Faunal Dynamics of Small Mammals through the Pit Sequence
Faunal Dynamics of Small Mammals through the Pit Sequence
Abstract and Keywords
The Pit locality in the Porcupine Cave contains a stratified, circa-2-m-thick sequence of sediment that has yielded more than 7,200 identified fossils representing more than 1,500 individual animals. A minimum of 1 amphibian species, 2 reptile species, 1 bird species, and 57 mammal species have been recognized. This chapter describes and interprets the faunal dynamics through the entire sequence and for most taxa. Tracing the taxa through the sequence provides insight into the relative abundance of the small mammal community in the middle Pleistocene, how the fauna of interglacials differed from that of glacials, and how environments may have subtly fluctuated within the major climatic intervals defined by sedimentary criteria.
The Pit locality, located about 35 m inside Porcupine Cave (see figure 2.3), contains a stratified, circa-2-m-thick sequence of sediment that has yielded more than 7200 identified fossils representing more than 1500 individual animals. A minimum of 1 amphibian species, 2 reptile species, 1 bird species, and 57 mammal species have been recognized (see tables 10.9, 10.10). The fossils are spread through the top 13 of the Pit’s 15 discrete stratigraphic intervals, which are numbered 1–14, with level 8 separated into 8 and 8a (figure 23.1).
Characteristics of the sediments in the Pit have been used to divide the sequence into interglacial (levels 1–3, 6–9, probably 11) and glacial (levels 4–5, 10) deposits (Barnosky and Rasmussen, 1988; Barnosky and Bell, 2000; chapter 7). Previous studies (Barnosky and Rasmussen, 1988; Wood and Barnosky, 1994; Barnosky et al., 1996; Bell and Barnosky, 2000) pointed out the dramatic faunal change that accompanied the transition from the uppermost glacial (levels 5 and 4) to the uppermost interglacial (levels 3–1). The earlier work also recognized that, within each of these major climatic intervals, relative abundance of taxa fluctuated, and that species composition changed somewhat from lower to upper levels. With the more complete data set now available, it is possible to describe and interpret the faunal dynamics through the entire sequence and for most taxa, and that is the focus of this chapter. Tracing the taxa through the sequence leads to inferences about how the South Park mammal community assembled through the middle Pleistocene, how the fauna of interglacials differed from that of glacials, and how environments may have subtly fluctuated within the major climatic intervals defined by sedimentary criteria.
Materials and Methods
Details of the Pit excavation were provided by Barnosky and Rasmussen (1988) and Bell and Barnosky (2000), and in chapter 2. This analysis focuses on the mammal taxa, especially small mammals, because they are more adequately sampled than the other vertebrate groups. A total of 6981 identified specimens (numbers of identified specimens [NISP]) representing at least 1402 individuals (minimum numbers of individuals [MNI]) were included. Raw data are presented in tables 10.9 and 10.10.
Relative abundances of mammals were calculated using MNI. Relative abundance results based on NISP are not substantially different, although the MNI technique slightly overestimates the abundance of rare taxa. Percentages for relative abundances were calculated in Microsoft Excel 9.0.
To assess the minimum sample size at which rare taxa might be expected to appear in a given level, NISP at each level was plotted against species richness for the following groups of taxa: all Pit taxa (including reptiles, amphibians, and birds), mammals only, carnivores, rodents, lagomorphs, arvicolines, Neotoma, Spermophilus, and Cynomys. Rarefaction curves were computed by considering each stratigraphic level a single sample and using the Coleman rarefaction algorithm in Estimates 5.01 (R. Colwell, Storrs, Connecticut) (Colwell, 1997; Colwell and Coddington, 1994). One rarefaction curve included all mammal taxa; a second one was computed for only rodents and lagomorphs. Various problems arise in applying the EstimateS algorithms to paleontological samples (Barnosky and Carrasco, 2002), but the results are nevertheless useful in comparing taphonomically similar fossil deposits, as are the Pit levels.
To assess how well each level represented the expected species diversity based on the total Pit sample, the number of species expected for a given NISP based on the rarefaction analysis was plotted against the observed number of species at each stratigraphic level. If samples are perfectly homogenous (all levels equal in species diversity), the correlation approaches 1.0. Levels with anomalously high or low diversity plot as exceptionally high or low outliers, respectively. Correlation statistics were derived in StatView 5.0 (SAS Institute, Inc., Cary, North Carolina).
Results and Discussion
The fossil small mammals from the Pit very probably provide a reasonable representation of the animals that constituted the middle Pleistocene community around Porcupine Cave, given the taphonomic pathways detailed in chapter 2. Essentially, the taphonomic pathways resembled those analyzed by Hadly (1999) for Lamar Cave, Wyoming, which demonstrated remarkable fidelity to the sampling universe from which the fossils were taken. Thus the primary sampling concern, as in most paleontological samples, is whether adequate numbers of specimens exist to compare relative abundance and diversity patterns from one stratigraphic level to another.
When the Pit sample is considered as a whole, the rarefaction curve suggests that, even with nearly 7000 specimens, additional sampling can be expected to yield more species (figure 23.2). The continual rise in the curve probably reflects two factors. First, the entire sequence spans 200,000–300,000 years, each level is of slightly different geological age, and therefore some new species are added through time. This temporal effect is removed in the richness-per-level curve, which indicates that stratigraphic levels that contain more than 500 specimens gain few new species with additional sampling. The second factor contributing to the consistent rise in the rarefaction curve is that the artiodactyls, perissodactyls, and carnivorans are poorly sampled relative to the rodents and lagomorphs because of taphonomic considerations. The rarefaction curve for rodents and lagomorphs alone rises more steeply than that for all taxa, consistent with the poorer sampling of large mammals. The fact that it does not level off until more than 1000 specimens (versus 100–300 specimens for the richness-per-level-curve for rodents and lagomorphs) reflects the addition of species through time.
The richness-per-level curves for subgroups of mammals verify that the large mammals are inadequately sampled (figure 23.3). However, for rodents and lagomorphs there is no correlation between sample size and species richness above a total NISP of about 100 for rodents and 300 for lagomorphs. This also holds true at the generic level for rodents. Therefore the sample for these groups is adequate in levels 1–6 (NISP = 1580, 1442, 1020, 1154, 656, 434, respectively); marginally adequate but less than desirable in levels 7, 8, 8A, and 10 (NISP = 283, 104, 125, 119, respectively); and clearly inadequate in levels 9, 11, and 12 (NISP = 25, 36, 3, respectively).
Small-Mammal Community Composition
Taxa present in most levels of the Pit include Ochotona (pika), leporids (rabbits), Marmota (marmot or woodchuck), Spermophilus (ground squirrel), Cynomys (prairie dog), Thomomys (pocket gopher), Neotoma (wood rat), Peromyscus (deer mouse), and arvicolines (voles, lemmings, and muskrats). Species within each genus are united by generally similar ecological preferences. Ochotona lives in rocky alpine environments, which are in the vicinity of Porcupine Cave today and presumably were there throughout the Quaternary. Leporids and Peromyscus are ecological generalists, occupying a variety of habitats in grasslands, forests, and semiarid regions. Spermophilus and Cynomys are typically found in grasslands and semiarid shrublands. Marmota, although similar to ground squirrels and prairie dogs in aspects of life history (burrowing and long periods of hibernation or aestivation), generally occurs in more mesic areas. Most arvicoline rodents inhabit (p.321) relatively moister microhabitats than Spermophilus or Cyno-mys; an exception is Lemmiscus curtatus, characteristic of semiarid regions where sagebrush is abundant. Neotoma species typically build nests in caves when they are available, but spend most of their life foraging outside the cave as ecological generalists.
Many of these same genera dominate in numbers of species and numbers of individuals in the fauna around Porcupine Cave today (Fitzgerald et al., 1994). The fact that a similar assemblage of genera (or numbers of genera within a subfamily, in the case of arvicolines) has persisted largely unchanged in the Porcupine Cave region for more than 800,000 years suggests remarkable stability in montane communities at the generic level, even in the face of the major climate changes that have taken place during the Pleistocene. However, two of the most abundant taxa in the Porcupine Cave record, Mic-tomys (bog lemming) and Lemmiscus curtatus, do not occur in the region today, and many of the species are extinct, demonstrating that a modicum of fluidity operates within the basic community stability.
Relative Abundance of Small Mammals and Its Climatic Implications
The taxa that occur in the highest relative abundances, in order of abundance in most levels, are as follows: arvicolines, Neotoma, Spermophilus and Marmota, leporids, and Thomomys (figures 23.4, 23.5). The high relative abundance of Neotoma may reflect a taphonomic bias relative to the modern landscape, in that the fossil accumulation is basically a stacked series of wood rat nests in which the residents occasionally died. However, because this bias was constant throughout the deposit, it is still valid to compare relative abundances between levels. The high number of arvicolines in the fossil deposits parallels the situation on the landscape today at Porcupine Cave, as do the high numbers of Spermophilus and Marmota. This suggests that the same kinds of species that are most often sighted (or found in modern owl pellets and carnivore scat) on the landscape today were also most abundant on the landscape during the middle Pleistocene.
Despite the stability in the presence and rank-order abundance of genera, some clear changes in relative abundances of individuals within each genus occurred. Marmota first appears coincident with a slight decrease in Spermophilus and Cynomys at level 8 (figure 23.4). This probably represents an immigration of Marmota into the Porcupine Cave region, which took place during a climatic interval that the sediments suggest was an interglacial. Interestingly, this immigration coincides with a time that the arvicoline fauna suggests was among the wettest intervals represented in the entire record. It seems unlikely that Marmota’s first appearance in the Pit sequence is solely the result of sampling bias, given the reasonably large sample in level 10, the abundance of Marmota in levels 8 and above, and the coeval dip in Spermophilus and Cynomys. The temporary depression of those two xeric taxa would be consistent with expansion of the mesic microhabitats that favored Marmota and may even indicate competition for burrow habitat, given that all species share similar life history traits that focus on hibernating from fall to spring. Marmota continued to be abundant throughout this interglacial and the ensuing glacial, then decreased dramatically as Spermophilus and Cynomys increased with the onset of what was probably the warmest, driest interglacial.
Arvicolines demonstrate an inverse relationship with Sper-mophilus (figure 23.6). Hadly (1996) suggested that a similar inverse relationship at Lamar Cave reflected the shifting abundance of xeric (Spermophilus) and mesic (arvicolines) micro-habitats. This hypothesis has been confirmed by subsequent trapping and statistical analyses of bones found in raptor pellets from xeric versus mesic environments in and around Yellowstone National Park (Hadly, 1999; E. A. Hadly, pers. comm.). The peak of arvicolines in the uppermost glacial (at level 5) represented at Porcupine Cave thus suggests more mesic microhabitats relative to the upper part of the preceding or the entire subsequent interglacial. Interestingly, ar-vicolines peak during level 8, in the midst of an interglacial according to the sedimentary record. This peak coincides with the arrival of Marmota, substantiating a relatively wet time during the interglacial. However, it is noteworthy that one of the arvicolines that peak during this time is Lemmiscus curtatus, the sagebrush vole, today confined to the arid Great Basin and adjoining areas, where sagebrush abounds (Carroll and Genoways, 1980). The other arvicoline that contributes to the high abundance is Mictomys, the bog lemming, indicating an increase in very wet areas around the cave.
Arvicolines also show an inverse relationship with Neo-toma (figure 23.5). The Neotoma peak in level 9 may not be meaningful, given the small sample size. However, both of the Neotoma peaks occur immediately after the transition from a glacial into an interglacial. Today only N. cinerea occurs around the cave, and at both Neotoma peaks that species is most abundant. Also present are N. floridana and N. micropus in level 8A, and N. floridana and N. stephensi in level 3. The species other than N. cinerea are all found at lower elevations and warmer effective temperatures than characterize the Porcupine Cave region today (Repenning, chapter 18). All except N. stephensi occur within 80 km of the cave. Thus their increased numbers at the beginning of interglacials would be consistent with climatic warming. Repenning (chapter 18) speculated that tectonic changes might explain Neotoma distributions, but little geological evidence exists to support that suggestion.
The arvicoline rodent fauna of Porcupine Cave Pit contains 10 species and abundant specimens; thus it is diverse enough to meaningfully examine relative abundance changes at the species level (figure 23.7). All the arvicoline species except Lemmiscus curtatus and probably the Microtus specimens designated 5T are extinct. For reasons given by Wood and Barnosky (1994), Mictomys is probably an ecological analogue for the living bog lemming, Mictomys borealis. For most of the record, Lemmiscus curtatus and Mictomys dominate in abundance. Given the contrasting environmental preferences (p.322) of these two species—Lemmiscus indicating relatively xeric microhabitats and Mictomys indicating relatively mesic microhabitats—they provide a guide to environmental fluctuations within the glacial and interglacial episodes demarcated by the sediments. Below level 10 sample sizes are too small for meaningful discussion.
The basic pattern that appears above level 10 is as follows: high abundance of Mictomys in the lower interglacial (levels 9–6) coupled with fluctuating abundances of Lemmiscus; in the ensuing glacial (level 5–4), intermediate but declining values for Mictomys with the inverse for Lemmiscus; and finally, in the upper interglacial (levels 3–1), very low percentages of Mictomys with high, then low, then high abundances of Lemmiscus (figure 23.7). The general climatic inference from Mictomys is that mesic microhabitats were most abundant during the lower interglacial, slightly less widespread during the glacial, and sparse during the uppermost interglacial. The climatic inference from Lemmiscus is that sagebrush grasslands were at first abundant during the lower interglacial (level 9), then fluctuated until they were less abundant during the upper part of that interglacial (levels 8A-6). The addition of morphologically modern Microtus (with five triangles) in level 6 may suggest that the decline in sagebrush-dominated areas was accompanied by an increase in mesic grassy areas, but not as mesic as those preferred by Mictomys, because Mic-tomys declines beginning in level 6. During the glacial (levels 5 and 4), the relatively high percentage of Lemmiscus suggests increasing coverage by sagebrush grassland, which in turn implies a drier climate than prevailed during much of the preceding interglacial. In the uppermost interglacial (levels 3–1), the fluctuation between Lemmiscus, Microtus meadensis, and Microtus 5T (these are possibly extant species characterized by ml with 5 or 6 alternating triangles, but they cannot be identified to species) and the decline in Mictomys both suggest the driest times in the record, with levels 3 and 1 dominated by sagebrush grassland, and level 2 slightly more mesic, with grassland spreading at the expense of sagebrush.
Taken in concert, the arvicoline fluctuations imply that the lower interglacial (levels 6–9) was generally moist relative to the subsequent glacial (levels 4 and 5) and interglacial (levels 1–3). Temperatures were cooler in the lower interglacial than in the uppermost glacial, fluctuating from warm (level 9), to cool (level 8A), to warm again (levels 8–6). The level 4–5 glacial appears dry relative to the preceding interglacial, but wet relative to the subsequent interglacial. It was perhaps nearly as warm as the preceding interglacial but cooler than the uppermost interglacial (levels 1–3). The high abundance of Mictomys in level 10, combined with the relatively low abundance of Lemmiscus, suggests a cool, relatively moist glacial in the lowest level that produced reasonably abundant fossils.
Information from sciurids is in agreement (Goodwin, chapter 17). The perennially low abundance of Tamiasciurus (red squirrel or chickaree), a woodland indicator, drops after level 6, and it all but disappears by level 1.
These data agree with climatic interpretations derived from pollen, isotope, and invertebrate fossil data recovered about 200 km south of Porcupine Cave at the Hansen Bluff site at 2300 m, an elevation somewhat lower than that of Porcupine Cave but still reasonably high (Rogers et al., 1992). Assuming the correlation presented in figure 7.5, the uppermost interglacial sediments at Porcupine Cave were probably deposited during oxygen isotope stage 21; the middle glacial, during stage 22; the preceding interglacial, during a slight warming excursion that lasted from about 900 to 910 Ka ago; and the lower glacial, coincident with a cold spell between about 910 and 960 Ka ago. The Hansen Bluff climatic reconstruction suggests that stage 21 was marked by the warmest, driest conditions, but that fluctuations toward cooler, wetter conditions took place within it; this is very similar to the pattern indicated by the arvicolines in levels 1–3 at Porcupine Cave. Glacial stage 22 is interpreted as somewhat cooler and wetter than interglacial stage 21 at Hansen Bluff but was still characterized as warm and dry. The warm, dry overall character parallels the arvicoline interpretation for levels 5 and 6 in Porcupine Cave, but the high abundance of Mictomys and the sediments of those levels (Bell and Barnosky, 2000) are consistent with generally wetter conditions relative to the overlying levels. The preceding interglacial at Hansen Bluff was characterized as cold and wet; this is consistent with the arvicoline interpretation at Porcupine Cave of an interglacial that was relatively cool and moist compared to levels 1–3, with some intervals that may have been as cool as the subsequent glacial.
Striking similarities are evident between the arvicoline relative percentages at Porcupine Cave and a pollen diagram from Hansen Bluff. The sagebrush (Artemisia) pollen percentages through stage 21 at Hansen Bluff (Rogers et al., 1992:figure 10) start out high, drop to a low in the middle, and then increase again, just as the sagebrush vole (Lemmiscus) percentages do in levels 3–1 at Porcupine Cave. During stage 22 at Hansen Bluff, Artemisia percentages fluctuate with a peak in the middle, as do the Lemmiscus percentages at Porcupine Cave. Poaece (grass) and Pinus (pine) pollen percentages, which correlate with more mesic times, dip at the stage 22-stage 21 boundary and at the top of stage 21, consistent with the decline in Mictomys percentages at the corresponding parts of the Porcupine Cave record (levels 4–1).
Assembly of Species: Immigration, Extinction, and Evolution
The species composition was not static through the Porcupine Cave record. Immigration events include Microtus meadensis at level 8A, Marmota at level 8, Microtus paroperarius at level 7, and Microtus 5T at level 6. All these immigration events took place within the interglacial represented by levels 9–6, but they were staggered through the climatic interval, rather than clustered at its beginning or end.
In addition to these clear-cut first appearances, Brachylagus coloradoensis is first identified in level 7, and Sylvilagus and Lepus first appear in level 6. Prior to these levels all identifiable leporid specimens are either Hypolagus or Aztlanolagus. But there are many leporid specimens in lower levels that are not diagnostic at the generic level and could represent one of the later-appearing lagomorphs. Within the sciurids, Cynomys cf. C. leucurus first appears in level 5, along with Tamias cf. T. minimus. Tamiasciurus hudsonicus is first found in level 7 (Goodwin, chapter 17). The latter two species are difficult to substantiate as true immigrants, given their rarity, and their absence in lower levels could simply reflect small sample sizes.
Extinction, in contrast to immigration, does seem to cluster at the major climatic boundaries. Allophaiomys pliocaenicus, Mimomys virginianus, and Phenacomys gryci make their last appearance at the level 4/3 boundary. All three of these disappearances cluster at the transition from the middle glacial into the uppermost interglacial. Hypolagus appears last in level 10, as does the extinct morphotype of Spermophilus (?Otospermophilus) sp. (Goodwin, chapter 17). The transition from level 10 to level 9 also represents a glacial-interglacial transition. Extinctions at the glacial-interglacial boundaries are of interest in the context of the end-Pleistocene extinction debate, which pits human-caused extinction against environmentally caused extinction (Martin and Klein, 1984; Bar-nosky, 1989; Alroy, 2001). These middle Pleistocene examples from Porcupine Cave demonstrate that extinctions are a common feature at glacial-interglacial transitions, even in the absence of humans, though the only extinctions documented at Porcupine Cave are of small mammals.
Even with the immigrations, faunal turnover, and extinctions, the breakup of so-called disharmonious or non-analogue assemblages—that is, the sympatric occurrence of taxa that are presently allopatric—is not observed in the Porcupine Cave Pit sequence. This is in contrast to the situation at the latest Pleistocene-to-Holocene transition, where the breakup of non-analogue assemblages has been observed in mammals (Graham and Grimm, 1990). The only non-analogue association that characterizes the assemblage is the co-occurrence of Mictomys and Lemmiscus (Wood and Bar-nosky, 1994), and the association persists across all the major climatic boundaries.
Evolutionary patterns have been studied in marmots (Bar-nosky et al., chapter 25; Polly, 2003), other sciurids (Goodwin, chapter 17), and Lemmiscus curtatus (Barnosky and Bell, 2003). Marmots demonstrate no statistically significant morphological changes through the Porcupine Cave sequence. Goodwin (chapter 17) suggested that Cynomys cf. C. leucurus exhibited changes in the third lower molar, and Spermophilus cf. S. elegans in the fourth lower premolar, that potentially correlated with increased aridity. Barnosky and Bell (2003) observed that a shift in the frequency of Lemmiscus specimens with four triangles on their lower first molar versus those with five triangles became statistically significant only in the uppermost interglacial, suggesting a correlation with increasing aridity (figure 23.8). Thus it appears that climate change may have contributed to evolutionary change in some species, but only when it was most extreme—that is, during the increased aridity in the uppermost interglacial, which was apparently warmer than any of the climatic fluctuations that occurred lower in the sequence.
Species Richness Patterns
Analysis of species richness patterns was confined to taxa and stratigraphic levels that the rarefaction analysis (figure 23.2) and species-richness-per-level plots (figure 23.3) indicated were adequately sampled: rodents and lagomorphs from levels 1–6. These levels contain the following numbers of rodent and lagomorph species: level 1, 24; level 2, 25; level 3, 25; level 4, 29; level 5, 27; level 6, 23. A plot of these observed richness (p.325) values against the values expected from Coleman rarefaction analysis suggests that the interglacial levels 1–3 are less species rich than glacial levels 4 and 5, and interglacial level 6 (figure 23.9). The r2 value for the lower three levels is 0.932, indicating very close agreement for expected and observed values when only the glacial and immediately preceding levels are considered. The upper interglacial has observed values so far below the expected values that r2 drops to 0.067 when they are included in the regression. Moreover, observed richness drops more and more in the progression from level 3 to level 1, suggesting that richness declines successively as the interglacial progresses.
The species that disappeared at the glacial-interglacial boundary or within the interglacial included the arvicolines Allophaiomys pliocaenicus, Mimomys virginianus, and Phena-comysgryci, the pocket gopher Thomomys aff. T. bottae, and the wood rats Neotoma micropus and N. mexicana (see figures 7.1–7.3). Species diversity within the sciurids Spermophilus and Cynomys remained constant throughout all the environmental changes indicated in the Porcupine Cave record (figure 23.3), even though the species within those genera changed through time (Goodwin, chapter 17). Likewise, lagomorph diversity remained constant, even though the species and even the genera changed, from Hypolagus and Aztlanolagus in the lower levels to Lepus, Sylvilagus, and Brachylagus in the upper levels.
Rodents and lagomorphs from the Porcupine Cave Pit sequence are adequately sampled and provide the basis for assessing how faunal dynamics correlate with independently assessed glacial-interglacial cycles; they also fill in details of climatic conditions within each glacial and interglacial episode. On the generic and subfamilial levels, remarkable stability through glacial-interglacial cycles is apparent. The same genera (or, in the case of arvicolines, subfamilies) that are abundant on the landscape today were abundant in each of the glacial-interglacial cycles represented over 780,000 years ago. However, the population densities of species within those genera fluctuated in response to environmental changes, both within each glacial-interglacial episode and across glacial-interglacial transitions.
Ecological interpretations of those fluctuating relative abundances suggest that the uppermost interglacial in the Pit sequence was by far the warmest and driest climatic episode; the preceding glacial was dry and warm in general character but cooler and wetter than the uppermost interglacial; and the middle interglacial was relatively cool and moist compared to the uppermost interglacial. Within at least the glacial (levels 4 and 5) and uppermost interglacial (levels 1–3), climatic fluctuations altered the abundance of sagebrush grassland relative to more mesic grassland and boggy areas, resulting in fluctuations in the relative abundance of the species that prefer the respective microhabitats: Lemmiscus curtatus as a sagebrush-grassland inhabitant, Microtus spp. as grassland specialists, and Mictomys as a wet grassland occupant. These interpretations parallel those based on the sediments of the Pit sequence in Porcupine Cave, and those of Hansen Bluff, which covers the same time period as the Porcupine Cave Pit sequence but yields climatic interpretations from fossil pollen, invertebrates, and isotopic data.
Species composition was not static through the 200,000 years or more represented by the Porcupine Cave Pit sequence. Immigration of at least 4 and possibly 10 species occurred during the middle interglacial interval. These first appearances were staggered through the interval, rather than clustered at any one stratigraphic level, suggesting gradual assembly of the species rather than a single community moving en masse.
In contrast, extinctions clustered at glacial-interglacial transitions: at least two species at the lower transition, and three species at or near the upper glacial-interglacial transition. Added to three species that went extinct at the upper transition are at least three more that were extirpated either at the transition or within the uppermost interglacial.
This loss of species significantly decreased species richness in the uppermost interglacial relative to the preceding glacial. Taxa whose species seemed most affected by diversity fluctuations include Thomomys, Neotoma, and arvicoline rodents. Taxa whose species richness remained constant through the numerous climate changes indicated in the Pit sequence include Spermophilus and lagomorphs.
In the taxa that have been studied, evolutionary changes correlating with climate change either are not apparent or are confined to the most pronounced climate change, the increasing warmth and aridity in the uppermost interglacial interval.
In sum, the faunal dynamics indicate both stability and change in montane communities over the past million years (p.326) or so. The stability is evident in the general structure of the community at the generic level: most of the genera that are found as fossils in the Porcupine Cave Pit sequence still have representatives in the fauna today. The taxa that are abundant on the landscape today are also the ones that were most abundant some 800,000 years ago at the generic and sub-familial levels. And the rank-order abundance of genera and subfamilies (in the case of arvicolines) was not much affected even by major climatic transitions. Within this overall context of higher-level stability, however, are important changes indicating that the players change through time at the species level. Most evident in this regard is the long co-occurrence of two taxa, Lemmiscus and Mictomys, that today do not live around Porcupine Cave but that, during the time of the Pit deposits, were among the most abundant small mammals on the landscape. Further stamps of faunal change on the basic pattern of stability are various minor fluctuations in the relative abundance of taxa, probably in response to changes at the microhabitat level, immigrations, extinctions, and, in some taxa, probable evolutionary changes. Of particular importance in view of current global warming is the question of whether the basic stability of communities or the overprints of faunal change are more likely to prevail. That issue is the subject of chapter 26.
The research reported in this chapter was partially supported by NSF grant BSR-9196082. This chapter is University of California Museum of Paleontology contribution 1815.