Historical biogeography of the neotropical Diaptomidae (Crustacea: Copepoda)
© Perbiche-Neves et al.; licensee BioMed Central Ltd. 2014
Received: 7 January 2014
Accepted: 14 April 2014
Published: 1 May 2014
Skip to main content
© Perbiche-Neves et al.; licensee BioMed Central Ltd. 2014
Received: 7 January 2014
Accepted: 14 April 2014
Published: 1 May 2014
Diaptomid copepods are prevalent throughout continental waters of the Neotropics, yet little is known about their biogeography. In this study we investigate the main biogeographical patterns among the neotropical freshwater diaptomid copepods using Parsimony Analysis of Endemicity (PAE) based on species records within ecoregions. In addition, we assess potential environmental correlates and limits for species richness.
PAE was efficient in identifying general areas of endemism. Moreover, only ecoregion area showed a significant correlation with diaptomid species richness, although climatic factors were shown to provide possible upper limits to the species richness in a given ecoregion.
The main patterns of endemism in neotropical freshwater diaptomid copepods are highly congruent with other freshwater taxa, suggesting a strong historical signal in determining the distribution of the family in the Neotropics.
Diaptomid copepods are among the main trophic links between primary producers and consumers in aquatic webs and represent the dominant family of Calanoida in inland waters of Europe, Asia, North America (NA), Africa, and the northern part of South America (SA) . Their origin, diversification, taxonomy, and evolution are still poorly understood in the Neotropics - particularly in South America - and there have been few studies dealing with general aspects of the regional biogeography of the group (but see [2, 3]). Although the origin of the order Calanoida is ancient, dating back at least from the Silurian period (439-416 MYA) , the initial colonization by diaptomids of continental waters from marine origins is hypothesized to have taken place in the northern supercontinent of Laurasia sometime after the break-up of Pangaea (around 160 MYA) . The success of this colonization and subsequent diversification process is reflected by the fact that nearly half of the current 440 known species in the family are distributed in the Palearctic and Nearctic regions . It has been assumed that the colonization of the Neotropics is probably much more recent, but no theory has yet received wide acceptance. According to the scenario hypothesized by Boxshall & Jaume , the presence of diaptomids at low altitudes in the northern and central parts of SA resulted from a late invasion from NA, occurring after the closure of the Panama gap in the Pliocene about 3 MYA. After invading from the North, the diaptomids would have spread rapidly, colonizing the interconnected lowland river systems. In this scenario the rapid colonization of multiple river basins would conceal biogeographical differences among regions, as well as a possible N-S gradient in species richness , given that the southern areas would have had less time to accumulate species. In fact, patterns that would be congruent with this scenario can be observed in freshwater cyclopoid copepods , but they are inconsistent with the remarkably high diversity of SA diaptomids. Alternatively, Suárez-Morales  and Suárez-Morales et al.  suggested the development of independent diaptomid faunas in Central/North America and in SA, resulting from their long isolation prior to the closing (in the Pliocene) of the Panama isthmus. Such a recent connection would help to explain the large differences in diversity between the SA diaptomid fauna and that of Central America (CA).
An understanding of diaptomid biogeography is still in its infancy. Most studies to date either include restricted geographical regions [3, 8] or are based on panbiogeography [9, 10], which has been heavily criticized for its excessive reliance on ancient vicariance (to the expense of alternative hypotheses) and its limitations in terms of reproducibility (see  for a recent discussion). Given the lack of comprehensive phylogenetic information on neotropical diaptomids, a valuable first approximation can be obtained through a Parsimony Analysis of Endemicity (PAE, ). One important recent criticism of PAE is use of biologically unrealistic limits to areas, such as geopolitical boundaries . Another, criticize the inability of PAE to detect perfect vicariance or dispersal to explain biogeographical histories trough phylogenetic biogeography . In contrast, one advantage of the study of freshwater fauna is that hydrological basins provide natural geographical boundaries between areas (e.g. ), thus improving considerably the explanatory power of the analyses. A detailed discussion for and against the use of PAE can be found at Morrone .
The goal of the present study is to provide a large-scale study of the biogeography of neotropical diaptomid species using the most comprehensive dataset on species occurrences compiled to date. Our specific goals are (1) to use PAE as a tool to provide a general overview of the biogeographical relationships among diaptomid copepod faunas of different neotropical ecoregions; (2) to map the geographical variation in diaptomid species richness in the Neotropics; (3) to test the role of climatic factors as potential correlates of diaptomid species diversity.
Multiple regression analysis of potential correlates of neotropical freshwater diaptomid species richness
Annual mean temperature
Maximum temperature of warmest month
Minimum temperature of coldest month
Precipitation of driest month
Precipitation of wettest month
The present study represents the most comprehensive continental-scale biogeographical investigation of freshwater copepods to date and the first to use ecoregion divisions as biogeographical units. Our results provide valuable insight into the distribution patterns of the organisms at different geographic scales, building on the earlier works of Brandorff , Dussart  and Suárez-Morales et al.  for the Neotropics. In particular, ecoregions provided congruent physical limits to explain the copepod distributional patterns through the PAE, as previously demonstrated for teleosts . The main ecoregions in terms of endemicity are Amazonas Lowland, Amazonas Estuary and Coastal Drainages, Lower Paraná, Chaco, Orinoco, Llanos, Paraguay, Iguassu, Madeira Brazilian Shield, Tocantins Araguaia, Northeastern Mata Atlantica, and Lower Uruguay. Some of these areas might represent diversity hotspots, and faunal differences between those regions should be taken into account in conservation initiatives.
The historical scenario elaborated by Boxshall & Jaume  assuming an extensive colonization and rapid speciation following the closure of the Panama gap in the Pliocene (about 3 MYA) would suggest that insufficient time had been available to allow biogeographical differentiation between regions, but our results do not support those predictions. In contrast, we found extensive differentiation between the diaptomid faunas of the ecoregions throughout the Neotropics. The high species richness of South American diaptomids is highly congruent with those observed for other freshwater taxa such as teleost fish (e.g. [20, 21]). The species richness of diaptomids was positively correlated with ecoregion area, and in this taxon the highest values were found in the largest hydrographic basins (the Amazonas and Paraná rivers), a pattern shared also with fish . The high diversity and endemicity of diaptomids in SA suggest an ancient occupation of this part of the neotropical region, prior to the Pliocene. The lowlands of Central and part of mainland Central America were subject to repeated marine transgressions, such that the current diaptomid fauna of the region was established during the Holocene, 8,000 years ago, by successive local colonization and extinction events . We infer that diaptomid copepods were distributed in SA over period of time long enough for the group to have undergone major diversification but also to remain isolated from Central America. The weak influence of the SA diaptomid fauna in Central America and Mexico, being represented only by the presence of Prionodiaptomus colombiensis[3, 22], together with the biogeographic affinity between the northernmost sector of South America and Central America (Figure 1) provides additional support for this inference. This interpretation is also supported by the species richness patterns of teleost fish and freshwater harpacticoid copepods. In both taxa the representation of SA species in CA is weak and attributed to recent, post-Pleistocene invasions [20, 23].
Central American Diaptomids are also known to occur in the Nearctic region, and may possibly be the origin of the main colonizing populations. Some species are also shared with the Palaearctic region [2, 3]. This suggests a weak influence of SA diaptomids, with exception of only two genera. The same occurs with the order Harpacticoida from high altitudes  as well as for other aquatic taxa, such as fish , a pattern attributed to desiccation processes in these regions . The opposite pattern has been revealed for cyclopoid copepods, with a strong SA influence in Central America , emphasizing that the appearance of these organisms is regarded as relatively recent , from the Neogene (15 MYA). Suárez-Morales & Reid  suggested at least five biogeographic groups for American diaptomids, which are supported by the present study using PAE. The actual number of diaptomid species might be underestimated in some genera that presumably radiated in Central America, such as Leptodiaptomus and Mastigodiaptomus. New taxonomically-oriented efforts and biogeographic data will reveal more detailed local patterns in this subregion. Nevertheless, it is likely that the general trend showing relatively lower diaptomid diversity with respect to South America will prevail.
In general, Central American ecoregions have smaller areas than the SA counterparts - a difference that could explain the lower species richness in CA and also in SA regions with restricted areas along the coastal basins of the Atlantic Ocean and Brazil. These ecoregions also have low teleost species richness . The higher species richness of diaptomids found in larger ecoregions is in accord with the results of previous analyses estimating the relationships between species number and area for planktonic crustaceans . Highly diverse areas might be characterized as biodiversity hotspots, as suggested for other groups of organisms , because of the increasing anthropic activities and potential disturbance of these sites. Such information should also be used to inform policies and conservation actions at different levels .
The Amazon ecoregion is impacted by an intense and increasing deforestation  and the construction of dams for hydroelectric energy along the main rivers (i.e. Madeira, Tapajós, Teles-Pires, Araguaia, Tocantins and Xingu) has had an additional impact. The reservoirs bring about a permanent change in surrounding habitats, such as lakes and marshes, and some studies suggest that copepod diversity is lower in reservoirs compared to lotic stretches of rivers . The Paraná River basin is extensively dammed and is also the location of the largest cities in SA, which contribute to raised water pollution and eutrophication. Only a few tolerant species thrive under such conditions and they are found in ecoregions that include the Paraná River basin. Some endemic species occur only in smaller ecoregions, and this highlights an important issue when considering species conservation: populations within smaller regions may be more susceptible to extinction brought about by various factors, including anthropogenic impact.
Little is known about the ecological drivers of species richness in freshwater copepods at larger scales, although climate has been recognized as an important variable impacting most components of regional fauna and flora . Variables such as lake area, conductivity, pH, dissolved inorganic carbon, and chlorophyll can all show correlations with copepod abundance (e.g. [25, 30]). In SA there are several polythermic species, which are likely to tolerate wide variation in temperature . In addition, the progressive increase of temperature driven by global warming has been implicated as responsible for changes in the centers of distribution of some crustacean species at higher latitudes . Our results suggest that a combination of climatic factors such as low precipitation and temperature, might determine an upper limit to the number of species that can be harbored within a given ecoregion, but more moderate climatic conditions within a particular geographical area do not necessarily lead to higher species richness. The highest levels of diaptomid diversity were linked with the largest ecoregions, supporting the inference that suitable conditions for the maintenance of high species richness were more likely to be stable in larger ecoregions. Based on the results of the BSS correlations undertaken in this study, we infer that minimum temperatures are a boundary condition for species dwelling in high temperature ecoregions, because of the high thermal variation in colder places. However, these conclusions should be interpreted with care, given that sampling effort has not been homogeneous throughout the Neotropics and therefore could have affected some of the observed patterns and/or masked potential correlates of species richness.
Despite the high environmental diversity and the complex biogeographic history of the Neotropics, diaptomid copepods show relatively well-defined patterns that are informative their underlying diversification processes. In particular, the relationships between areas of endemism and the high diversity in SA both support a history of diversification among copepods that preceded considerably the connection between the Nearctic and the Neotropics in the Pliocene. Finally, climatic conditions might provide boundaries to the level of diaptomid species richness that can be attained in a given ecoregion.
Records of freshwater diaptomid species occurrences were compiled from an extensive survey of the literature (Additional file 1: Appendix S1), as well as from new and recent data . The resulting dataset involves 98 species recorded in the Neotropics, although some are also found in the Nearctic region. This number of species differs from previous accounts (116 species of ; 82 from ). Each species was then scored as either present or absent in each of 72 ecoregions delimited according to the FEOW (Freshwater Ecoregions of the World, ) classification (, available at http://www.feow.org) (Additional file 1: Table S1). The resulting incidence matrix was investigated using Parsimony Analysis of Endemicity (PAE). Parsimony analyses were carried out using PAUP* v 4.0b10 . Heuristic methods consisted in 1000 random taxon addition searches followed by tree-bisection reconnection. The 100 best trees from each replicate were retained and the final tree was represented as a majority-rule consensus among all replicates. Data of occurrence of Diaptomids used in PAE are show in Additional file 1: Table S3.
Drivers of diaptomid species richness were investigated by multiple regression analysis of the number of species in each ecoregion against potential correlates, namely altitude, area, annual mean temperature, maximum temperature of warmest month, minimum temperature of coldest month, annual precipitation, precipitation of wettest month, and precipitation of driest month. Estimates of these variables were obtained from WorldClim v 1.4  by generating 5000 random coordinates in each ecoregion, extracting their respective values at a resolution of 2.5’ using ArcGIS v 9.3 , and calculating their averages. Finally, we tested whether diaptomid species richness is constrained by environmental conditions by using the Boundary Sum of Squares (BSS) test implemented in EcoSim v 7.0 , which is applied as follows. First, the ranges of the observed richness and environmental variables across all ecoregions are used to delimit a rectangular region in bivariate space. A lower right triangle is created by connecting the data points (min x, min y), (max x, max y), and (max x, min y), and the average distance of the points falling outside the lower right triangular from the boundary is calculated. If this value is unusually small, the points are concentrated near that boundary. Statistical significance is obtained by random resampling (1000 times) of the original dataset. The entire procedure was repeated for the following variables: annual mean temperature, maximum temperature during the warmest month, and minimum temperature during the coldest month. The effect of altitude was tested using a similar approach, except that a lower left triangle was used by connecting the data points (min x, min y), (max x, min y), and (max x, max y).
Gilmar Perbiche Neves is a post-doctoral researcher of University of São Paulo (USP, Brazil). He works on freshwater copepods, investigating several areas as biogeography, ecology, taxonomy, evolution, molecular and morphological phylogeny, etc., and his academic interests center on understanding the evolution of freshwater copepods.
We would like to thank to Professor Edinaldo Nelson dos Santos Silva (INPA, Brazil) for useful insight during this study. We also thank FAPESP (process 2008/02015-7, 2009/00014-6, 2011/18358-3) for financial support to GPN; and CNPq for financial support to DP (process 141702/2006-0) and MRP (process 304897/2012-4).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.