Megafaunal Extinction in the Neotropics: A Landscape Approach

Abstract

A 20 million years reign of large terrestrial mammals, known as megafauna (animals over 45 kg), came to an end during the Pleistocene-Holocene transition. Particularly apparent in the Americas, the sudden extinction of Pleistocene megafauna has no equivalent in the fossil record as it only targeted large terrestrial animals. Early hypotheses explaining this extinction focused on climatic forcing and anthropogenic interactions, which included extensive hunting and/or landscape transformation. More recently, a third plausible explanation of megafaunal extinction has been the combined effects of climate-change and human influences; commonly referred to as the ‘synergistic effect of climate-humans’ hypothesis. The study of the ecology and extinction of Pleistocene megafaunal populations in the Neotropics has been hampered by the scarcity of well-dated fossil records. To date, there is almost no information regarding the timing of megafaunal extinction in Central America, and just one record from the tropical Andes. This lack of information is even more profound if we ask how they lived, rather than when they died. This study focuses on i) the long-term populations dynamics of Neotropical megafauna covering the last 45,000 years, ii) investigating the relationship between megafaunal abundance and vegetation changes, and iii) the timing and cause(s) of megafaunal extinction. The study follows megafaunal populations through paleo-proxies from regions with contrasting climatic and vegetation histories: the lowlands of the Yucatan Peninsula (Lake Petén-Itzá), and the highlands of the tropical Andes (Lakes Qoricocha and Junin). Megafaunal histories were determined by quantifying the occurrence through time of a suite of coprophilous fungus (e.g., Sporormiella, Podospora, and Cercophora), recovered from well-dated lake sediments. Changes in the climatic system were assessed with the analysis of fossil pollen, and diatoms, whereas the influence of humans in non-fire prone ecosystems was determined by charcoal analysis. The record from Lake Petén-Itzá spanned 45,000 years and provided a history of changing populations of coprophilous fungi from the lowlands of the Yucatan Peninsula. These data indicated that megafaunal dynamics responded to changes in the vegetation system driven by major change in climate associated with Atlantic Ocean forcing. In this tropical setting climates fluctuated between warm-wet interstadials and cold-dry stadials known as Heinrich events. Thus, the vegetation system alternated between open woodlands and scrubby grasslands. Pleistocene megafauna appeared to exist at maximum densities in the warmer, wetter woodlands of interstadials. These woodlands were rather open, with high abundances of grasses, and may be better thought of as woody-grasslands rather than forest environments. During cold, dry stadials, the vegetation transitioned to dry scrubby grasslands, that did not support strong megafaunal populations. Similarly, pine-dominated woodlands did not support megafauna. A further finding was that rapid climate change was detrimental to megafaunal populations. The control that climate change exerted on the dynamics of megafaunal populations was further observed in the tropical Andes. In this region, located more than 2000 km south of the Yucatan Peninsula, the climate was predominantly cold and wet during the Pleistocene, favoring the abundance of grassland or Polylepis (rosaceous treelet) scrubland vegetation above 3000 m elevation. In southern Peru, Heinrich events were perceived as wet events, although there was no marked temperature signal. The vegetation reconstruction from Lake Qoricocha indicated that Polylepis scrublands gave way to grasslands during the Pleistocene-Holocene transition, whereas the pollen analysis from Lake Junin suggested the prevalence of grasslands for the last 20,000 years. The Sporormiella analysis from the tropical Andean region showed that declines in megafaunal populations were linked to periods of warm-wet conditions (i.e.,Lake Qoricocha) with positive responses of megafaunal populations to relatively dry events during the deglacial period (i.e., Lake Junin). These contrasting responses of megafauna to climate might be explained by differences in the faunal composition. Some additional new insights into the ecology of these animals was revealed through the fossil pollen analysis. Sporormiella and fossil pollen from Lake Qoricocha, indicated a long co-existence of megafauna and Polylepis, a tree species highly sensitive to herbivore browsing. In the absence of physical barriers that could have prevented Polylepis browsing, I hypothesized that the survival of Polylepis might have resulted from solitary or small group behavior, as opposed to herd behavior, of megafaunal herbivores. Moreover, the low abundances of coprophilous fungal spores observed at Lake Junin and Lake Petén-Itzá were consistent with these megafauna exhibiting solitary behavior. The timing of extinction appears to have been similar in all the areas studied. In the Yucatan Peninsula a prolonged absence of coprophilous fungi after 13.2 k cal BP heralded the megafaunal functional extinction. This extinction coincided with increasingly dry conditions and aligned with the period of probable human population expansion in Central America. A hiatus in the 26,000 record from Lake Qoricocha obscured the precise point when megafaunal extinction occurred. However, projections of Sporormiella abundances suggest that megafaunal populations might have become functionally extinct around 12.5 k cal BP. In the central Andes of Peru, a 20,000-year record from Lake Junin, indicated that megafaunal extinction began at 13.0 k cal BP. In both cases the extinction coincided with local and regional signals of climatic instability, and occurred after unprecedented increases in fire activity suggesting human intervention in Andean ecosystems. The records from all the sites were consistent with the combined effects of climate change and humans as the cause of megafaunal extinction in these Neotropical settings. The surviving megafauna were a blend of forest (solitary) and grassland (herd-forming) taxa. It is unlikely that any of the forest megafauna were herd-forming, but in the grasslands both herd-forming and solitary lifestyles could have been adopted. Humans are effective predators in many settings, but probably most effective in open or grassland settings than forest. Perhaps solitary megafaunal behavior in open woodland/ grassland settings may have critically reduced survival chances when faced with human predators, accounting for the behavioral dichotomy of surviving species.