We investigate changes in small-mammal richness and diversity in southwestern Europe (Iberian Peninsula) during the late Pleistocene–Holocene transition in order to evaluate whether they follow a climatic pattern or are predominantly determined by human impact, especially after the emergence of agriculture in the Neolithic period. We selected 6 late Pleistocene and Holocene sites that correspond to 18 different layers dated to between ca. 22 and 3 kyr B.P. Using indices of species richness and evenness diversity, we show that climate played an important role at some sites during the late Pleistocene and at the beginning of the Holocene, in that the richness and diversity of small mammals were closely related to the mean annual temperatures and landscape changes, and varied according to the different climatic fluctuations detected (Heinrich Event 1, Bølling-Allerød, and Preboreal-Boreal). However, at the beginning of the mid-Holocene, the small-mammal richness and diversity no longer seem to follow any kind of climatic pattern, and the observed changes in some studied sites are more closely related to human activities. By contrast with similar studies carried out in other parts of the world, the changes in diversity in the Iberian Peninsula do not seem to follow a constant pattern during the late Pleistocene and beginning of the Holocene. Some of the changes detected appear to be related to climate (late Pleistocene), and others appear to be related to human influence (Holocene) on the landscape.


Analyses of the species diversity of small nonflying mammals are common in ecological studies describing present-day communities (McCain, 2004). The determinants of the species richness of present-day nonflying small mammal associations are characterized by numerous variables, including temperature, precipitation, productivity, habitat heterogeneity, area and restrictions peculiar to the distribution of the species, as well as the mid-domain effect or topography, and human influence (McCain, 2004). All these variables can be simplified into three (Fløjgaard et al., 2011): biogeographical history, habitat heterogeneity, and human influence.

The literature on the above-mentioned determinants of patterns of species richness and diversity includes a vast number of hypotheses. Climate has found the most general support as the main driver of species richness and diversity at large scales (e.g., Andrews and O’Brien, 2000; Hawkins et al., 2003; Tognelli and Kelt, 2004; Blois et al., 2010), but there is also great support for habitat heterogeneity (e.g., Kerr and Packer, 1997; O’Brien et al., 2000) and long-term historical processes of human influence (e.g., Araújo et al., 2008).

Taking these background considerations into account, we try for the first time to link different, well-dated late Pleistocene to Holocene sites in the Iberian Peninsula with small-mammal studies in order to observe the variation in the richness and diversity of species over this period of time. We compare our data with global climate changes, shown by the 18O isotopic curve, to ascertain whether the different climatic moments in the late Pleistocene and Holocene are associated with an increase or a decrease in small-mammal richness and diversity, and to determine how and when human influence came to have an effect on small-mammal faunas. The final aim of this paper is to establish whether richness and diversity in the late Pleistocene–Holocene follow a climatic pattern, as suggested by Blois et al. (2010), or whether there are other factors that influence the changes in small-mammal richness and diversity during these periods.


Study Area

We selected 6 late Pleistocene and Holocene sites (caves) throughout the Iberian Peninsula (Fig. 1) that correspond to 18 different layers dated to between ca. 22 and 3 kyr B.P. (see Appendix DR1 in the GSA Data Repository1).

Small-Mammal Assemblage

The systematic attributions and descriptions of the small-mammal material from these different sites have been published (Cuenca-Bescós et al., 2009, 2010; Bañuls and López-García, 2010; Bañuls et al., 2012; López-García et al., 2008, 2010a, 2010b, 2011; Appendices DR1 and DR2 in the Data Repository).

14C Dating

The dates were obtained by radiocarbon 14C accelerator mass spectrometry in different laboratories (Bañuls et al., 2012; López-García et al., 2010b; Straus and Gonzáles Morales, 2003; Vaquero et al., 2009; Vergés et al., 2002), and were calibrated and correlated with the 18O isotopic Hulu curve by means of the CalPal program (Weninger et al., 2007) at 2σ (Appendix DR3).

Determining Richness and Diversity

Despite the large number of indices proposed for evaluating the biodiversity of a sample, species richness and evenness are fundamental for assessing the homogeneity of an environment (Magurran and McGill, 2011). Evenness is a diversity index that is used to quantify how equal the various communities are numerically. The evenness of a community can be represented by the Simpson index of diversity, which is also equivalent to the probability of interspecific encounter (Appendices DR1 and DR4) (Simpson, 1949; Blois et al., 2010): i.e., Simpson index of diversity = 1 – Σ(pi2), where pi is the proportion of individuals in the ith species.

The richness of species in a sample is a function of the sample size. There are several methods of solving this problem; rarefaction is one of the most commonly used (Hammer and Harper, 2006). Rarefaction provides a measure of species diversity that is robust to sample size effects, permitting comparison between communities where, for example, the densities of animals are very different (Hurlbert, 1971; Simberloff, 1972) (Appendices DR1 and DR4).

Determining the Climatic Parameters

The paleoclimatic parameters were obtained using the mutual climatic range (MCR) method, by means of the small-mammal associations (e.g., Agustí et al., 2009; Blain et al., 2009; López-García et al., 2010b) (Appendix DR1).

Determining the Landscape Parameters

The landscape parameters were obtained by the habitat weighting method (Evans et al., 1981; Andrews, 2006; Appendix DR1).


In the light of the relationship between the species and the 18O curve, diversity and species richness generally seem to follow a climate-based pattern for the late Pleistocene to early Holocene. The diversity values vary during the Last Glacial Maximum (LGM), Heinrich Event 1 (H1), and Bølling-Allerød (B/A) periods, associated with the fluctuations in mean annual temperature (MAT) and landscape (Appendix DR5). Moreover, a decline in species diversity is observed during the B/A (levels 106–102.1) and Preboreal-Boreal (Pr/B; level 10.1); this appears to be related to Interstadial 1 (IS1) in the first case, and the beginning of Holocene warming in the second case (Fig. 2).

For the species that are well represented in the sites studied, it can be seen that the species associated with moist environments (Palomo and Gisbert, 2005), such as Arvicola terrestris, show greater representation during the late Pleistocene than the Holocene (Appendices DR6 and DR7). By contrast, the species associated with dry environments (Palomo and Gisbert, 2005), such as Crocidura russula, and nonmoist environments, such as Microtus arvalis, with the exception of P1 (see Fig. 2), show greater representation from the beginning of the Holocene than during the late Pleistocene. Furthermore, the representation of Apodemus sylvaticus, a species associated with woodland resources (Palomo and Gisbert, 2005), indicates the predominance during the late Pleistocene of open environments and an increase in tree cover throughout the mid-Holocene (Appendices DR6–DR8; Fig. 3).

However, we have not observed any trend in diversity that bears a relation to climate in the mid-Holocene (Appendix DR5). The variations detected in richness and evenness in the mid-Holocene (from the Neolithic period on) seem likely to be associated with human activities at the studied sites. All of them are sites with human occupation during the Neolithic period (as shown by animal dung accumulations; these levels have basically been formatted by anthropic sedimentation) (Vergés et al., 2002; Straus and Gonzáles Morales, 2003; Oms et al., 2008). However, the studied sites are less anthropized during the Bronze Age, showing just sporadic human occupation (burial or pit sites) (Vergés et al., 2002; Straus and Gonzáles Morales, 2003; Oms et al., 2008). The relation between diversity and human influence is reflected also in the evolution of the most abundant taxa. In the Neolithic and Chalcolithic periods, there is a relative increase in species such as Eliomys quercinus and Terricola sp., and the percentage representation of these species follows the fluctuations in evenness (Appendices DR6–DR8; Fig. 3). This can be explained by the fact that the presence of human populations is associated with the rearing of livestock, and this activity destroyed the high grasslands, a habitat type avoided by E. quercinus (Bertolino and Currado, 2001), and also that the presence of livestock would have increased the populations of geophytes, which are the main food for species of Terricola (Noy-Meir and Oron, 2001).

Climatic and Landscape Change

Climate has been considered the most general factor influencing changes in species richness and diversity (e.g., Andrews and O’Brien, 2000; Hawkins et al., 2003; Tognelli and Kelt, 2004; Blois et al., 2010). Throughout the late Pleistocene and early Holocene, the climate was the dominant factor controlling paleoenvironmental conditions, whereas humans played an unimportant role (Zolitschka et al., 2003). As recently argued for California (United States) by Blois et al. (2010), in the late Pleistocene–early Holocene of the Iberian Peninsula, the MAT and landscape fluctuations (Appendix DR5; Fig. 3) were closely related to small-mammal diversity. Small-mammal communities were highly even in some sites and taxonomically rich during the cold and relatively humid late Pleistocene moments (LGM and H1) and, as suggested by Blois et al. (2010), the evenness declined sharply with the warming of the B/A period (Liu et al., 2009). In our case, this coincides with the first turnover pulse (Interstadial 1) and the warming of the Pr/B period (Fig. 2).

Human Influence

Even during what is known as the mid-Holocene climate optimum (ca. 8–6 kyr B.P.), when central Europe was covered by closed forest, human influence on the environment does not seem to have been a determinant factor influencing species diversity (Zolitschka et al., 2003). It is from the advanced Neolithic and the advent of agriculture onward that human activity in the environment is accentuated and becomes a determinant factor in landscape formation, and thus also in the richness and diversity of species (Zolitschka et al., 2003; Carrión et al., 2010). However, evidence of the human impact on the environment and its effect on species seems to become more striking, through pollen and charcoal studies, from the Chalcolithic and Bronze Age onward (Carrión et al., 2010).

Taking these considerations into account, and contrary to Blois et al. (2010), the changes in small-mammal richness and diversity during the mid-Holocene (Neolithic period) in the Iberian Peninsula do not appear to follow a pattern related to climate. It is possible that at some of the studied sites these changes were related to the impact of humans. In the Neolithic and Chalcolithic periods, the sites studied display an increase in the representation of species considered to be synanthropic (Mistrot, 2000), and the percentage representation of these species follows the fluctuations in evenness (Fig. 3). This is due to the action of humans on the environment in the vicinity of the sites surveyed in the mid-Holocene (Neolithic period).


The study of small mammals expounded here shows that human impact on the environment in the immediate vicinity of the studied caves can be seen from the Neolithic period on. As argued previously, during the latest Pleistocene and the beginning of the Holocene climate played an important role in small-mammal diversity, constituting probably the predominant factor in the changes in micro-mammal diversity. In contrast with similar studies undertaken in other parts of the world, however, we show that the human impact in a region with as long a history as the Iberian Peninsula was an important factor in small-mammal diversity from as early as the Neolithic period, during which time landscape was modified through the implementation of agriculture.

We thank all members and directors of the excavations mentioned in this work for allowing us to study the small-mammal material. We are grateful to Ellen Thomas and three anonymous reviewers for their valuable comments on the manuscript. This paper is part of projects CGL2012-38358 and SGR2009-324.

1GSA Data Repository item 2013066, Appendices DR1–DR8 (description of study area, material and methods, standardized data of minimum number of individuals identified in the layers of the analyzed sites, layer data, richness and evenness data, relation of temperature and precipitation with evenness diversity, quantitative evolution of the most highly represented species in the 18 layers analyzed, standardized data of the most representative taxa, comparison of evenness with the percentage representation of synanthropic species, and percentage representation of the landscape in the studied site), is available online at www.geosociety.org/pubs/ft2013.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.