At large scales in the Amazon, it is often hypothesized that species distributions are restricted by geographical barriers, such as huge rivers (river-barrier hypothesis). This hypothesis has been offered to describe the spatial-distribution boundaries of species and to indicate endemism areas for number of phylogenetic lineages. We tested the capacity of the river-barrier hypothesis to describe patterns of varieties diversity and spatial-distribution borders for 1952 easily-detected varieties in 14 taxonomic groups that occur approximately the Madeira River, and also our results show that the theory that the Madeira river is the border between endemism areas and also explains much of the diversity uncovered in the an ar is inappropriate for >99% the species. This shows that alternative hypotheses must be propose to describe the boundaries of distributions of varieties around the Madeira River, as well as a revision of the criteria that are provided to recognize species-endemism areas.

You are watching: How is the amazon river a barrier

Download PDF


Presence or lack of individuals of a species in the Amazon can be attributed to multiple factors. At regional scales, habitat characteristics have been established as the main components of the distribution of various of plants1,2,3,4, lizards5, anurans6,7, snakes8, ants9, mammals10,11,12, termites13 and birds14,15. However, at wider scales, the is frequently hypothesized that species distributions are greatly related come dispersal limitation caused by geography barriers, together as huge rivers16,17. This explanation is frequently referred to as the “river-barrier hypothesis”.

Wallace18 was one of the first to hypothesize the the distribution of Amazonian species could be limited by big Amazonian rivers, such as the Negro, Amazon and also Madeira Rivers. Follow to the contemporary interpretation the this hypothesis, huge rivers room expected come subdivide a population to the allude of staying clear of gene flow in between individuals in different areas and to promote hereditary divergence in between them, boosting the opportunity for allopatric speciation19,20. If this theory is correct, the is expected that (i) sister species or lineages will certainly be top top opposite flow banks21,22,23, (ii) the similarity in varieties composition will certainly be greater in localities on the same financial institution (adjacent sites) 보다 sites on opposite banks separated by the very same distance24,25,26,27 and (iii) the limits of types distributions will certainly coincide with large rivers21,22,23,24,25,26,27,28.

The river-barrier hypothesis has been offered to describe the spatial-distribution limits of species and to indicate feasible endemism areas29,30 for number of phylogenetic lineages in the several taxa in the Amazon (e.g. Primates23,24, lizards17,28, anurans16,17,25, butterflies21, birds22,26,27,31). The hypothesized endemism areas delimited by rivers have actually been supplied as surrogates in conservation planning30. However, this hypothesis is not constantly accepted and also the function of rivers as the boundaries to endemism areas has been questioned for numerous taxa13,17,26,27,31,32,33,34,35,36,37,38,39,40,41. For example, the impacts of the Tapajós river (for amphibians and squamates17) and the Amazon flow (for birds26) as barriers depend ~ above the life-history attributes of the species. Dambrós et al.13 showed that sites separated by large geographic ranges had unique termite-species composition and also most the the broad-scale variation in varieties composition can be defined either by spatial predictors or distinctions in ecological conditions between regions, and also not by huge rivers, such together the Madeira, Negro, Branco and also Amazon.

In the bulk of the studies that welcomed the river-barrier hypothesis, the conclusions were based on studies with few species20,23 and on the assumed lack of species on one bank16,17,24. In addition, rivers vary in discharge and width, and these two factors have been considered important in identify when huge rivers function as geographic obstacles to species dispersal24,36,42. Therefore, the acceptance or rejection of the hypothesis may depend mostly on the types and river investigated. This two factors together make it daunting to generalize the importance of big rivers as reliable geographical obstacles to the distribution of Amazonian varieties and as a feasible hypothesis to define the species diversity uncovered in the region.

In this study, we approximated the relationship of species in various taxonomic groups that have actually their distributions limited by a flow (the generic theory of the river together a barrier) and also the number of species because that which over there is evidence (sister varieties on opposite financial institutions of the river) the this river functioned as a vicariance barrier causing speciation (the hypothesis of visibility of endemism areas based on large rivers). We supplied only types for which false absences room unlikely to describe the figure of the river together a barrier. We performed the research on the Madeira River, which has actually been postulated together a barrier to dispersal for types of assorted taxa16,18,22,38,43,44,45,46,47 and the border between two endemism areas29,30, and also we learned an area in the mid will where plenty of studies have indicated the it is an effective biogeographic barrier. Ours results show that the hypothesis that the Madeira river is the border that separates two endemism locations (Inambari and also Rondonia) and that the river-barrier hypothesis describes much the the diversity found in the an ar is inappropriate for most species, and we imply that alternative hypotheses must be proposed to explain the limits of distributions of most varieties found in the region, and also a review of the criteria that are supplied to identify species-endemism areas.

Generic theory of big rivers as barriers

The hypothesis that the circulation of types around the Madeira flow is greatly related come dispersal limitation led to by flow barriers, to be rejected for most varieties studied (Fig. 1, Supplementary Table S1). Of the 1952 varieties with detection probabilities sufficiently high that false absences room improbable, only 0.10% (Primates: Saguinus labiatus labiatus and Aves:Lepidothrix coronata) had their distributions restricted by the river (Supplementary Figs S1 and S2).


Because the relationship of species limited to one side of the river counts on our decision regarding which varieties the detection probability was high enough for a precious test, our results could underestimate the number of species limited to one side of the flow if the varieties that are minimal by the river room those that are challenging to detect. Therefore, us report the variety of species in each taxonomic or functional group that had actually their distributions restricted by the flow considering various other P intended in Supplementary Table S1, and give your distributions in Supplementary Figs S1 and also S2 (only for species with detection probability ≥ 0.40). The variety of species supposedly separated by the flow was low in all cases, other than when us made for sure no correction for probable false absences (Supplementary Table S1).

Hypothesis that the presence of endemism areas based on big rivers

Evidence the the Madeira River functions as a vicariance obstacle causing speciation (presumption of the endemism-areas hypothesis) was not discovered for 713 (99.45%) that the types investigated because that which us could attain data to erect robust phylogenetic hypotheses (Supplementary Figs S3–S7). We uncovered evidence suggesting that the river had worked as a vicariance barrier only for 4 (0.55%) the the species .


Evidence saying that the Madeira River could have functioned as a vicariance barrier for Callicebus brunneus and also Callicebus dubius. (a) Phylogenetic theory of small, large and non-flying mammals (72 spp); (b) Vicariance hypothesis; and (c) types distributions follow me the Madeira River; black color squares represent recognized occurrence that C. Brunneus, and also gray squares represent known occurrence of C. Dubius; the black color solid line represents the Madeira River; the red solid line represents the Madre de Dios flow in Bolivia and the dashed heat represents the Amazon River. See Supplementary Fig. S7 for in-depth phylogenetic hypotheses connected with species distributions follow me Madeira river (right or left financial institution of the river). Map generated using QGIS v2.18 (


Evidence arguing that the Madeira River can have functioned as a vicariance barrier for Psophia viridis and also Psophia leucoptera. (a) Phylogenetic hypothesis of Aves (446 spp); (b) Vicariance hypothesis; and (c) varieties distributions along the Madeira River; black squares represent well-known occurrence that P. Viridis and gray squares represent known occurrence of P. Leucoptera; the black color solid heat represents the Madeira River; red solid heat represents the Madre de Dios flow in Bolivia; and also the dashed heat represents the Amazon River. Check out Supplementary Fig. S8 for comprehensive phylogenetic hypotheses connected with species distributions along the Madeira flow (right or left financial institution of the river). Map generated using QGIS v2.18 (

The theory of the Madeira River together the border of circulation was not supported for most species, for this reason our outcomes are not concordant through the river-barrier theory explaining the origin20,22 or spatial-distribution borders of species16,17, no one of the visibility of endemism areas29,30, for most of the species that occur approximately the Madeira River. Even if the hypothesis is correct that the efficiency of the river together a barrier depends on the features of life backgrounds of the species for a small proportion of part taxa17,26, this would explain only a very small component of the organic diversity of the Amazon40,48,49.

In many studies that embraced the hypotheses about the impacts of rivers16,17,24, the apparent absence of a species on the opposite bank to the sampled was offered to conclude the a big river was a geographical barrier. However, any type of species-sampling method has some bias and the absence of a varieties in a certain location could indicate that the species was just not detected50.

For example, Dias-Terceiro et al.16 uncovered that the distribution of Ameerega trivittata (Anura:Dendrobatidae) was restricted to the left financial institution of the Madeira river (accepting the generic theory of large rivers). This varieties was recorded on both financial institutions in the Madeira flow in our study and also on the left financial institution of the Tapajós flow in the research by Moraes et al.17. The Tapajós river is located surrounding to the right bank of the Madeira River, and also the existence of a types on the left financial institution of the Tapajós River implies the presence of A. Trivittata top top the right financial institution of the Madeira River. Fecchio et al.47 concluded that the composition of helminth in birds to be dependent on endemism locations in the Amazon, however some that the host species that supported this conclusion emerged in our samples live independence of the endemism area. It is feasible that the conclusion of these authors was biased through the false absence of the hold in among the locations of endemism. This feasible bias in conclusions has been observed because that other varieties in Cracraft29, a reference that has been widely supplied to support and also justify research studies that identify endemism areas in the Amazon, based only on the apparent lack of a varieties on the opposite financial institution of a huge river. It is possible that these are not the only cases of doubtful outcomes in the literature, since this type of potential error to be detected many times in ours analyses. In roughly 40% of the species, the detectability analysis indicated that sampling was poor to draw a conclusion. It might be that only hard-to-detect varieties are impacted by rivers, however this seems unlikely since the river-barrier hypotheses were raised based on easily-detected species.

It is unquestionable that large rivers space the distribution borders of part Amazonian species, yet the big number that exceptions shows that the clues of the Madeira River as a border between endemism areas may be inappropriate for most species. The is crucial to emphasize the rivers can role as varieties limits without necessarily indicating the they represent barriers that caused vicariance speciation51, an assumption of the presence of endemism areas based on large rivers. Alternatively, sympatric speciation via sex-related selection52,53, eco-friendly differences54,55,56 or eco-friendly interactions57,58; linked with dispersal limitation51,59 and also competition60 might produce the same patterns that allopatric distribution observed in Figs 2 and 3, and additionally in Ribas et al.22, Fernandes et al.20, Boubli et al.23 and also are likely an ext important mechanisms because that generating and also maintaining Amazonian biodiversity than rivers. However, these alternative hypotheses are frequently ignored in researches that expropriate the hypothesis of huge rivers together the reason of speciation. Moreover, many of the conclusions relating to the river-barrier hypotheses assume that the geographical distribution of a varieties does not change over time, but there is proof that plenty of distributions in the past were various from present distributions61,62,63,64,65.

The absence of evidence discovered to assistance the river-barrier hypotheses (generic hypothesis of big rivers as barriers, and the hypothesis of centers that endemism based on huge rivers) in a stretch the river commonly postulated together the border between endemism areas16,18,29,30,38,44,45,46,47, suggests that the theory of visibility of endemism locations based just on the distribution of a few species and very large rivers, should be reevaluated because that the majority of species. V the reevaluation of this limits, the need for brand-new hypotheses will certainly arise to define the Madeira River’s role in the origin and distribution of Amazonian biodiversity. An ext importantly, in the lack of information on the distribution of many species, the proposed endemism locations are being offered as surrogates in preservation planning30. Substitutes must only be supplied when over there is solid evidence of the relationship in between the majority of targets and the proposed substitute66. In the instance of centers that endemism, this evidence is not available for many Amazonian rivers, and specifically because that the Madeira River, the evidence that the is a border between endemism areas applies to a very small proportion that biodiversity.

Our results are for only one area and there are taxonomic issues relating to types boundaries that have to be resolved for numerous taxa. Many of the types we learned are recognized on morphological criteria and also with the applications of molecule methods more species might be uncovered that have actually the Madeira River as a border to their distributions. Nevertheless, our results show that the functions of large rivers in promoting organic diversity and also the usage of postulated endemism locations as convenient surrogates for conservation planning in the Amazon still must be tested for the particular taxonomic group and conservation concern being addressed.


Data source

To estimate the proportion of types whose distribution are efficiently delimited through the river, we took benefit of an intensive study that the fauna linked with the implantation that a hydroelectric dam ~ above the Madeira River. Sampling was brought out on both financial institutions of the river, adhering to the RAPELD protocol67 (Supplementary Fig. S8). Some varieties may be restricted by rivers however not occur on the immediate banks due come habitat-type (e.g. Flooded area) avoidance. However, the field infrastructure comprised two parallel 5-km trails (Supplementary Fig. S8) and likewise sampled non-flooded area. The variety of samples every bank and also taxonomic groups surveyed are noted in Supplementary Table S2. In this study, us investigated only the distribution of animal species, due to the fact that none of the evidence used to propose the river–barrier hypothesis was based on information about plants or microorganisms.

Data analysis

Generic hypothesis of large rivers together barriers

It was not possible to check the theory for all the varieties of the region, because small is known around the distribution of numerous species, and also many Amazonian types have no yet to be described. As surveys of every taxonomic or functional team were make by the exact same researchers, we could include non-described varieties (hereafter described as morphospecies), for those types for i beg your pardon detectability analyses suggested that the absence of documents on one bank of the river had small chance that being due to false absences.

In stimulate to attain an unbiased estimate of the relationship of species in every taxonomic or functional group that had their distributions limited by the river, we thought about that the flow was a potential geographical obstacle only when detectability analyses suggested that the expected probability (P supposed ) of the species truly being lacking from among the financial institutions (right or left) to be P expected  ≥ 0.50. This criterion allows us come conclude that the absence of a species on the opposite bank to which that was current is i can not qualify to be as result of false absences resulted in by failures in the detection the the species. This meant probability was estimated according to the formula:

$$ mP_ mexpected=1-<1-( mN/ mN_ msampleBank)/ mN_ msampleBank>^ mNsampleOppositeBank$$

where: P meant , is the meant probability that the varieties occurring on the bank opposite to the on i m sorry it to be recorded; N is the variety of samples whereby the species occurred; NsampleBank is the total variety of samples on the bank where the types was current (right or left bank); NsampleOppositeBank, is the total number of samples on the opposite financial institution to which the varieties was recorded.

Hypothesis of the presence of centers the endemism based on large rivers

The hypothesis of the existence of endemism locations based on huge rivers was tested for 717 types (no false absences taken into account) the vertebrates for which it was feasible to achieve phylogenetic information. To show if the river worked as a vicariance barrier independent the the taxonomic or functional group, we built a phylogenetic hypothesis individually for each group (Figs S3–S7). Because that small, big and non-flying mammals (72 spp), line (66 spp), lizards (35 spp) and frogs (98 spp), the phylogenetic relationship were derived with the R package “rotl”68, and for birds (446 spp) the info was derived through the website birdtree.org69,70,71.

To determine the number of sister types or lineages for which the flow was an noticeable vicariance barrier, we connected each types in the phylogenetic hypotheses (referring come the different taxonomic or sensible groups) through their location of event (right or left bank of the river). If sister species or lineages (indicated by the phylogenetic hypothesis) were present on opposite banks (allopatric distribution), this an outcome could be an indication the the river operated as a vicariance barrier.

Avoiding potential sample biases

Before agree the generic theory of huge rivers together barriers, and also the theory of visibility of endemism locations based on big rivers, and also to minimization the impact of sampling top top the results, we confirm the distribution of each varieties that apparently arisen only top top one bank based ~ above the data indigenous Santo Antônio with documents in the literature and also in the websites of the global Biodiversity details Facility (, speciesLink (, information system that integrates in genuine time, main data of scientific collections), Portal da Biodiversidade (, this site gives data and information on Brazilian biodiversity generated or obtained by the set of the Environment and related institutions) and also the Smithsonian national Museum that Natural history (

Data Availability

The datasets analyzed throughout the current study were accumulated during the environmental-impact researches for the Santo Antônio hydro-electric reservoir and are that open-access with the web site that the Instituto Brasileiro perform Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA) net site. However, because of some inconsistencies in the data base, the data used below were listed by Santo Antônio Energia and were further quality checked. Castle are easily accessible from the corresponding writer on request.


Cintra, R. & Naka, L. N. Spatial sports in bird community composition in relation to topographic gradient and forest heterogeneity in a main amazonian rainforest. Int. J. Ecol. 2012 (2012).


Ribas, C., Aleixo, A., Nogueira, A. C. R., Miyaki, C. Y. & Cracraft, J. A palaeobiogeographic design for biotic diversification within Amazonia end the past three million years. Proc. R. Soc. B. 279, 681–689 (2011).


Cracraſt, J. Historic biogeography and also patterns that differentiation in ~ the south American avifauna: locations of endemism. Ornithol. Monogr. 49–84 (1985).


Williams, B. K., Nichols, J. D. & Conroy, M. J. Evaluation and administration of animal Populations (Academic Press) 1–817 (Oxford, 2002).


Magnusson, W. E. Et al. Biodiversity and Integrated environmental Monitoring (Áttema Editorial) 1–351 (Santo André, 2013).

Download references

This study was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível superior - CAPES; Coordenação de Pesquisas em Biodiversidade – COBIO of the Instituto Nacional de Pesquisas da Amazônia - INPA; Programa de Pós Graduação em Ciências Biológicas (Biologia de Água Doce e Pesca Interior); Santo Antônio Energia – SAE; nationwide Institute for Amazonian Biodiversity (INCT-CENBAM) and also the regime for Biodiversity research study in west Amazonia (PPBio-AmOc). The data were accumulated by many biologists as part of the environmental-impact studies undertaken by SAE. We specifically thank Albertina Lima because that indications around the visibility of data for various groups.

Author notes

William E. Magnusson and Claudia P. Deus added equally to this work.


Programa de Pós graduação em Ciências Biológicas, Instituto Nacional de Pesquisas da Amazônia, Av. André Araújo, 2.936, Petrópolis, CEP 69.067–375, Manaus, Amazonas, Brazil

Sergio Santorelli Jr.

Centro de Estudos Integrados da Biodiversidade Amazônica, Av. André Araújo, 2.936, Petrópolis, CEP 69.067–375, Manaus, Amazonas, Brazil

William E. Magnusson

Coordenação de Pesquisas em Biodiversidade, Instituto Nacional de Pesquisas da Amazônia, Av. André Araújo, 2.936, Petrópolis, CEP 69.067–375, Manaus, Amazonas, Brazil

William E. Magnusson & Claudia P. Deus


S.S.J., C.P.D. And W.E.M. Design the study, S.S.J. Analyzed the data with contribution that W.E.M., S.S.J. Wrote the very first draft the the manuscript, and all authors contributed substantially come revisions.

See more: How Much Is 20000 Crowns In Us Dollars ? 20000 Crw To Usd

Corresponding author

Correspondence to Sergio Santorelli Jr..