments whereas in regions
where the two species are overlapping (i.e. sympat-
ric) within the Mediterranean climate zone, house
mice are mainly restricted to settlements (Auffray
et al. 1990). This indicates competitive exclusion
between house and outdoor mice in their region of
sympatry. The evidence for the exclusion of house
mice from non-settlement environments by outdoor
mice suggested to Tchernov (1991a, b) that only
the creation of a new type of habitat in highly
sedentary Natufian settlements could have opened
up a niche to be filled by a specialized commensal
species. Current evidence from DNA of the various
members of the Mus group inhabiting regions of
West and Central Asia and Europe show that the
ancestor of the house mouse in southwest Asia
originated in Central Asia during earlier periods
of the Late Pleistocene and underwent a westerly
expansion event which coincided with the post-gla-
cial period of the Final Pleistocene (Boursot et al.
1993; Rajabi-Maham et al. 2008). It is conceivable
that both climate change and the appearance of
more sedentary human occupation sites could have
contributed to the spread and establishment of the
present day house mouse in regions which were for-
merly inhabited by other, currently non-commensal
species of common mice. It still remains unclear
precisely what level of sedentism and related
habitat alteration would have been necessary for
the establishment of the commensal house mouse
in Natufian settlements. A fundamental question
that remains open is whether commensal species
can also occur in the settlements of seasonally
mobile communities (see Edwards 1989; Tangri
and Wyncoll 1989).
The ecological mechanism underlying the
commensalism model is a universal one. In eco-
logical theory, commensalism is a formal type of
interspecies interaction involving a positive effect

704
Lior Weissbrod et al.
of one species on the equilibrium population size of
another without immediate consequences for itself
and is denoted a +/0 interaction (Dickman 1992,
2006). A key aspect of commensalism is that in con-
trast to more direct ecological interactions such as
predation (+/-) and competition (-/-) it often occurs
indirectly through the contribution of one species
to the suitability of the environment of another.
Recent advances in ecological theory suggest that
commensal interactions can be subsumed within
a more generalized framework of indirect interac-
tions involving environmental modification that is
a fundamental and pervasive process in ecosystems
worldwide (Dickman 2006). This process has been
termed environmental engineering and is typical-
ly identified in situations where the activities of
environmentally more dominant organisms such
as trees, elephants, and dam-building beavers
alter physical conditions in the environment and
indirectly affect the supply of critical resources to
a host of other species (Jones et al. 1994, 1997; see
also Smith 2007 on archaeological implications). In
this framework, the creation and maintenance of
settlements that is a basic characteristic of human
culture may be seen as a specific form of environ-
mental engineering.
In his discussion of the development of commen-
salism in the context of sedentarization in Natufian
settlement occupation Tchernov (1991a) relates the
activities of human settlement occupation to habitat
alteration and associated ecological dynamics. By
drawing on principles of mammalian behavioral
ecology and optimal and central-place foraging,
Tchernov ties human occupation of settlements to
mobility, resource acquisition, and their ecological
consequences (Tchernov 1984, 1991a, b). Here,
increased permanency in use of settlements is pre-
dicted to occur in conjunction with reduced mobility
and enhanced and more spatially bounded effort of
resource extraction from the environment surround-
ing settlements. Such an effort of concentrated daily
resource extraction activities that are sustained for
long periods of time through intensified seasonal
occupation and under conditions of growing human
populations in settlements is expected to reduce
the abundance of many species in the immediate
environment and increase the abundance of a few
others. This ecological process implies a reduction
in biological diversity and a process of commen-
salization (Tchernov 1991a: Fig. 2). According to
Tchernov (1984, 1991a) the process of commensal-
ization is predicted to accompany the development
of sedentary settlements from less sedentary ones
and should represent a threshold in the nature
and scale of ecological interactions among human
inhabitants of settlements and coexisting species
of small animals. The exact levels of commensalism
or of human settlement occupation that define such
a threshold remain largely unknown.
Aspects of habitat engineering that could have
impacted ecological interactions in Natufian set-
tlements should include the extraction of subsis-
tence resources from the immediate environment,
collection of materials for construction and fuel,
the erection of various structures (e.g. dwellings,
installations, and fences or terrace walls), human
traffic, and the accumulation of refuse and storage
of foodstuffs. It can be envisaged that some of these
activities constituting substantial disturbance of
the natural environment of the settlements would
have contributed to depressing the abundance
of certain species whereas other activities could
have at the same time enhanced the abundance of
other species thereby leading to commensalism.
Theoretically, factors that can contribute to for-
mation of commensal bonds include the supply of
food and shelter and reduction of pressures from
competing species or predators (Dickman 1992).
These improved habitat conditions will be indi-
rectly provided by the unaffected partner to the
commensal species. A range of ecological studies
in present day settlement environments have
generally shown that the conditions provided by
human modified habitats can mitigate various
limitations on population size that pertain in less
modified habitats. Such research also indicates that
only species with suitable pre-adaptations and a
competitive advantage can benefit from the unique
circumstances in such environments (Andrzejewski
1978; Braithwaite 1980; Berry 1981; Canova and
Fasola 1994; Chernousova 2001, 2002; Courtney
and Fenton 1976; Ganem 1991, 1993; Pocock et al.
2004).
Tchernov (1991a) already observed that the bi-
ological and ecological characteristics of commensal
species may be equivalent to those of invasive or
colonizing species and that an invasive or colonizing
ability should represent a pre-adaptation to com-
mensalism. A recent summary by Sakai et al. (2001)
of the characteristics of invasive species includes: 1)
tolerance of environmental stress, disturbance, and
heterogeneity based on behavioral and/or phenotyp-
ic plasticity, 2) high productivity and reproductive
flexibility (i.e. r-selected life-history strategies), 3)
competitiveness through aggressive behavior, 4)
broad dietary preferences, and 5) wide distribution
across varied habitats. It can be expected that such
pre-adapted species occupying Natufian settlement

705
Commensalism: Was it Truly a Natufian Phenomenon?…
environments would have benefited from enhanced
availability of food from organic refuse or stored
supplies and of shelter from the construction of
various structures. In addition, many competitors
and predators of commensal species should have
been excluded to some extent from Natufian set-
tlement environments because they were either
directly exploited by humans, depended on other
exploited resources, or could otherwise not tolerate
escalating levels of disturbance from human activ-
ities. The exact manner in which different species
may react to specific human engineering activities
as well as to the overall impact of increasing levels
of occupation in sedentarizing human settlements
is little understood for most species of animals,
however.
Answers to some of the questions that remain
open regarding the commensalism-sedentism link-
age and the relevance of the commensalism model
for identifying sedentism in the Natufian depend
on the availability of additional empirical data.
These data can only be obtained from observations
on interactions between humans and micromam-
mals in a wide range of contemporary settlement
settings. Extrapolating ecological and cultural in-
formation from contemporary settings back in time
and across wide geographic regions is problematic.
This endeavor may be useful, nonetheless, for es-
tablishing generalized boundary conditions for the
past such as the distinction between the ecological
impact of seasonally mobile and sedentary human
communities.
Ethnoarcheological and ecological
contributions to the commensalism
question
Direct ecological observations in settlements
of complex hunter-gatherers equivalent to those
of the Natufians would have been ideal for our
purposes. Such societies do not exist in the modern
world, however. Relevant data is presently available
from only a few studies in other types of settlement
contexts. Courtney and Fenton (1976) studied the
influence of a seasonal holiday retreat in Canada
on populations and community composition of
micromammalian species. Their study focused on
the garbage dump of the retreat which was used
during May-October of each year by 4-5 families
dumping >200 liter/week of garbage. Monitoring
of micromammals in the dump and in an off-site
control revealed greater overall abundances, levels
of female fecundity, and juvenile survival in the
dump than in the adjacent control site. These data
indicate increase in populations in the dump.
In addition, it was shown that a local and high-
ly abundant rodent species – white-footed mouse
(Peromyscus leucopus) – was an early colonizer of
the dump during the seasonal cycle but that it was
eventually out-competed by introduced commensal
house mice. These house mice were introduced sea-
sonally with garbage trucks and died out following
the termination of dumping each year. Courtney and
Fenton (1976) hypothesized that white-footed mice
initially benefited due to their omnivorous feeding
habits but that house mice had the advantage of
greater flexibility in their social structure and an
ability to maintain denser populations. It was sug-
gested that the seasonal and discontinuous use of
the dump prevented establishment of stable house
mouse populations and more broadly that the degree
of human occupation influenced the structure and
composition of the micromammalian community.
This would appear to support the inference that
high abundance of commensals in archaeological
contexts will indicate more continuous human
occupation (e.g. Hesse 1979; but see Tangri and
Wyncoll 1989).
The garbage dump situation in a seasonal
holiday retreat in temperate Canada may not be
readily comparable to that of Natufian settlements.
Nonetheless, this case highlights the current lack
of criteria for identifying the parameters of com-
mensalism in human settlement environments
and how varying levels of human occupation may
affect the ecology of associated micromammalian
communities. It may be asked, for example, whether
the association of indigenous white-footed mice with
the seasonal holiday retreat may in fact constitute a
valid case of commensalism or whether year-round
rather than seasonal occupation by only 4-5 families
and anticipated permanent establishment of house
mice would have.
A more relevant approach may be based on
comparison of ecological data on micromamma-
lian communities of settlements from sedentary
and seasonally occupied settlement environments
among small-scale societies. Such a comparison
allows us to better assess how varying signals of
commensalism may relate to specific levels of settle-
ment occupation along a sedentarization gradient.
Available data are taken from a recent study by one
of us (Weissbrod 2010a, b) on seasonal settlements
of Maasai pastoralists in southern Kenya and from
an earlier study (Misonne 1963) on agricultural
villages in the northeastern Democratic Republic
of Congo (DRC). The study of Maasai settlements
was conducted in a semi-arid region where farming

706
Lior Weissbrod et al.
is not viable and mobile herding of cattle, sheep,
and goats is the predominant economic pursuit.
Settlements can remain in use for many years (>
40 years in some cases) but are occupied on a sea-
sonal basis, a pattern that may be referred to as
seasonal sedentism (Kent 1989). Typically, a part
of or the entire population of Maasai settlements
undertakes movements with herds between wet
and dry seasons.
Maasai settlements may be considered small,
typically consisting of less than 10 houses and
having a relatively small and stable human pop-
ulation (Weissbrod 2010a, b). Important aspects
of engineering of the settlements may include the
construction and maintenance of houses and branch
fences for livestock enclosures, considerable buildup
of dung from the corralling of livestock, and grazing
by livestock and collection of materials for house
and fence construction and for fuel from surround-
ing habitats. There is little or no storage of food or
accumulation of organic food refuse. In contrast,
the villages in Misonne’s (1963) study were located
in one of the most productive agricultural regions
of Africa within the Rwenzori mountain range and
near Lakes Edward and Albert in northeastern
DRC. It was reported that the studied villages
consisted of between 10 and 100 houses and were
adjacent to crop fields (Devignat 1946; Misonne
1963). In these circumstances we may expect sig-
nificant sedentism in village occupation, relatively
large and growing human populations, and storage
of food and accumulation of organic refuse on a
large scale, which constitute important aspects of
engineering.
The two studies of pastoralists and farmers vary
greatly in terms of the environmental setting and
cultural context but are comparable in terms of the
characteristics of micromammalian communities of
the settlements due to the way in which the studies
were designed. In both cases ecological monitoring
of micromammalian communities was done inside
the settlements as well as in outside control sites al-
lowing for a controlled comparison between the two
contexts. Data on micromammalian communities
were collected through systematic trapping in six
settlements in the Kenya Maasai study (Weissbrod
2010a, b) and in 58 villages in the DRC study (Mi-
sonne 1963). In addition, in the former case control
sites were situated at a distance of 200-400 m from
each of the settlements and in the latter a series
of control sites were deployed along a transect of
increasing distance from the villages beginning
at 50 m in the adjacent crop fields and extending
to less intensely utilized habitats beyond 200 m
from the villages (see Fig. 2). The overall numbers
of captures of micromammals in the two studies
were 352 and 30,731 individuals, respectively.
Figure 2 presents bar graphs that summarize
the proportions of captures for each species of mi-
cromammal across settlement and control sites for
the agricultural villages and pastoral settlements.
A stark difference can be observed when comparing
the agricultural villages and pastoral settlements
and their respective control sites. The agricultural
villages exhibit dominance by a single species –
multimammate rats (Mastomys sp.) – and are ex-
ceedingly less heterogeneous than adjacent control
habitats. Multimammate rats have a frequency of
>98% inside the villages but drop to below 2% in the
adjacent crop fields and are absent from habitats
further away. The difference in absolute abundances
between the villages and outside fields cannot be
determined from Misonne’s (1963) published data.
Nonetheless, it was stated by the author (Misonne
1963:106) that whereas multimammate rats far
exceeded in abundance all other species within
the villages they were practically absent in the
outside fields. This pattern presents a clear case
of commensalism in the villages and accords with
the well known status of multimammate rats as
one of the leading indigenous commensals in Africa
(Kingdon 1974).
On the other hand, the pastoral settlements
have only as much as 50% of the most abundant
species – spiny mice (Acomys sp.) – and show
greater heterogeneity than in adjacent control sites
(Fig. 2). In addition, the absolute abundances of
most micromammalian species in the settlements
exceeded those in the controls. Table 2 shows a
ratio of 120:75 for spiny mice, 25:1 for multimam-
mate rats, and intermediate ratios for four other
species. Only Tatera gerbils (Tatera sp.) had fewer
numbers in the settlements than in the controls.
From a theoretical standpoint, this pattern supports
a case of environmental engineering with overall
positive effects for a wide variety of species that
may involve commensal interactions (Weissbrod
2010a). The case of pronounced commensalism of
multimammate rats in the sedentary agricultural
villages (frequency of 98%) may represent an ex-
treme along a continuum of commensalization. The
association of spiny mice with Maasai settlements
(frequency of 50%) may be considered a case of
“subtle commensalism” that can be related to the
seasonal use of the settlements and reduced mag-
nitude of habitat modification. The inference on
commensalism of spiny mice in Maasai settlements
may be further supported by data on the association

707
Commensalism: Was it Truly a Natufian Phenomenon?…
of these mice with human settlements in the Nile
Valley (Setzer 1959) and northern Kenya (Canova
and Fasola 1994).
The findings regarding seasonal and high-
ly sedentary settlement environments provide
benchmarks against which the relative frequency
of potential commensal species in Natufian micro-
mammalian assemblages may be evaluated, and a
basis for more accurate assessment of the level of
occupation at different sites or temporal phases.
Fig. 2. Comparison of living micromammalian communities in settlements of seasonally mobile pastoral-
ists in Kenya and sedentary agricultural villages in Democratic Republic of Congo (figure adapted
from Weissbrod 2010a: fig. 10.1).
Table 2. Results of live-trapping of micromammals in Maasai settlements and adjacent control sites
showing numbers of captures
No. captures
Genus: common name (scientific name) Settlements Controls
Spiny mouse (Acomys sp.) 120 75
Multimammate rat (Mastomys sp.) 25 1
Zebra mouse (Lemniscomys sp.) - 9
Elephant shrew (Elephantulus sp.) 26 3
Naked-soled gerbil (Tatera sp.) 7 15
White-toothed shrew (Crocidura sp.) 9 1
Teterillus gerbil (Teterillus sp.) 7 -
Common gerbil (Gerbillus sp.) 31 3
Narrow-footed woodland mouse (Grammomys sp.) 17 3
Total 242 110

708
Lior Weissbrod et al.
Variability in Natufian
micromammalian assemblages
Figure 3 shows relative frequencies of micro-
mammalian taxa based on minimum numbers of
individuals (MNI) in samples from Early and Late
Natufian deposits from the sites of el-Wad Terrace
(EN and LN) and Raqefet Cave (LN) in Mount
Carmel and the open air site of Eynan (FN) in the
Upper Jordan Valley. Species from 12 different
taxa are represented in the different assemblages
including 11 from the order Rodentia: common
mouse (Mus spp.), wood mouse (Apodemus spp.),
spiny mouse (Acomys cahirinus), black rat (Rattus
rattus), field vole (Microtus guentheri), water vole
(Arvicola terrestris), hamster (Mesocricetus aura-
tus), jird (Meriones tristrami), mole rat (Spalax
ehrenbergi), squirrel (Sciurus anomalus), dormouse
(Gliridae), and one from the order Insectivora: shrew
(Soricidae).
The majority of the samples in Fig. 3 show
relatively high taxonomic richness and heteroge-
neity. Of the 18 samples 16 have between six and
nine different taxa each. A measure of taxonomic
heterogeneity that is based on the Shannon-Wiener
index (Krebs 1999) shows a fairly heterogeneous
distribution of frequencies among the different taxa
in these 16 samples and Shannon-Wiener values
range between 1.443 and 1.926. Only two of the sam-
ples – one from el-Wad Terrace EN (Str. II) and the
other from Eynan (Str. 225) – have as few as three
taxa. These two taxonomically poor samples also
show relatively low taxonomic heterogeneity with
values of 1.061 and 0.842, respectively. Although the
two samples with low numbers of taxa and hetero-
geneity also have low MNI’s suggesting a sample
size effect, the correlation between heterogeneity
values and MNI is not significant (r=.19; p=.442). In
addition, these two samples with low heterogeneity
values are outliers in this relationship showing
lower than expected heterogeneity values given
their MNI (Fig. 4). This is based on z-scores, which
are higher than two standard deviations above the
mean (z=-2.017 and z=-2.738, respectively).
Taxonomic composition does not seem to vary
greatly among the samples in Fig. 3. Six of the 12
taxa including common mice, field voles, jirds, mole
rats, squirrels, and shrews are present in more than
70% of samples. Other taxa including wood mice,
spiny mice, black rats, water voles, hamsters, and
Fig. 3. Variability in taxonomic composition and relative frequencies in samples of micromammalian
assemblages from four Natufian sites/temporal phases.

709
Commensalism: Was it Truly a Natufian Phenomenon?…
dormice are less common and occur in 17-56% of
samples. Given the relatively low frequencies of
most of these latter taxa (<10%) their more restrict-
ed distribution across samples may be attributed
to chance. The occurrence of water vole only in the
Eynan samples, on the other hand, fits with recent
ecological observations showing that the geographic
distribution of the species in the region is restricted
to the Jordan Valley (Dor 1982).
Variability among the samples in terms of
the frequencies of different taxa is more marked.
Patterns of variability that can be detected in Fig.
3 may be attributed to differences between tem-
poral phases of the Natufian, differences among
geographic and environmental zones, differences
among spatial contexts within sites, and differenc-
es in mode of accumulation between cave/terrace
and open air sites. The only conspicuous difference
between samples from the EN and LN phases at
el-Wad Terrace is the relatively low frequency of
common mice in samples from the EN phase (<10%).
In addition, the frequency of field voles appears to
decrease from the site of el-Wad Terrace, which is
situated adjacent to the coastal plain, to inland
Raqefet Cave and Eynan. Squirrels appear to in-
crease in frequency along the same gradient. These
patterns may in part reflect temporal differences
given that the Eynan samples are from the FN
phase.
Perhaps one of the most interesting patterns
emerging from Fig. 3 is the importance of spatial
variability within sites. This is examined in Fig. 5
through a formal index of similarity in taxonomic
frequencies (Morisita’s index; Krebs 1999) based on
the average similarity among samples from each
of the sites/temporal phases. We may expect to see
greater similarity among samples from el-Wad
Terrace LN than among samples from the other
sites/temporal phases. This is because seven of
the samples from el-Wad Terrace LN (I-VII) were
collected from deposits that were concentrated in a
single location (ca. 0.25 m2) and were not differen-
tiated by context spatially or with depth whereas
samples from the other sites/temporal phases were
collected from different contexts or activity areas.
These include the inside and outside of a structure
(el-Wad Terrace EN), loci of burial activity and a
bedrock mortar (Raqefet), and various structures
and features (Eynan). Fig. 5 shows that the sim-
ilarity value is highest among the contextually
undifferentiated samples from el-Wad Terrace LN.
An analysis of variance of the differences among
the four means in Fig. 5 shows that it is significant
(F3,36=18.650; p=.000).
It is reasonable to expect that man-made struc-
tures may in certain circumstances provide discrete
settings for the accumulation of micromammalian
remains within Natufian sites. Some support for this
expectation can be obtained from the data in Fig. 3.
The el-Wad Terrace EN sample from inside structure
II and the Eynan sample from structure 225 reveal
the lowest taxonomic heterogeneity values among
all the Natufian samples in the study. These two
samples have especially high frequencies of remains
of mole rats (>40%), which likely form part of the
food debris in Natufian sites (see Weissbrod et al.
2005). Another of the Eynan samples from a cluster
of features (Str. 200/222/240) shows the highest
frequency of common mice among the Natufian
samples in the study (29%). Remains of common
mice that coexisted with people as commensals
within Natufian settlements and may have been
associated with concentrations of food debris could
Fig. 4. The relationship between MNI and taxonomic
heterogeneity (Shannon-Wiener index) for the
18 Natufian micromammalian assemblages.
Fig. 5. Averages of within-site similarity (Morisita’s
index) among micromammalian samples from
four Natufian sites/temporal phases. Vertical
bars represent 95% confidence intervals.

710
Lior Weissbrod et al.
have accumulated in situ within structures (see
also Yeshurun et al. herein). More detailed spatial
and taphonomic analysis of larger samples from
these sites would be required to assess the cultural
significance of the observed patterns of variability,
however.
Varying modes of accumulation of micromam-
malian remains in Natufian sites may also be re-
lated to differences in the depositional environment
between cave/terrace and open-air sites. In the
former case we can expect a significant contribution
of predators such as owls roosting in rock cavities
(Weissbrod et al. 2005) whereas in open air sites
predator activity is considered less likely and in
situ accumulation may constitute a more significant
contributor to the formation of micromammalian
assemblages (Tchernov 1984). A comparison of
samples from the open-air site of Eynan with those
from Raqefet Cave and el-Wad Terrace EN and
LN does not reveal marked differences that may
specifically relate to site type and the depositional
environment, however. Taxonomic heterogeneity
(Shannon-Wiener index) of the Eynan samples
ranging between 0.868 and 1.554 falls within the
range of the other samples from cave/terrace sites.
Moreover, intra-site differences in frequencies of
specific taxa among the samples from Eynan seem
just as marked as inter-site differences between
samples from Eynan and the cave/terrace sites. For
example, the frequency of common mice among all
Natufian samples is highest in the Eynan sample
from structures 220/222/240 whereas this taxon
is absent or has an especially low frequency in
two other Eynan samples (Str. 225 and 228/242,
respectively).
A fifth potential source of influence on variability
among Natufian micromammalian assemblages is
the ecological impact of Natufian settlements on
their immediate surroundings and the formation
of commensal micromammalian communities in
association with the settlements. The Natufian
micromammalian samples in Fig. 3 can be com-
pared with living micromammalian communities
from contemporary seasonal settlements of Maasai
pastoralists and sedentary agricultural villages in
Africa (see Fig. 2). The fossil and living communities
may be compared in terms of the level of diversity
and degree of dominance of expected commensal
taxa. None of the fossil samples exhibit the extreme
of depressed heterogeneity and dominance by a
single taxon that characterizes micromammalian
communities of the sedentary agricultural villages.
House mice are considered a key commensal
taxon of Natufian settlements. In Fig. 3, house
mice are represented by the taxonomic category
of common mouse (genus Mus) including both the
house mouse (M. musculus domesticus) and outdoor
species (M. macedonicus). The maximum relative
frequency of this taxon among the Natufian sam-
ples is approximately 30%. This is markedly lower
than both the frequency of spiny mice in seasonal
Maasai settlements (50%) and of multimammate
rats in sedentary agricultural villages (98%). Except
for five samples with especially low frequencies or
absence of common mice including the three sam-
ples of el-Wad Terrace EN and two of the Eynan
samples, all Natufian samples have frequencies of
common mice ranging between 10 and 30%.
Discussion: An evaluation of the case
for Natufian commensalism
In this section we assess the variability among
micromammalian assemblages from Natufian
sites in light of insights from ethnoarcheological
and ecological data on commensalism. The data
presented here on living micromammals of mobile
and sedentary communities in Africa suggests that
the frequency of a key commensal of southwest
Asia – the house mouse – in a number of important
Natufian sites is relatively low. This may be taken as
an indication of relatively low duration and intensity
of human occupation in the Natufian settlements.
Given that the majority of the analyzed samples
were collected from deposits of the late phase of the
Natufian (LN) such an indication would correspond
with expectations for increased mobility during this
phase (Bar-Yosef 2002; Belfer-Cohen and Bar-Yosef
2000; Byrd 2005; but see Weinstein-Evron 2009).
Several important factors may complicate the
direct use of micromammalian assemblages from
Natufian sites as indicators of commensalism and
human level of occupation, however. It was shown
that data on variability in micromammalian assem-
blages from Natufian sites can contain information
on geographic differences in environmental condi-
tions that occur on a larger scale than the localized
ecological impact of settlements on their immediate
surroundings. Additional influencing factors include
the accumulation of remains from food-related ac-
tivities involving micromammals. This does not rule
out the potential for preservation of information on
ecological processes such as commensalism, which
are expected to have occurred at the local scale of
the Natufian settlement. It does imply, however,
that micromammalian assemblages from Natu-
fian sites represent palimpsests of remains that
accumulated by different means and represent

711
Commensalism: Was it Truly a Natufian Phenomenon?…
environmental, ecological, and cultural processes
occurring at varying spatial and temporal scales.
An important fact to consider is that a substan-
tial proportion of micromammalian assemblages
from Natufian sites and particularly those situated
in cave/terrace settings is likely to have been accu-
mulated by predators such as owls (Weissbrod et al.
2005). It is unlikely that the hunting ranges of var-
ious predators contributing micromammalian prey
remains to Natufian deposits were confined to the
settlements or their immediate surroundings where
the ecological impact of the settlements would have
been felt. A study by one of us (Weissbrod 2010a)
on recent prey assemblages (owls and mongoose) in
the environment of Maasai settlements in southern
Kenya has shown marked differences in taxonomic
composition between the prey assemblages and
living micromammalian communities from inside
the settlements. Several taxa that were frequent-
ly recorded through trapping in the settlements
were either absent or under-represented in the
prey assemblages of predators, which occurred in
proximity (< 1 km) to the settlements. This suggests
that prey assemblages deposited within small-scale
settlements or their vicinity will represent the wider
environmental background of the settlements and
the landscape rather than the localized ecological
impact of the settlements (Weissbrod 2010a).
Other such recent prey assemblages of owls
from Israel provide a useful comparison to Natu-
fian micromammalian assemblages. Dor (1982)
analyzed 22 prey assemblages that were collected
during the first half of the twentieth century from a
wide range of geographic regions in the central and
northern parts of the country. These assemblages
show considerable variability in the frequency of
common mice ranging between 0 and 75% (Dor
1982). Of the 22 assemblages only three had a fre-
quency of common mice exceeding 30%, however,
one from the Judean Mountains of central Israel
(Har-Tov) and two from the Sea of Galilee in the
north (Dgania and Tiberias). Additional assemblag-
es collected more recently in settlement settings
in Israel including a suburban site (Charter et al.
2007) and an agricultural village and crop fields
(Charter et al. 2009) show a more restricted range
of frequencies of 30-50%. This is decidedly greater
than the frequencies of common mice in the Natufian
samples (10-30%). Trapping studies have also shown
that especially high frequencies of common mice of
50-90% can occur in secondary scrubland habitats
(e.g. garigue), post-fire habitats, homogeneous tree
plantations, and crop fields (Kahila Bar-Gal et al.
2000; Kupstein 2001; Lehmann and Perevolotsky
1992). Relatively intact scrub forest of various
formations including oak (Quercus calliprinos),
mock privet (Phillyrea media) or carob-pistacio
(Ceratonia siliqua-Pistacia lentiscus) in the Mount
Carmel region support intermediate frequencies
of common mice of between 20 and 40% (Kupstein
2001; Lehmann and Perevolotsky 1992).
If we view Natufian micromammalian assem-
blages through the taphonomic lens of predator
accumulation we may see evidence for environ-
mental conditions at the broad spatial scale of the
landscape. From this perspective, common mouse
frequencies in Natufian contexts (10-30%) corre-
spond to what we would expect in environments
little disturbed by human activities (see also Lev-Ya-
dun and Weinstein-Evron 2005). The EN samples
included in the study from el-Wad Terrace with
especially low frequencies of common mice may be
too few and represent a still limited extent of the
spatial variability in the excavated area to allow a
reliable comparison with the LN and FN samples.
In addition, it is important to keep in mind that
remains of certain taxa in the assemblages may
mainly represent food debris. Extracting remains
of such taxa from the analysis will alter the rela-
tive frequencies. For example, extracting mole rats
and squirrels from the computations increases the
frequencies of common mice to 40% in one of the
samples from Raqefet Cave (Loc. III) and to 50%
in two of the Eynan samples (Structures 235 and
200/222/240). In the remainder of the assemblag-
es common mouse frequencies remain below or
increase to only slightly above 30% in this case.
Extracting certain taxa from frequency calculations
should be based on more detailed taphonomic data
from each of the assemblages, however.
Tchernov (1984) reported on relatively high
frequencies of common mice from two Natufian
sites: Hayonim Cave – 48% and Eynan – 39%. In
both instances the report does not differentiate
chronologically the deposits containing the assem-
blages and it is unclear whether they represent the
Natufian as a whole or one or more of the separate
phases. It is therefore difficult to compare these
earlier data with our own or to interpret them in
terms of temporal developments in levels of human
settlement occupation within the Natufian.
An additional taphonomic scenario for the for-
mation of micromammalian assemblages in Natu-
fian sites is the in situ accumulation of remains of
taxa that coexisted with people within and in the
immediate surroundings of Natufian settlements.
Such in situ assemblages can be expected to reveal
more direct information on the ecological impact of

712
Lior Weissbrod et al.
Natufian settlements at the local scale and in turn
provide indications on commensalism and the level
of human settlement occupation. Identification of
in situ micromammalian assemblages in Natufian
sites and their separation from predator accumulat-
ed assemblages or human food debris will depend on
a combination of taphonomic criteria and examina-
tion of the spatial association of the remains with
discrete contexts of accumulation such as structures
or various installations where the introduction of
predator derived material is less likely. Relevant
taphonomic criteria can include the presence or
absence of predator modifications or charring of
skeletal elements due to food preparation activities
on the one hand and a high level of completeness
in representation of skeletal elements and/or high
survivorship of individuals which may characterize
in situ deposition of commensal animal remains,
on the other (see Weissbrod et al. 2005). Data from
our study showing considerable variability in mi-
cromammalian assemblages within Natufian sites
and especially in relation to structures supports
such a focus on identification of in situ assemblages
and on their implications for cultural and social
activities.
Conclusions
We examined variability in micromammalian
assemblages from Natufian sites to assess the
potential for preservation of information on the
ecological impact of Natufian settlements and the
level of human occupation of the settlements. Our
examination shows that such assemblages can
represent highly complex palimpsests of data on
multiple environmental and cultural processes that
may be analyzed at varying spatial and temporal
scales. The two main findings of this study include
the indication for a low level of human disturbance
of the environment at the landscape scale during
the Late Natufian and the significance of spatial
variability in taxonomic composition and hetero-
geneity within Natufian sites for investigating the
question of commensalism and human mobility
versus sedentism. We suggest a framework for
the interpretation of information on the ecological
impact of Natufian settlements combining consid-
erations of taxonomic diversity, taphonomy, and
spatial association of the remains within sites.
Thus, the combination of low taxonomic diversity,
indication for in situ accumulation of the remains,
and association of the remains with living floors
inside structures, graves, bedrock mortars, hearths
and possibly other engineered features may be used
to distinguish assemblages containing information
on the ecological impact of settlements at a localized
spatial scale from other assemblages which were
accumulated by predators and represent wider
environmental variability at the landscape scale.
Continued research on the question of commen-
salism and changing level of human settlement
occupation in the Natufian should focus on identi-
fication of in situ accumulated micromammalian
assemblages and analysis of variability on a larger
scale. This should encompass additional material
from Early Natufian contexts and other contexts
from periods both preceding and succeeding the
Natufian. Such an approach is currently needed if
we are to establish a more detailed understanding
of both evolutionary developments and variability
within Natufian settlement systems.
Acknowledgments
We wish to thank the editors, Ofer Bar-Yosef
and François Valla, for inviting us to contribute
to this important volume. We extend a special
thanks to François Valla and Hamoudi Khalaily
for providing micromammalian material from
their excavations at the site of Eynan. The study of
variability in micromammalian assemblages from
Natufian sites was conducted during the tenure
of L.W. at the Zinman Institute of Archaeology
supported by a post-doctoral research fellowship
from the University of Haifa. L.W. would also
like to thank the Kenya government, Ministry of
Science and Technology, the National Museums of
Kenya, and Maasai residents of Eselenkei Group
Ranch in Kajiado District, Kenya for allowing his
study of the living micromammalian communities
of Maasai settlements. His research in Kenya was
supported in part by a grant from the National
Science Foundation (no. BCS-0536507).
References Cited
Andrzejewski, R., Babińska-Werka, J., Gliwicz, J.
and J. Goszczyński
1978 Synurbization processes in population of
Apodemus agrarius I. Characteristics of
populations in an urbanization gradient.
Acta Theriologica 23:341-358.
Auffray, J-C., Tchernov, E., Bonhomme, F., Heth,
G., Simson, S. and E. Nevo
1990 Presence and ecological distribution of
Mus “spretoides” and Mus musculus do-

713
Commensalism: Was it Truly a Natufian Phenomenon?…
mesticus in Israel, circum-Mediterranean
vicariance in the genus Mus. Zeitschrift
für Säugetierkunde 55:1-10.
Auffray, J-C., Tchernov, E. and E. Nevo
1988 Origine du commensalisme de la souris
domestique (Mus musculus domesticus)
vis-à-vis de l’homme. Comptes- Rendus
de l’Académie des Sciences Paris 307:517-
522.
Bar-Matthews, M., Ayalon, A., Gilmour, M.,
Matthews, A. and C. J. Hawkesworth
2003 Sea-land oxygen isotopic relationships
from planktonic foraminifera and spe-
leothems in the Eastern Mediterra-
nean region and their implication for
paleorainfall during interglacial inter-
vals. Geochimica et Cosmochimica Acta
67:3181-3199.
Bar-Oz, G., Dayan, T., Kaufman, D. and M.
Weinstein-Evron
2004 The Natufian economy at el-Wad Terrace
with special reference to gazelle exploita-
tion patterns. Journal of Archaeological
Science 31:217-231.
Bar-Yosef, O.
1983 The Natufian in the southern Levant.
In The Hilly Flanks and Beyond: Essays
on the Prehistory of Southwestern Asia,
edited by T. C. Young, P. E. L. Smith
and P. Mortensen, pp. 11-42. Oriental
Institute, University of Chicago,
Chicago.
1991 The archaeology of the Natufian layer at
Hayonim Cave. In The Natufian Culture
in the Levant, edited by O. Bar-Yosef
and F. R. Valla, pp. 81-92. International
Monographs in Prehistory, Ann Arbor.
2002 Natufian and the Early Neolithic: Social
and economic trends in southwestern
Asia. In Examining the Farming/Lan-
guage Dispersal Hypothesis, edited by P.
Bellwood and C. Renfrew, pp. 113-126.
McDonald Institute for Archaeological
Research, Cambridge.
Bar-Yosef, O. and A. Belfer-Cohen
1989 The origins of sedentism and farming
communities in the Levant. Journal of
World Prehistory 3:447-498.
Bar-Yosef, O. and E. Tchernov
1966 Archaeological finds and the fossil
faunas of the Natufian and microlithic
industries at Hayonim Cave (Western
Galilee, Israel). Israel Journal of Zoology
15:121-140.
Belfer-Cohen, A. and O. Bar-Yosef
2000 Early sedentism in the Near East: A
bumpy road to village life. In Life in
Neo lithic Farming Communities: Social
Organization, Identity, and Differentiation,
edited by I. Kujit, pp. 19-37. Kluwer
Academic Publishers, New York.
Berry, R. J.
1981 Town mouse, country mouse: adaptation
and adaptability in Mus Domesticus (M.
musculus domesticus). Mammal Review
11:92-130.
Binford, L. R.
1968 Post-Pleistocene adaptations. In New
Perspectives in Archaeology, edited by S.
R. Binford and L. R. Binford, pp. 313-341.
Aldine, Chicago.
Bouchud, J.
1987 La Fauna du Gesement Natufien de
Mallaha (Eynan), Israel. Mémoires et
Travaux du Centre de Recherche Français
de Jérusalem 4. Association Paléorient,
Paris.
Boursot, P., Auffray, J-C., Britton-Davidian, J.
and F. Bonhomme
1993 The evolution of house mice. Annual
Review of Ecology and Systematics
24:119-152.
Braithwaite, R. W.
1980 The ecology of Rattus lutreolus III*. The
rise and fall of a commensal population.
Australian Wildlife Research 7:199-215.
Byrd, B. F.
1989 The Natufian: settlement variability and
economic adaptation in the Levant at the
end of the Pleistocene. Journal of World
Prehistory 3:159-197.
2005 Reassessing the emergence of village life
in the Near East. Journal of Archaeolog-
ical Research 13:231-290.
Canova, L. and M. Fasola
1994 Population density and diet of the Spiny
Mouse Acomys cf. cahirinus in a desertic
area of northern Kenya. Revue d’Écologie
49:87-90.
Charter, M., Izhaki, I., Shapira, L. and Y. Leshem
2007 Diets of urban breeding barn owls (Tyto
alba) in Tel Aviv, Israel. The Wilson
Journal of Ornithology 119:484-485.
Charter, M., Izhaki, I., Meyrom, K., Motro, Y. and
Y. Leshem
2009 Diets of barn owls differ in the same
agricultural region. The Wilson Journal
of Ornithology 121:378-383.

714
Lior Weissbrod et al.
Chernousova, N. F.
2001 Specific features of the dynamics of
murine rodent communities under the
effects of urbanization. 1. Dynamics of
species composition and abundance.
Russian Journal of Ecology 32:122-125.
2002 Specific features of the dynamics of
murine rodent communities under the
effects of urbanization: 2. Maintenance
of population size. Russian Journal of
Ecology 33:23-27.
Courtney, P. A. and M. B. Fenton
1976 The effects of a small rural garbage dump
on populations of Peromyscus Leucopus
Rafinesque and other small mammals.
Journal of Applied Biology 13:413-422.
Devignat, R.
1946 Aspects de l’ épidémiologie de la peste au
Lac Albert. Annales de la Société Belge
de Médecine Tropicale 26:13-54.
Dickman, C. R.
1992 Commensal and mutualistic interactions
among terrestrial vertebrates. Trends in
Ecology & Evolution (Amsterdam) 7:194.
2006 Species interactions: Indirect effects.
In Ecology: An Australian Perspective,
edited by P. Attiwill and B. A. Wilson,
pp. 303-315. Oxford University Press,
Oxford.
Dor, M.
1982 The barn owl as an aid to science. The
Torgos 5:29-35 (in Hebrew).
Edwards, P. C.
1989 Problems of recognizing earliest seden-
tism: The Natufian example. Journal of
Mediterranean Archaeology 2:5-48.
Evans, J. G.
1978 An Introduction to Environmental
Archaeology. Paul Elek, London.
Flannery, K. V.
1972 The origins of the village as a settlement
type in Mesoamerica and the Near East:
A comparative study. In Man, Settlement,
and Urbanism, edited by P. J. Ucko, R.
Tringham and G. W. Dimbleby, pp. 23-53.
Duckworth, London.
Ganem, G.
1991 A comparative study of different popula-
tions of Mus musculus: Emotivity as an
index of adaptation to commensalism.
Comparative Biochemistry and Physiol-
ogy 99A:531-536.
1993 Ecological characteristics of Robertsonian
populations of the house mouse. Is their
habitat relevant to their evolution?
Mammalia 57:349-357.
Garrod, D. A. E.
1957 The Natufian culture: the life and
economy of a Mesolithic people in the
Near East. Proceedings of the British
Academy 43:211-227.
Garrod, D. A. E. and D. M. A. Bate
1937 The Stone Age of Mount Carmel. Claren-
don Press, Oxford.
Grosman, L.
2003 Preserving cultural traditions in a period
of instability: The Late Natufian of the
Hilly Mediterranean Zone. Current
Anthropology 4:571-580.
Grosman, L. and N. D. Munro
2007 The sacred and the mundane: Domestic
activities at a Late Natufian burial site
in the Levant. Before Farming 4: article
4.
Grosman, L., Munro, N. D. and A. Belfer-Cohen
2008 12,000-year-old Shaman burial from the
southern Levant (Israel). Proceedings
of the National Academy of Sciences
105:17665-17669.
Henry, D. O.
1985 Preagricultural sedentism: The Natu-
fian example. In Prehistoric Hunt-
er-Gatherers: The Emergence of Cultural
Complexity, edited by D. T. Price and J.
A. Brown, pp. 365-384. Academic Press,
San Diego.
Hesse, B.
1979 Rodent remains and sedentism in the
Neolithic: Evidence from Tepe Ganj
Dareh, Western Iran. Journal of Mam-
malogy 60:856-857.
Jones, C. G., Lawton, J. H. and M. Shachak
1994 Organisms as ecosystem engineers. Oikos
69:373-386.
1997 Positive and negative effects of organisms
as physical ecosystem engineers. Ecology
78:1946-1957.
Kahila Bar-Gal, G., Aviel, S., Aram, E., Ginsburg,
Z. and E. Tchernov
2000 Introduction of barn owl (Tyto alba) as an
effective biological control agent of rodent
populations in Israel. Israel Journal of
Zoology 46:373.
Kaufman, D.
1992 Hunter-gatherers of the Levantine Epi-
palaeolithic: The socioecological origins
of sedentism. Journal of Mediterranean
Archaeology 5:165-201.

715
Commensalism: Was it Truly a Natufian Phenomenon?…
Kent, S. (editor)
1989 Farmers as Hunters: The Implications of
Sedentism. Cambridge University Press,
Cambridge.
Kingdon, J.
1974 East African Mammals: An Atlas of
Evolution in Africa. Vol II, Part B: Hares
and Rodents. Academic Press, New York.
Krebs, C. J.
1999 Ecological Methodology. Benjamin/
Cummings, Menlo Park.
Kupstein, H.
2001 The use of small-mammals community
composition as a bio-indicator to the
interference extent in ecological systems
hosting outdoor recreation activities. M.A.
thesis, University of Haifa.
Lehmann, T. and A. Perevolotsky
1992 Small mammals in coniferous plantations
and native environment in southern Mt.
Carmel, Israel. Mammalia 56:575-585.
Lengyel, G. and F. Bocquentin
2005 Burials of Raqefet Cave in the context of
the Late Natufian. Journal of the Israel
Prehistoric Society - Mitekufat Haeven
35:271-284.
Lengyel, G., Nadel, D., Tsatskin, A., Bar-Oz,
G., Bar-Yosef Mayer, D. E., Be’eri, R. and I.
Hershkovitz
2005 Back to Raqefet Cave, Mount Carmel,
Israel. Journal of the Israel Prehistoric
Society - Mitekufat Haeven 35:245-270.
Leroi-Gourhan, Arl.
1984 L’environnement de Mallaha (Eynan) au
Natoufien. Paléorient 10:103-105.
Lev-Yadun, S. and M. Weinstein-Evron
2005 Modeling the influence of Natufian
el-Wad on the forest of Mount Carmel.
Journal of the Israel Prehistoric Society
- Mitekufat Haeven 35:285-298.
Marder, O., Yeshurun, R., Smithline, H.,
Ackermann, O., Bar-Yosef Mayer, D. E., Belfer-
Cohen, A., Grosman, L., Hershkovitz, I., Klein N.
and L. Weissbrod
herein Hof Shahaf: A New Natufian Site on
the Shore of Lake Kinnere, edited by O.
Bar-Yosef and F. R. Valla, International
Monographs in Prehistory, Ann Arbor.
Misonne, X.
1963 Les Rongeurs du Ruwenzori et des Régions
Voisines. Institut des Parcs Nationaux du
Congo et du Rwanda, Brussels.
Nadel, D., Lengyel, G., Bocquentin, F., Tsatskin,
A., Rosenberg, D., Yeshurun, R., Bar-Oz, G.,
Bar-Yosef Mayer, D.E., Beeri, R., Conyers, L.,
Filin, S., Hershkovitz, I., Kurzawska, A. and L.
Weissbrod
2008 The Late Natufian at Raqefet Cave: The
2006 Excavation Season. Journal of the
Israel Prehistoric Society - Mitekufat
Haeven 38:59-131.
Nadel, D. and G. Lengyel
2009 Human-made bedrock holes (mortars
and cupmarks) as a Late Natufian social
phenomenon. Archaeology, Ethnology and
Anthropology of Eurasia 37:37-48.
Nadel, D., Lengyel, G., Cabellos, T., Bocquentin,
F., Rosenberg, D., Yeshurun, R., Brown-Goodman,
R., Tsatskin, A., Bar-Oz, G. and S. Filin
2009 The Raqefet Cave 2008 excavation season.
Journal of the Israel Prehistoric Society
- Mitekufat Haeven 39:21-62.
Perrot, J.
1968 Premiers villages de Syrie et de Pales-
tine. Comptes-Rendus des Séances de
l’Académie des Inscriptions et Belles-
Lettres 112:161-177.
Pocock, M. J. O., Searle, J. B. and P. C. L. White
2004 Adaptations of animals to commensal
habitats: population dynamics of house
mice Mus musculus domesticus on farms.
Journal of Animal Ecology 73:878-888.
Rajabi-Maham, H., Orth, A. and F. Bonhomme
2008 Phylogeography and postglacial expan-
sion of Mus musculus domesticus inferred
from mitochondrial DNA coalescent,
from Iran to Europe. Molecular Ecology
17:627-641.
Reitz, E. J. and E. S. Wing
1999 Zooarchaeology. Cambridge University
Press, Cambridge.
Sakai, A. K., Allendorf, F. W., Holt, J. S., Lodge, D.
M. M.,Jane, With, K. A., Baughman, S., Cabin, R.
J., Cohen, J. E., Ellstrand, N. C., McCauley, D. E.,
O’Neil, P., Parker, I. M., Thompson, J. N. and S. G.
Weller
2001 The population biology of invasive
species. Annual Review of Ecology and
Systematics 32:305-332.
Setzer, H. W.
1959 The spiny mice (Acomys) of Egypt.
Journal of the Egyptian Public Health
Association 34:93-101.
Smith, B. D.
2007 Human niche construction and the be-
havioural context of plant and animal
domestication. Evolutionary Anthropol-
ogy 16:188-199.

716
Lior Weissbrod et al.
Stein, M., Torfstein, A., Gavrieli, I. and Y. Yechieli
2009 Abrupt aridities and salt deposition in
the post-glacial Dead Sea and their North
Atlantic connection. Quaternary Science
Reviews 29:567-575.
Tangri, D. and G. Wyncoll
1989 Of mice and men: Is the presence of
commensal animals in archaeological
sites a positive correlate of sedentism?
Paléorient 15/1:85-94.
Tchernov, E.
1984 Commensal animals and human seden-
tism in the Middle East. In Animals and
Archaeology, 3: Early Herders and Their
Flocks, edited by J. Clutton-Brock and C.
Grigson, pp. 91-115. BAR International
Series 202. Oxford.
1991a Biological evidence for human sedentism
in Southwest Asia during the Natufian.
In The Natufian Culture in the Levant,
edited by O. Bar-Yosef and F. R. Valla,
pp. 315-340. International Monographs
in Prehistory. Ann Arbor.
1991b Of mice and men. Biological markers for
long-term sedentism: A reply. Paléorient
17/1:153-160.
Valla, F. R.
1998 Natufian seasonality? A guess. In Season-
ality and Sedentism: Archaeological Per-
spectives from Old and New World Sites,
edited by T. R. Rocek and O. Bar-Yosef,
pp. 93-108. Peabody Museum Bulletin
6. Peabody Museum of Archaeology and
Ethnology, Harvard University, Cam-
bridge.
Valla, F. R., Bar-Yosef, O., Smith, E., Tchernov, E.
and J. Desse
1986 Un nouveau sondage sur la terrasse d’el
Ouad, Israël. Paléorient 12/1:31-36.
Valla, F. R. and H. Khalaily
1997 The first sedentary peoples in Israel:
Mallaha (Eynan) 1996. Bulletin du Centre
de Recherche Français de Jérusalem 1:72-
82.
Valla, F. R., Khalaily, H., Samuelian, N.,
Bocquentin, F., Delage, C., Valentin, B., Plisson,
H., Rabinovich, R. and A. Belfer-Cohen
1998 Le Natufien final et les nouvelles fouilles
à Mallaha (Eynan), Israël, 1996-1997.
Journal of the Israel Prehistoric Society
- Mitekufat Haeven 28:105-176.
Valla, F. R., Khalaily, H., Samuelian, N. and F.
Bocquentin
2002 From foraging to farming: The contribu-
tion of the Mallaha (Eynan) excavations,
1996–2001. Bulletin du Centre de Recher-
che Français de Jérusalem 10:71-90.
Valla, F. R., Khalaily, H., Valladas, H.,
Kaltnecker, E., Bocquentin, F., Cabellos, T., Bar-
Yosef Mayer, D. E., Le Dosseur, G., Regev, L.,
Chu, V., Weiner, S., Boaretto, E., Samuelian, N.,
Valentin, B., Delerue, S., Poupeau, G., Bridault,
A., Rabinovich, R., Simmons, T., Zohar, I.,
Ashkenazi, S., Delgado Huertas, A., Spiro, B.,
Mienis, H., Rosen, A. M., Porat, N. and A. Belfer-
Cohen
2007 Les fouilles de Ain Mallha (Eynan) de
2003 à 2005: Quatrième rapport prélim-
inaire. Journal of the Israel Prehistoric
Society - Mitekufat Haeven 37:135-383.
Weinstein-Evron, M.
1998 Early Natufian el-Wad Revisited. ERAUL
77, Liège.
2009 Archaeology in the Archives: Unveiling
the Natufian Culture of Mount Carmel.
Brill and American School of Prehistoric
Research Monograph Series, Boston.
Weinstein-Evron, M., Kaufman, D., Bachrach,
N., Bar-Oz, G., Bar-Yosef Mayer, D. E., Chaim, S.,
Druck, D., Groman-Yaroslavski, I., Hershkovitz,
I., Liber, N., Rosenberg, D., Tsatskin, A. and L.
Weissbrod
2007 After 70 years: new excavations at the el-
Wad Terrace, Mount Carmel. Journal of
the Israel Prehistoric Society - Mitekufat
Haeven 37:37-134.
Weinstein-Evron, M., Kaufman, D. and R.
Yeshurun
herein Spatial Organization of Natufian el-Wad
through Time: Combining the Results
of Past and Present Excavations. In
The Natufian Culture in the Levant II,
edited by O. Bar-Yosef and F. R. Valla,
International Monographs in Prehistory,
Ann Arbor.
Weissbrod, L.
2010a The small animals of Maasai settlements:
Ethnoarchaeoogical investigations of the
commensalism model. Ph.D. dissertation,
Washington University in Saint Louis,
Saint Louis.
2010b Biological indicators of occupation inten-
sity: An environmental ethnoarchaeology
of Maasai settlements. In The Archaeology
of Anthropogenic Environments, edited
by R. M. Dean, pp. 295-320. Center for
Archaeological Investigations, Southern
Illinois University, Carbondale.

717
Commensalism: Was it Truly a Natufian Phenomenon?…
Weissbrod, L., Dayan, T., Kaufman, D. and M.
Weinstein-Evron
2005 Micromammal taphonomy of el-Wad Ter-
race, Mount Carmel, Israel: distinguish-
ing natural from cultural depositional
agents in the Late Natufian. Journal of
Archaeological Science 32:1-17.
Yeshurun, R., Bar-Oz, G., Kaufman, D. and M.
Weinstein-Evron
herein Domestic Refuse Maintenance in the
Natufian: Faunal Evidence from el-Wad
Terrace, Mount Carmel, edited by O.
Bar-Yosef and F. R. Valla, International
Monographs in Prehistory, Ann Arbor.