Anthropogenic extinctions explain most size and trophic mismatches between large mammalian primary consumers from Amazonian and African rainforests

Present-day Amazonia has an impoverished fauna of large folivorous mammals by comparison with African rainforests, but recent fossil discoveries suggest that late-Quaternary anthropogenic extinctions could account for most size and trophic faunal mismatches. The extent to which modern Amazonian ecology may be an artifact of prehistoric human intervention merits periodic re-evaluation as paleontological evidence of the Pleistocene megafauna continues to emerge. Lowland tropical rainforests in the Americas, Africa, and Asia share many climatic, phytostructural, and phenological attributes. Among others, these biomes are characterized by persistently warm temperatures, high annual rainfall, evergreen vegetation, tall trees, woody climbers, gap-colonizing giant herbs, and prolonged annual fruit production (Jordano, 2000; Muller-Landau & Hardesty, 2005; Richards, 1952; Whitmore, 1998). Such similarities have led many zoologists to compare lowland rainforest vertebrate communities on different continents in the expectation that, despite differences in phylogenetic composition, they will be convergent in ecologically important respects. Examples of such community convergence include vertical stratification, taxonomically widespread frugivory, and extraordinary species richness (Basham et al., 2023; Fleming et al., 1987; Jenkins et al., 2013; Pillay et al., 2021; Thiel et al., 2021). Mammalian communities in Amazonian and African rainforests are similar in additional respects, notably with respect to trophic-guild occupancy (Dubost, 1987). Whereas similar niches in Amazonian and African rainforests are sometimes occupied by the same higher taxon—apex predators in both ecosystems are felids, for example, and diurnal-arboreal frugivores are anthropoid primates—other niches are occupied by morphologically and behaviorally convergent taxa (Charles-Dominique, 1983; Dubost, 1968). Additionally, body-size similarities among occupants of homologous trophic guilds in Amazonia and Africa have been interpreted as evidence for “basic [ecological] rules which govern the formation of mammalian communities in tropical forests regardless of taxonomic origins” (Dubost, 1987). Nevertheless, morphological and trophic comparisons among the largest primary consumers in Amazonian and African rainforests reveal striking differences. Ranked by body mass in each fauna, African rainforest primary consumers are consistently much larger than Amazonian primary consumers (Figure 1, Table 1). Additionally, the five largest primary consumers in African rainforests are all folivores—forest elephant (Loxodonta cyclotis), forest buffalo (Syncerus caffer), bongo (Tragelaphus eurycerus), gorilla (Gorilla gorilla), and giant forest hog (Hylochoerus meinertzhageni)—whereas the five largest primary consumers in Amazonia include only two folivores, the tapir (Tapirus terrestris) and the capybara (Hydrochoerus hydrochaeris). Therefore, by comparison with African rainforest, Amazonia has a conspicuously impoverished fauna of large folivorous mammals. Anthropogenic late Quaternary extinctions are known to have truncated the size distribution of mammalian faunas in North America, South America, and Australia: continents that had not been exposed to hominin predators prior to the arrival of modern Homo sapiens; by contrast, the mammalian faunas of Africa and tropical Asia, long exposed to predatory hominins—with which they presumably coevolved—are not known to have been depleted by anthropogenic extinctions (Sandom et al., 2014; Smith et al., 2018; but see Jukar et al., 2021). Although these disparate extinction histories are widely acknowledged, their significance for ecological comparisons of Amazonian and African rainforest faunas has been difficult to assess in the absence of information about the identities of extirpated species. Recent paleontological research, however, has discovered several candidate taxa. As reported by Asevedo et al. (2021), Lujanian (latest Pleistocene) fossils from southwestern Amazonia include the gomphotheriid proboscidean †Notiomastodon platensis (ca. 6300 kg), the ground sloth †Eremotherium laurillardi (ca. 3500 kg), the toxodontid notoungulates †Trigodonops lopesi (ca. 1900 kg) and †Toxodon platensis (ca. 1800 kg), the camelid †Palaeolama major (ca. 280 kg), and the pampatheriid armadillo †Holmesina rondoniensis (ca. 120 kg). Dental morphology suggests that these species were all folivores, and stable-isotopic (δ13C, δ18O) data suggest that they occupied C3-dominated habitats, presumably forests (Sage & Sultimanis, 2016). Adding these recently extinct species to the size-ranked graph of Amazonian primary consumers (Figure 2) eliminates or minimizes most body-size mismatches with similarly ranked species in the African rainforest fauna and suggests that end-Pleistocene extinctions were largely responsible for the deficit of large folivorous mammals seen in modern Amazonia. Despite reasonably close species-for-species matching in most cases, size differences between reconstructed-Amazonian and African rainforest taxa in the second-, third-, fourth-, and fifth-ranked pairs (P2–P5) are conspicuous. For P2 (†Eremotherium/buffalo), the size difference is 3230 kg; for P3 (†Trigodonops/bongo), it is 1700 kg; for P4 (†Toxodon/gorilla), it is 1685 kg; and for P5 (†Palaeolama/giant forest hog), it is 200 kg. These differences are too large to be explained by methodological issues associated with estimating body size of fossil species (Hopkins, 2018), so it seems likely that late-Quaternary Amazonian faunas may have included several species of folivores that were substantially larger than any but the elephant currently found in African rainforest. The abundance of large primary consumers in Africa and their scarcity in South America have been invoked as possible explanations for marked continental differences in hectare-scale rainforest tree diversity (Terborgh et al., 2016). Whereas only 50–75 tree species typically occur on one-hectare plots of African rainforest, 200 or more tree species are typically found on one-hectare plots in Amazonia. Terborgh et al. (2016) hypothesized that the low hectare-scale diversity of African rainforest trees, together with other “quite profound” continental differences in rainforest structure and diversity, could be attributable, in part, to the presence of megaherbivores in African rainforest and their absence in Amazonia. That megaherbivores (species with body masses >1000 kg; Owen-Smith, 1988) might have substantial effects on continental habitats is plausible because adult megaherbivores are effectively immune from predation: instead, their populations are food-limited and exert strong top-down pressure on sympatric vegetation. Cristoffer and Peres (2003) argued at some length that there was little evidence for megaherbivores in prehistoric Amazonia, but absence of evidence is not evidence of absence, and Asevedo et al.'s (2021) results—together with other fossil discoveries referenced in their report—provide the paleontological evidence that was previously lacking. If, in fact, late Pleistocene Amazonia was populated by several species of megaherbivores, then the vegetational differences between Amazonian and African rainforests noted by Terborgh et al. (2016) must either (1) have arisen in the Holocene, or (2) are ecologically unrelated to megafaunal presence/absence. Recent paleobotanical and ecological research (e.g., Currano et al., 2021; Miranda et al., 2022) has tended to support the second alternative. Continental comparisons aside, there can now be little doubt that Pleistocene Amazonia differed substantially from the modern biome in several megafaunal-related aspects, plausibly including nutrient cycling (Doughty et al., 2013; Pedersen et al., 2023), seed dispersal (Blake et al., 2009; Doughty et al., 2016), carbon storage (Berzaghi et al., 2019; Doughty et al., 2016), and canopy-gap size (Johnson, 2016). The extent to which the ecology of modern Amazonia is, in effect, an artifact of prehistoric human intervention merits periodic re-evaluation as paleontological evidence of the Pleistocene fauna continues to emerge. The data that support the findings of this study are openly available in the Dryad Digital Repository: https://doi.org/10.5061/dryad.q573n5tsg.