Australia: The Land Where Time Began

A biography of the Australian continent 

Latest Palaeocene Thermal Maximum – Mammalian Community response, An Isotaphonomic Study in the Northern Bighorn Basin, Wyoming

In this study new stratigraphic and palaeontological evidence from the McCullough Peaks, northern Bighorn Basin, Wyoming, was incorporated into an isotaphonomic faunal database that was used to investigate the impact of the Latest Palaeocene thermal maximum and coincident Wasatchian immigration event on the local mammalian community structure. It is shown by rarefaction that as a result of the immigration event there was an abrupt and dramatic increase in species richness and evenness. Diversity tapered off to more typical Wasatchian levels, though they still remained higher than those in the preceding Clarkforkian, following this initial increase. The new community organisation rapidly incorporated Wasatchian immigrants, which represented about 20% of the taxa and about 50% of the individuals. The immigrant taxa generally were of larger body sizes and had dietary habits that were more herbivorous and frugivorous compared to endemic taxa, which caused significant turnover in body size structure. In many lineages there was a significant decrease in body size that may have been prompted by the elevated temperatures and/or decreased latitudinal thermal gradients during the latest Palaeocene thermal maximum. Rapid short-term climate changes (transient climates) and associated biotic dispersal can have abrupt effects that are long lasting on mammalian community evolution.

For more than 100 years the Palaeocene-Eocene boundary has been recognised as an important period in mammalian evolution (Gervais, 1877). In North America and Europe Eocene mammalian faunas have differed from those of the Palaeocene in that artiodactyls, perissodactyls, and primates, new orders that dominate mammalian faunas from the beginning of the Eocene to the present. In spite of an intense effort at collecting, there has not been even a single occurrence of these important modern orders of mammals, which have been well documented in Palaeocene age deposits. The existence of a rapid short-term warming event near the Palaeocene-Eocene boundary that reduced significantly the latitudinal temperature gradients has been indicated by palaeoclimatological studies ([LPTM]; Zachos et al., 1993). It has been established that climate anomaly correlates precisely with the long documented turnover near the Palaeocene-Eocene boundary in North America (Clarkforkian-Wasatchian land mammal age boundary) by isotope studies of palaeosol carbonates, mammalian tooth enamel, and fish scales (Koch et al., 1992, 1995; Fricke et al., 1998). Earlier suggestions that increased equability opened high-latitude mammal migration corridors, which allowed members of the modern orders to disperse across the entire holoarctic, and presumably further, is strongly supported by these results (McKenna, 1983; Kraus & Maas, 1990). The geographic origins of these Eocene orders have, however, remained a mystery.

In this paper Clyde et al. combined new information from the McCulloch Peaks region with records from elsewhere in the Bighorn Basin that were similarly detailed in order to evaluate the effect of the LPTM and the coincident earliest Wasatchian immigration event on the diversity of the local mammal population, body size structure, and trophic structure. This study differs from previous studies on the Bighorn Basin mammalian turnover (Rose, 1981; Gingerich, 1989; Gunnell et al., 1983; Maas et al., 1995; Gunnell et al., 1995) in 2 important ways:

1.    The focus is on the ecological impact of the LPTM and the coincident immigration event, and

2.    The use of a database that includes all material (including elements that were labelled (“miscellaneous”) that were previously unidentified that had been collected from a single taphonomic mode (overbank assemblages that were surface collected.

This isotaphonomic approach (Behrensmeyer et al., 1992) is especially important in this type of study because they often represent a composite of several taphonomic settings that are not distinct uniformly through time.

First and last appearances were calculated for genera by the used of 10 m intervals in the new northern Bighorn Basin database of Clyde & Gingerich. If taxa were identified in the study area from older or younger deposits they were not counted as first or last appearances. As many of the species recorded belong to single lineages (Chronospecies), and therefore would not afford reliable first or last appearance information first or last appearances were recorded at the genera level.

There are 15 first appearances during Wa-0 (the earliest faunal zone of the Wasatchian) that characterise the Wasatchian immigration event and this is the largest first appearance event in the record. Only a slight increase across the Palaeocene-Eocene boundary interval, which suggests that the immigration event did not precipitate (or require) significant extinction or emigration. Species of immigration genera represent 18% of the total fauna at their initial appearance in Wa-0, a figure that doesn’t change much over the rest of the Wasatchian. Immigrant species represent 21% of all Wa-0 individuals, though in the next sampling interval they increase to 40%, where their abundance remains more or less constant for the remainder of the record. At present the establishment of dominance by 1 or a few introduced species over a relatively short period of time is a common pattern that is observed, though it has only rarely been resolve in fossil studies. It is suggested by the existence of an observable temporal lag between the immigration of the Wasatchian species and the establishment of their numerical dominance that the intervening period (Wa-0) was a transient interval of competition between Clarkforkian endemic holdovers and Wasatchian immigrants.

Diversity

Rarefaction compensates for the strong relationship between diversity and the size of the sample by standardising to the smallest sample that is being compared (Raup, 1975). Relatively low species richness for a given sample size is shown by Clarkforkian samples, whereas relatively high species richness for a given sample size is shown for Wasatchian samples. Anomalously high species richness values are shown by the Wa-0 sample. Earlier conclusions (Rose, 1981; Gingerich, 1989) that the Clarkforkian faunas generally have fewer species than the Wasatchian faunas that overlie them, are supported by these results, and differences in sampling are ruled out as the cause. Species evenness, which tracks the distribution of individuals within species, follows a pattern that is almost identical with Wasatchian faunas being distributed more evenly than Clarkforkian faunas, and with Wa-0 faunas being more evenly distributed than both. There are 36% of the individuals in the Wa-0 sample being represented by 3 taxa (Hyopsodus loomisi, Copecion davisi and Ecdocion parvus), whereas almost the same percentage of a typical Clarkforkian sample is represented by a single taxon (Ectocion osbornianus). It has previously been noted (Gingerich, 1989) that there was relatively high diversity in the Wa-0, though when sampling (compared to other sampling intervals) has been relatively poor it is even more extreme. It is also estimated that the Wa-0 interval is also estimated to be the shortest amount of time among the intervals that were included in this study. Diversity would be expected to be lowered by less temporal averaging, which would make the high Wa-0 diversity even more significant at this scale of sampling resolution. If Wa-0 represented a mixed sample of Clarkforkian and Wasatchian assemblages would be a simple explanation for this pattern. Given the differences in composition between Wa-0 and typical Clarkforkian and Wasatchian assemblages (e.g., 12 of the 39 species at Wa-0 are known only from this level, and there is only a single Clarkforkian species present in the Wa-0 fauna, however, faunal mixing cannot be an important cause of the abrupt rise in diversity at the beginning of the Wasatchian.

Body size structure

In order to investigate changes in body-size structure across the LPTM, estimates of body weight were calculated for taxa in each of the sampling intervals by use of body size regressions tooth size (Legendre, 1989). Where more than 1 M₁ was measured for a given species, the mean was used to estimate body weight. When there were no M₁ specimens that were available from the specimens used in the study, the measurements were taken from published sources. 10 species that lacked known M₁s were excluded from the study. The focus of the study was on relative changes in body-weight distributions, because the absolute accuracy of these distributions is highly skewed by the taphonomic filters that were mentioned earlier. All body-weight estimations were calculated were based on the same tooth (M₁), and relative body-weight should not be affected by errors introduced by regression.

In the fauna at Wa-0 there is a clear long-term increase in the mean species size, but a clear short-term decrease in the mean size of individuals. The disappearance of several small Clarkforkian species at Wa-0, and the immigration of new Wasatchian species with a large body size, is a reflection of the long-term pattern of mean species size. The temporary Wa-0 decrease in mean individual size, which is superimposed on the increase in mean species size, is more complicated.

The relative numerical importance of the different sized species in a sample is represented by the mean individual size. Many of the species (~50%) at Wa-0, immigrant as well as endemic, are unusually small when compared to other members of their genus that are present elsewhere in the record (Gingerich, 1989: Figs. 14, 36, 26, 28, 33 and 41). These relatively small taxa (for their genus) in Wa-0 are especially abundant and represent ~60% of the individuals. About half of the lineages that are recorded at Wa-0 are represented by species that are relatively small, with some being abundant, whereas the smallest Clarkforkian lineages (i.e, genera) are no longer recorded at Wa-0 and afterwards (in part causing the shift to higher mean species body size in the Wasatchian). There is substantial evidence that the LPTM had a direct short-term impact on the body size structure of local mammal communities, Given:

1.    That both immigrant (e.g., Hyracotherium) and endemic (e.g., Ectocion) taxa were affected by this short term size change,

2.    That it coincides precisely with a ~4^oC warming event in mean annual temperature (Frick et al., 1998), and

3.    That modern mammal individual (and population) body size is known to correlate negatively with temperature (the warmer the smaller) (Searcy, 1980), there is substantial circumstantial evidence that the LPTM had a direst short-term impact on the body size structure of local mammalian communities.

Whether this short-term body-size phenomenon resulted from in situ evolution or cline migration   is difficult to determine, but in either case smaller individual body sizes are consistent with warmer temperatures (Koch, 1996).

Trophic Structure

As with trophic structure, body-size structure is another taxon-free ecological parameter that can be evaluated for ancient mammalian communities. It is evidence of one of the sets of interactions that constitute the complex ecological web termed a community, and ultimately determines, in large part, the transfer of energy from one community to another. Visual inspection of tooth morphology, body size, postcranial elements, and masticatory muscle reconstruction, were the basis for dietary determinations for Clarkforkian and Wasatchian species. Only 5 trophic categories were considered in this study: herbivore, carnivore, omnivore, insectivore, and frugivore, given the uncertainty of this subjective coding system. Each trophic category is represented by a certain fraction of the total number of species, and a certain fraction of the total number of individuals, for each sampling interval.

A noticeable trend across the Clarkforkian-Wasatchian boundary is shown by a principal components analysis that was performed on the combined species and individual data. Relatively low scores were achieved by the Clarkforkian assemblages on the Principal Component 1, and relatively high scores by the Wasatchian assemblages on Principal Component 1, with Wa-0 falling out as intermediate. The high Wasatchian scores are associated with   abundances of species that are relatively high of herbivores, carnivores, and frugivores, as well as high individual abundances of herbivores and frugivores. On the other hand, the low Clarkforkian scores are associated with relatively high abundances of species and individual abundances of omnivores as well as insectivores. Influx of Wasatchian immigrants, which are characterised by a higher fraction of herbivores, carnivores and frugivores, and a lower fraction of omnivores and insectivores, compared to endemics (ꭓ² test p <0.001 for both species and individuals) is largely responsible for driving this trophic turnover. The trophic shift at Wa-0 is due, therefore, to the influx of immigrants, and not a direct in situ response of the community to the LPTM climate perturbation.

Discussion and conclusions

It is shown by these palaeoecological observations that the LPTM and the coincident earliest Wasatchian immigration event had significant short-term and long-term effects on diversity, body-size structure and trophic structure in Bighorn mammalian communities. In the high species level diversity and small size of individuals of the Wa-0 assemblage, short-term effects are apparent. Among the long-term effects, which continued through the entire Wasatchian record that was studied in this investigation, are higher diversity, larger average size of species, and more herbivory and frugivory in Wasatchian assemblages compared to the preceding Clarkforkian assemblages.  Most changes resulted from the immigration event of the latest Wasatchian, with the distinct ecological attributes being rapidly imposed on the existing Clarkforkian community. The short-term, within lineage, decrease in the size of the body that is exhibited by a number of immigrant and endemic lineages at Wa-0 is the single change that does not appear to relate directly to the immigration event. It is likely that this short-term phenomenon represents a direct effect (by in situ evolution or cline migration) of the LPTM.

The term transient climate to describe the short-term climate response to crossing a critical physical of chemical threshold during longer periods of climate change was coined by Zachos et al. (1993). Aberrant, unstable climatic conditions that return rapidly to background levels after geologically short durations are represented by transient climates. It seems that the latest Palaeocene thermal Maximum represents a classic example of this phenomenon. However, the biological effect of these transient climates can be profound and irreversible. The transient climate, in the case of the latest Palaeocene, was characterised by especially low latitudinal temperature gradients that permitted holoarctic dispersal of mammalian taxa. The ecological framework of local holoarctic communities was changed permanently by immigrant taxa. The climatic shift in the Bighorn Basin and coincident dispersal resulted in transient ecological conditions (Wa-0), which were followed rapidly by a new, stable, ecological framework that was quite different from that known before. The latest Palaeocene transient climate returned rapidly to normal, but its impact on biological communities continues to be felt today.

Clyde, W. and P. Gingerich (1998). "Mammalian community response to the latest Paleocene thermal maximum: An isotaphonomic study in the northern Bighorn Basin, Wyoming." Geology 26: 1011-1014.