Distribution of the cladoceran Bosmina huaronensis Delachaux , 1918 and niche differentiation among populations from different biogeographic regions

Described from a high-altitude lake in Peru, Bosmina huaronensis Delachaux, 1918 has been recorded in diverse kinds of water bodies in South and North America, suggesting that this species has high environmental tolerance and a wide niche breadth. The present study surveyed the occurrence records of B. huaronensis from 55 localities and, using multivariate analysis, investigated the niche differentiation among populations from different biogeographic regions, based on altitude and seven climatic variables. The first two principal components (PC) explained 73% of the overall variance. PC1 was related to annual mean temperature, precipitation of driest quarter, and annual precipitation. PC2 was related to temperature seasonality and precipitation of wettest quarter. The PCA diagram showed three population groups, occupying different climate niches: (1) tropical highlands; (2) Neotropical lowlands; (3) temperate zones of both hemispheres. These results point to the need for further evaluation of these populations under morphological, genetic and ecological aspects.


INTRODUCTION
Bosmina huaronensis Delachaux, 1918 is a member of the zooplanktonic community from water bodies in a different trophic state (e.g.Menu-Marque and Marinoni, 1986;Zanata, 2005;Di Genaro, 2010).Described from a high-altitude lake in Peru (Lake Huaron, about 4,640 m a.s.l.), it has been recorded in diverse kinds of water bodies in South and North America (e.g.Paggi, 1979;De Melo and Hebert, 1994).In the South American tropical zone, Green (1995) only found B. huaronensis in lakes placed above 3,800 m a.sl.However, many records include South American ponds, lakes, reservoirs, and rivers in subtropical and temperate lowlands (e.g.Paggi, 1979;Velho et al., 2000;Serafim Junior et al., 2006;Rocha et al., 2011).
The niche is considered an abstract multidimensional space, in which a set of In the absence of a more complete data set, altitude and climatic variables, now easily available from global climate databases, have been used for niche differentiation studies for a variety of organisms and objectives (e.g.Knouft et al., 2006;Ibarra-Cerdeña et al., 2009;Batista and Gurgel-Gonçalves, 2009;Kamilar and Muldoon, 2010;Giannini et al., 2011;Schnitzler et al., 2012).
The present study aims to survey the occurrence records of B. huaronensis and investigate the niche differentiation among populations from different biogeographic regions, based on altitude and climatic variables.

MATERIAL AND METHODS
Occurrence records of B. huaronensis were obtained from literature and internet databases (Tab.1).Two original records from Brazil were also included: a pond at São Simão Beach, Mostardas, Rio Grande do Sul state, 14m a.s.l (col.LMAEL, Jan.01.2002) and Lake Paranoá, Brasília, Federal District, c. 1,000m a.s.l (col. LMAEL, Feb.19.2012).When literature did not provide geographical coordinates, they were obtained from Google Earth, and only verifiable locations were included in the analysis.
The niche differentiation among B. huaronensis populations was investigated by conducting a principal components analysis (PCA), which also was used to identify climatic variables potentially important in the geographic distribution.The abiotic variables (19 climate and one topographic) were obtained from the WorldClim database (http://www.worldclim.org/) at 5 arcminutes resolution.The "extract value by point" function in DIVA 7.5.0(Hijmans et al., 2001) was used to obtain the climate and altitude data for each locality.
Before conducting the PCA, a correlation matrix among all variables was constructed; this allowed the degree of multicollinearity in our dataset to be minimized by removing highly correlated variables (r > 0.8).Using this criterion, 12 variables were removed from initial dataset, and the PCA included the following eight variables: altitude, annual mean temperature, mean diurnal range, temperature seasonality, mean temperature of driest quarter, annual precipitation, precipitation of wettest quarter, and precipitation of driest quarter.The data were log10-transformed to standardize data for PCA.Statistics were performed using PAST (Hammer et al., 2001).
To better explore the possible aggregation of occurrence records, these were classified by biogeographic regions and transition zones (Tab.1), following Morrone's criteria (2006).

RESULTS
The data collected in the present study indicated that B. huaronensis is reported from 55 localities in South and North America (Tab.1), ranging about 45° S to 35° N.
In the multivariate analysis, the first and second principal components explained 44.9 and 28.1% of the overall variance, respectively.The climatic variables most associated with the distribution of occurrence records along axis 1 of the PCA diagram were annual mean temperature, precipitation of driest quarter, and annual precipitation (Tab.2).Altitude had the strongest negative relation with axis 1.For axis 2, the most related variables were temperature seasonality and precipitation of wettest quarter.
The ordination of occurrence records in the PCA diagram (Fig. 1) showed that populations from the biogeographic regions aggregate in three groups.The first group was represented by the records from South American and Mexican transition zones, including Lake Huaron (AS01), the type locality.The second one was formed by Neotropical populations.Nevertheless, there is an overlapping area between groups 1 and 2, represented by records AS24, AS27, AS28, AS37, AS40, AS41, and AS42 (Fig. 1; for exact localities, see Tab. 1).Finally, the third group aggregated the populations from Andean and Nearctic regions.One of the Neotropical records, from Lake Paranoá (AS08), was placed outside of the 90% density ellipses.

DISCUSSION
According to the surveyed records, B. huaronensis has a wide distribution range, including tropical and temperate zones of the American continent, from both hemispheres.In its range, this species occurs in a variety of water bodies, located at different altitudes and under diverse climate regimes.
The current distribution of species results from the interaction of large spatio-temporal processes (e.g.speciation, dispersal, climatic and geographic developments) and small-scale processes (e.g.competition, predation, local disturbances).The former processes act on a regional scale and govern species availability, but the local processes act as definite filters (Winkler and Kampichler, 2000).
The present study investigated B. huaronensis distribution only on regional approach, once only climatic and geographic aspects were considered.To better understand the processes acting on this species local distribution, it is also important to consider limnological variables and aspects of zooplankton community structure.However, this approach was outside the scope of this paper.
Despite this limitation, based on altitude and climate variables considered here, the occurrence localities of B. huaronensis were gathered into three groups (Figure 1).Group one represents the tropical high-altitude water bodies, located in the Central Andes and Central Mexico.Populations of B. huaronensis in these regions occur at low annual mean temperature, temperature seasonality, and precipitation, which is concentrated in the wettest quarter.As the type locality belongs to this group, this climatic niche should be considered the most typical for the species.
The second group, represented by Neotropical records, corresponds to localities in tropical and subtropical lowlands (below 1,000 m a.s.l.).The climate niche of this group is characterized by low temperature seasonality and high annual mean temperatures and precipitation in the wettest quarter.The overlapping area between groups 1 and 2 corresponds to localities on the eastern side of the Andes, with lower precipitation in the driest quarter.It could represent a transitional niche between the former two groups, but, alternatively, it could mean that the Morrone regions and transition zones are not the best criterion under which to investigate B. huaronensis distribution.
High temperature seasonality and low annual mean temperature and precipitation characterize the occurrence records of the third group, located in temperate zones of both hemispheres.
As presented here, B. huaronensis populations were found under different climate regimes, supporting the idea that the dominant clones in each one occupy a different ecological niche.Many authors have showed clones of cladoceran species differing with regard to their physiological, reproductive, behavioral, and ecological traits (e.g.Reede and Ringelberg, 1995;Dodson et al., 1997;Innes and Singleton, 2000;Michels and De Meester, 2004;Forasacco & Fontvieille, 2010;Lehto and Haag, 2010).Such differences, therefore, allow the occupation of different niches and enable the dominance of certain clones subjected to a particular climate regime.
Nevertheless, the possibility should not be excluded that these groups of populations could represent a complex of species.In recent years, new cladoceran species have been recognized as deep morphological and genetic comparisons between populations from different regions were conducted (e.g.De Melo and Hebert, 1994;Kappes and Sinsch, 2002;Nilssen et al., 2007;Elías-Gutiérrez and Valdez-Moreno, 2008;Belyaeva and Taylor, 2009;Kotov et al., 2009;Sinev and Elmoor-Loureiro, 2010;Van Damme et al., 2011).Therefore, B. huaronensis populations should be compared in their morphology, genetics and ecology in order to access their taxonomic status.
In particular, the population from Lake Paranoá should be better investigated.For about 30 years, the planktonic community of Lake Paranoá has been studied, and this species has never been reported before (Elmoor-Loureiro et al., 2004;Mendonça-Galvão, 2005;Batista, 2007;see Padovesi-Fonseca et al., 2001 for previous records).Additionally, it represents the first record from Central Brazil (Elmoor-Loureiro, 2000) and, in general morphology, the individuals do not differ from the species description (Paggi, 1979).However, at a similar latitude (11° to 17°S), B. huaronensis had previously been reported only above 3,000m a.s.l.(Delachaux, 1918;Uéno, 1967;Paggi, 1979;Valdivia-Villar, 1988;Coronel et al., 2007), while Lake Paranoá is located at about 1,000m a.s.l.Such geographic uniqueness corresponds to a particular climate (Köppen's Aw, tropical savanna climate, with well defined wet and dry seasons), which is reflected in the PCA diagram, where the Lake Paranoá record is located outside of the 90% density ellipses (Figure 1).Although the possibility that this result arises from gaps in sampling effort cannot be discarded, a more detailed morphological comparison between Lake Paranoá and the typical population is recommended.
The present paper showed that B. huaronensis has a wide distribution in the Americas and that its populations occupy three different climate niches: tropical at high altitudes, tropical and subtropical in lowlands, and temperate.Comparative morphological, genetic, and ecological studies are needed to evaluate the taxonomic status of these population groups.

Figure 1 .
Figure 1.Principal component analysis (PCA) of environmental variables and geographic distributions of Bosmina huaronensis, with 90% density ellipses.The first component explains 44.9% of the variation and the second component 28.1%.T.Z means transition zones.

Table 2 .
PC1 and PC2 loadings from principal components analysis of environmental variables for populations of Bosmina huaronensis.