The detailed species-level inventories and associated phylogenetic and biogeographic data emerging from this project directly address hypotheses on the origins and maintenance of species-rich tropical aquatic faunas. Many important theoretical questions have previously remained intractable to empirical investigation – both because regional collections have been too scanty to permit detailed biogeographical comparisons, and due to the incomplete phylogenetic knowledge of most Amazonian aquatic taxa. These data provide a unique and powerful opportunity to illuminate aspects of the structure of Amazonian aquatic diversity at the meso to macro-scale (100-1,000 km). 

NEOTROPICAL SPECIES RICHNESS

     Regional species richness in aquatic ecosystems reaches a zenith in tropical freshwaters, which with less than 0.1% of the world's water harbors more than half its aquatic vertebrate species (Lundberg et al. 2000). This pattern is especially true for Neotropical fishes, which with about 5,700 known species represents a majority of the world’s c. 10,000 freshwater fishes, and perhaps 10% of all known vertebrate species (Vari and Malabarba 1998; Reis et al. 2003). The alpha diversity of Amazonian ichthyofaunas is especially high, with many floodplain faunas represented by more than 100 locally abundant resident species [12, 146, 158]. The aquatic faunas of Neotropical freshwaters, however, remain incompletely documented, especially at the species level. A recent review calculated that, as of 2003, ~ 25% of Neotropical fish species known in museum collections were undescribed [388]. Further, the current rate species discovery and publication suggests that the actual total of Neotropical fish species is substantially more than 7,000 (W. Eschmeyer, pers. comm.). Amazonian freshwaters therefore offer unparalleled -- but as yet poorly-developed -- opportunities for studying the origins and maintenance of species rich tropical animal assemblages.

SPECIES RICHNESS AND REGIONAL SPECIES POOLS

     The most celebrated and intensively studied cases of speciation focus on explosive episodes of diversification in geographically isolated areas (Harrison 1991; Schliewen et al. 1994; Coyne and Orr 1998; Johns and Avise 1998). Special insights have been derived from the study of species flocks that arise from such events including island faunas, lacustrine aquatic faunas, and recently, a riverine fauna (Sullivan et al. 2002). Despite the emphasis of explosive episodes of diversification in studies of speciation, much of life on earth exhibits evidence of a much broader geographical and temporal evolutionary history (Lundberg 1998). Accumulating evidence suggests that Neotropical fish species diversity is ancient, with regional species pools accumulating over tens of millions of years and over a geographical arena spanning multiple hydrogeographic basins. Similar patterns are also emerging for many elements of the Neotropical terrestrial biota, including amphibians (Gascon et al. 2000; Parra-Olea et al. 2004; Weigt et al. 2005; Roberts et al. 2006; Roberts et al. 2007), reptiles (Iturralde-Vinent and MacPhee 1999; Schulte et al. 2000; Doan and Arriaga 2002; Torres-Carvajal et al. 2006), birds (Bleiweiss 1998; Givnish et al. 2000; Pereira and Baker 2004; Fjeldsa and Rahbek 2006; Tavares et al. 2006; Weir 2006), and mammals (Salazar-Bravo et al. 2001; Galewski et al. 2005; Cozzuol 2006). In other words, the exceptional species richness of local Amazonian assemblages is not the result of local or even regional diversification. Rather, the exceptional species-richness of Neotropical faunas accumulated at a continental scale and over geological time frames. Patterns of biodiversity and biogeography in the species-rich Neotropical electric fish Gymnotus are consistent with this perception (Fig. 5). In this taxon sympatric species assemblages are of polyphyletic origin (i.e., are not the result of in situ radiations), comprise species with distributions that span far outside the area of sympatry, and comprise species which predate the Pleistocene climate oscillations. 

ALTO PURÚS ENDEMISM

     Preliminary investigations indicate the Alto Purús is an area of exceptional species-richness and endemism (INRENA 1999; Leite-Pitman et al. 2003). For many Neotropical taxa species richness and endemism reach their highest levels in the Southwestern Amazon Moist Forests ecoregion, including many plants (Ter Steege et al. 2000; Laurance et al. 2001; Pitman et al. 2001; Wittmann and Junk 2003; Burnham 2004; Kreft et al. 2004; Leimbeck et al. 2004; Parolin et al. 2004; Quijano-Abril et al. 2006; ter Steege et al. 2006; Wittmann et al. 2006) (but see (Leimbeck et al. 2004)), gastropods (DeJong et al. 2001), bivalves (Anderson 2006), oligochaetes (Lavelle and Lapied 2003), coleopterans (Erwin 1979; Pearson and Carroll 2001; Lucky et al. 2002; Scheffler 2005), dipterans (Hamada et al. 2002), lepidopterans (Brown and Freitas 2000; Racheli and Racheli 2003, 2004; Mallarino et al. 2005), fishes (Reis et al. 2003; Cox-Fernandes et al. 2004; Crampton et al. 2005; Albert et al. 2006; Hubert and Renno 2006), frogs (Eterovick 2003; Roberts et al. 2006), reptiles (Doan and Arriaga 2002), birds (Nores 2000; Brumfield and Braun 2001; Brumfield et al. 2001; Rahbek and Graves 2001) (Brumfield and Braun 2001; Brumfield et al. 2001; Brumfield et al. 2003a; Brumfield et al. 2003b; Brumfield et al. 2004; Racheli and Racheli 2004; Witt and Brumfield 2004; Brumfield 2005; Cheviron et al. 2005; Ribas et al. 2005), and mammals (Patterson et al. 1996; Peres 1997; Lambert et al. 2005, 2006). A recent rapid survey of the park recorded about 80 terrestrial mammal species, 157 reptiles and amphibians and over 100 fish species, a high proportion of which are unique to the region (Ortega 2003a). Among taxa for which species-level data are relatively well known for the region (i.e., birds (Bleiweiss 1998; Weigend et al. 2004) and butterflies (Mallarino et al. 2005)) the levels of endemism were found to be very high.

FITZCARRALDO PIEDMONT

     From a biogeographic perspective, the Alto Purús is part of the Fitzcarraldo Piedmont, a forearc basin associated with the Miocene-Pliocene rise of the Peruvian Andes (Campbell et al. 2001). The Fitzcarraldo Piedmont contains the headwaters of four of the largest tributary basins in the western Amazon—Ucayali, Juruá, Purús and Madeira (Fig. 2). Importantly, the upper reaches of these rivers are hydrologically isolated from one another, and the biotas of these waterways are separated from lowland Amazonia by rapids and cataracts. Analyses of radiometric and biostratigraphic data indicate that the Fitzcarraldo Piedmont changed from a depositional to an erosional setting during the Late Miocene to Pliocene (c. 9-3 Ma.) (Potter 1997; Campbell et al. 2001; Harris and Mix 2002; Campbell et al. 2006; Westaway 2006) providing minimum age estimates for the genetic divergence of populations/species inhabiting these isolated basins (Sivasundar et al. 2001; Hrbek et al. 2002; Montoya-Burgos 2003; Albert et al. 2006; Hrbek 2006; Lovejoy et al. 2006; Weir 2006). The Fitzcarraldo Piedmont therefore represents an exceptional biogeographic setting: it is the only geological region contained entirely within the Amazon Basin for which we have reliable estimates of the timing of headwater basin separation. Studies of aquatic taxa endemic to the Fitzcarraldo Piedmont provide unique data with which to test hypotheses on rates of diversification and the formation of regional species pools (Witman et al. 2004; McPeek 2007).

TIMING OF BASIN SEPARATION

     The unique geological setting of the Fitzcarraldo Piedmont provides opportunities for studying the timing of basin separation and the isolation of aquatic taxa in lowland aquatic Amazonian ecosystems. Geological and biostratigraphic evidence indicates that the basins of the Fitzcarraldo Piedmont were established in the Late Miocene-Pliocene (c. 9-3 Ma.) (Potter 1997; Campbell et al. 2001; Harris and Mix 2002; Rossetti et al. 2005; Campbell et al. 2006; Westaway 2006), constraining the minimum dates for estimates of the divergence times among populations and species restricted to these headwaters. Sedimentological analyses show a switch from depositional to errosional environments associated with the transition from mid-Miocene (Quechua phase) faulting to Pliocene (Diaguita phase) compressional deformation (Westaway 2006). Radiometric Ar-Ar dating of two volcanic tuffs from the Solimoes Formation were dated to c. 9 and c. 3 Ma. (Campbell et al. 2001). Mammalian biostratigraphy confirms these age estimates, as the top of the Solimoes Formation (Chapadmalan Stage) has no mammal fossils of North American origin (Cozzuol 2006). This indicates that sedimentation in the Solimoes Formation ceased before c. 4 Ma. Lastly, molecular dating of aquatic mammal and fish populations in the Upper Madeira (Cunha et al. 2005; Renno et al. 2006) suggests Late Miocene-Pliocene dates for the separation of the Upper Ucayali and Madeira basins.

BIOGEOGRAPHIC ANALYSIS

     Comparisons of taxa distributed across the headwater tributaries of the Fitzcarraldo Piedmont provide biogeographic and/or phylogeographic tests for the generality of models on the formation of regional species pools. We are compiling lists of candidate groups distributed among at least three of the constituent basins, and that exhibit measurable sequence divergence at the population and/or species-level. Candidate taxa with substantial genetic or phenotypic discontinuities are excluded from analysis to control for the implicit assumption (in the phylogenetic analysis of extant species) that extinction rates are relatively low compared with speciation rates (Albert et al. 1998; Albert et al. 2006). For some members of the target fauna (e.g., uninoid bibvalves, corbulid gastropods) the Miocene-Pliocene fossil record is sufficiently rich to permit direct tests of the influence of extinction on the formation of the regional species pools. Interspecific phylogenies and phylogeographic (intraspecific) data do not in isolation provide rigorous tests for alternative hypotheses concerning the geography of speciation, because of the lability of geographical ranges and the lack of correlation between the role of adaptive processes and geographical mode of speciation (Collins et al. 1996; Avise and Wollenberg 1997; Losos and Glor 2003). However, concordances in species-area relationships or phylogeographic patterns among multiple taxa do provide evidence pertaining to the sequence and relative timing of hydrological events that influence the diversification aquatic biotas (e.g., basin separation, headwater stream capture (Hrbek and Larson 1999; Lovejoy and Araújo 2000; Montoya-Burgos 2003; Hrbek et al. 2005; Albert et al. 2006; Hardman and Lundberg 2006; Hrbek 2006).