Figures
Abstract
We used species distribution modeling to investigate the potential effects of climate change on 24 species of Neotropical anurans of the genus Melanophryniscus. These toads are small, have limited mobility, and a high percentage are endangered or present restricted geographical distributions. We looked at the changes in the size of suitable climatic regions and in the numbers of known occurrence sites within the distribution limits of all species. We used the MaxEnt algorithm to project current and future suitable climatic areas (a consensus of IPCC scenarios A2a and B2a for 2020 and 2080) for each species. 40% of the species may lose over 50% of their potential distribution area by 2080, whereas 28% of species may lose less than 10%. Four species had over 40% of the currently known occurrence sites outside the predicted 2080 areas. The effect of climate change (decrease in climatic suitable areas) did not differ according to the present distribution area, major habitat type or phylogenetic group of the studied species. We used the estimated decrease in specific suitable climatic range to set a conservation priority rank for Melanophryniscus species. Four species were set to high conservation priority: M. montevidensis, (100% of its original suitable range and all known occurrence points potentially lost by 2080), M. sp.2, M. cambaraensis, and M. tumifrons. Three species (M. spectabilis, M. stelzneri, and M. sp.3) were set between high to intermediate priority (more than 60% decrease in area predicted by 2080); nine species were ranked as intermediate priority, while eight species were ranked as low conservation priority. We suggest that monitoring and conservation actions should be focused primarily on those species and populations that are likely to lose the largest area of suitable climate and the largest number of known populations in the short-term.
Citation: Zank C, Becker FG, Abadie M, Baldo D, Maneyro R, Borges-Martins M (2014) Climate Change and the Distribution of Neotropical Red-Bellied Toads (Melanophryniscus, Anura, Amphibia): How to Prioritize Species and Populations? PLoS ONE 9(4): e94625. https://doi.org/10.1371/journal.pone.0094625
Editor: Bruno Hérault, Cirad, France
Received: September 14, 2012; Accepted: March 19, 2014; Published: April 22, 2014
Copyright: © 2014 Zank et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Caroline Zank was supported by a doctorate fellowship from CAPES. MBM is supported by CNPq (research grant 312968/2013-2). Diego Baldo is grateful to Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) for Doctoral Scholarship and ANPCyT for their financial support: PICTs 1524/2011, 1895/2011, 2687/2012, PIP 112201101/00875. The field data was supported by Fundação Grupo o Boticário and Centro Nacional de Pesquisa e Conservação de Répteis e Anfíbios of Instituto Chico Mendes de Conservação da Biodiversidade (RAN/ICMBio). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Projections of future climate predict that major changes will take place in most subtropical regions, including an increase in average global temperature and a decrease in precipitation [1], [2]. Current global climate distribution is likely to change during the 21st century, and possibly some climate patterns will disappear and others will emerge [3]. Impacts of global climate change can already be observed in several physical and biological systems [4], [5], [6], and these impacts might change the distribution of suitable areas for a wide variety of organisms by the end of the century, increasing the risk of extinction for many species [7], [8], [9], particularly those with restricted geographical range [10], [11]. In addition, species that are already threatened might suffer further negative changes in conservation status [12].
Population persistence under climate change is dependent either on adaptation or dispersal capabilities that enable species to track suitable habitat conditions in other areas [13], [14]. Species that do not display either of these abilities will probably become extinct [15], so that biodiversity may decline and highly mobile and opportunistic species may thrive [16], [17]. Amphibians are experiencing accelerating worldwide population declines and species extinctions [18]. Approximately 30% of all amphibian species are currently listed as threatened to some degree by the International Union for Conservation of Nature [19]. Amphibian decline is considered a global problem with complex local causes rooted in climate change, and in habitat alteration and fragmentation caused by human activities [8], [18], [20], [21]. Amphibians are particularly vulnerable to climate change given their aquatic/terrestrial life histories and low dispersal abilities. Therefore, displacement and contraction of suitable climatic areas represents a major threat to the conservation of amphibian species with low dispersal capabilities [22], [23]. The potential effects of climate change can be evaluated by developing models that provide working hypotheses to support research and conservation strategies. In the particular case of Neotropical amphibians, it is crucial to prioritize species and populations for in situ monitoring, with the aim of supporting conservation decisions, but also for evaluating whether the real effects correspond to those projected by models.
In this study, we used species distribution modeling (SDM, sensu [24]) to investigate the potential effects of climate change projected for 2020 and 2080, for 24 species of anurans of the Neotropical genus Melanophryniscus. These toads are small in size (less than 40 mm), and limited in mobility, and a high percentage of species are endangered and/or restricted in their geographical distribution [19]. To compile the species data, we gathered a comprehensive taxonomic and georreferenced list of 4,000 records from 22 scientific collections in six countries. Our main objective was to estimate the potential effects of climate change on the distribution of Melanophryniscus species and to use the results for prioritizing species for conservation. Based on comparisons of present time modeled distributions and distributions projected to the future, we aimed to answer the following questions: (1) What is the potential magnitude of change in the size of suitable climatic area for each species?; (2) What is the potential number of currently known populations of each species that would be located in non-suitable climatic conditions in the future (i.e., the potential loss of known populations)?; (3) Is the reduction in suitable areas related to Major Habitat Type, original range size or phylogeny (inferred from species groups); and (4) Can we rank species and populations for prioritization in research and conservation?
Material and Methods
Study species
Melanophryniscus Gallardo, 1961, is a Neotropical genus that was recovered as the sister taxon of all remaining Bufonidae in several phylogenetic analyses (e.g. [25], [26], [27]). The distribution of this genus is restricted to subtropical and tropical South America, including northern Argentina, central and southern Brazil, Uruguay, Paraguay, and Bolivia (Fig. 1). Melanophryniscus is currently represented by 26 recognized species [28], [29], [30], [31]. The majority of these species exhibit naturally small distributions, and in some cases, the distribution of a species is limited to only one or two known locations. It is also remarkable that their distributions are mostly non-sympatric (Fig. 1). In addition, these toads are considered rare and/or difficult to record. They are normally difficult to find during most of the year and are usually recorded only during explosive reproductive events (sensu [32]), which occur over a short period of time in temporary aquatic environments created during- and immediately after intense rainfall [33], [34], [35]. Precipitation and temperature appear to strongly influence the activity patterns of the species in this genus (e.g. [33], [35]). Melanophryniscus is a group under strong conservation concern, with at least 10 species included on endangered species lists at regional, national or global levels and three others listed as Data Deficient [19], [36], [37], [38], [39], [40]. This genus therefore includes a group of vulnerable species that might experience shifts of climatic suitability throughout their actual distribution range.
Points in the map represent the known distribution of the studied Melanophryniscus species throughout the major habitat types of South America.
We used data from 24 species (Dataset S1), including 21 recognized species in the genus Melanophryniscus [28], [30] and three undescribed species, hereafter referred to as M. sp.1, M. sp.2, and M. sp.3. Five recently described species, M. admirabilis Di-Bernardo, Maneyro, and Grillo, 2006; M. alipioi Langone, Segalla, Bornschein, and de Sá, 2008; M. vilavelhensis Steinbach-Padilha, 2008; M. peritus Caramaschi and Cruz, 2011; and M. setiba Peloso, Faivovich, Grant, Gasparini and Haddad, 2012, were excluded from the analyses because they are each known from one single locality.
Presence localities were obtained primarily from the published literature (Dataset S2) and later validated by a review of voucher specimens. We reviewed approximately 4,000 voucher specimens deposited in 22 scientific collections in Argentina, Brazil, Uruguay, Paraguay, Germany, and France (Dataset S3), including all historical records. To determine the geographical coordinates for each record of species presence, we used data available in the scientific literature or from voucher specimens in museum collections. When the precise geographic locations were not available, we assigned approximate coordinates according to descriptions of sample localities as they appear in museum records and in the literature, or to the closest town. We checked all locations using the on-line tools available in the SpeciesLink project (http://splink.cria.org.br/)[41].
Environmental data
For each species, we generated models of current suitable areas using altitude and nine bioclimatic variables obtained from the WorldClim version 1.4 database (Dataset S4), which is based on the interpolation of climatic conditions recorded from 1950 to 2000, with a 30 s (≈1 km) resolution [42]. We selected these ten variables because they have low collinearity [43], and also because they are ecologically meaningful to amphibians: mean diurnal range, isothermality, maximum temperature of warmest month, annual temperature range, mean temperature of wettest quarter, mean temperature of the warmest quarter, precipitation of the wettest month, precipitation seasonality, precipitation of the driest quarter and altitude.
To assess the effects of climate change, we used climatic data projected to the years 2020 and 2080, from the global climate model of the Canadian Centre for Climate Modeling and Analysis (CCCMA), which was recently evaluated as a top performing model [44]. In addition, we used data specified by emission scenarios (SRES) A2a and B2a described in the 2001 Intergovernmental Panel on Climate Change (IPCC) Third Assessment Report, which simulated climate system responses to increasing levels of greenhouse gases based on different hypotheses about projected population size, technological advances and socioeconomic trends. Scenario A2a projects relatively large changes based on recent observations suggesting that climate change will be more severe than previously expected [5], [45], [46]. Scenario B2a projects intermediate climate changes, based on projections of a world with intermediate population and economic growth [5]. Projected data for the maximum temperature, minimum temperature and average precipitation for the years 2020 and 2080 were obtained from the WorldClim database and processed in the software DIVA-GIS, version 5.2 [47], to generate the same nine bioclimatic variables used for modeling the current distributions.
Species distribution models
We generated species distribution models using the MaxEnt algorithm (MaxEnt version 3.3.3k), which searches for the maximum entropy density using Robust Bayes Estimation and requires only presence points as input data [48], [49], [50]. Ultimately, MaxEnt estimates the relation between species presence and environmental variables in a particular geographic space and draws a model of environmental suitability for the occurrence of a given organism. We estimated the current distribution of climatic suitability for Melanophryniscus species, projected the models on future climatic conditions and then compared the current and future distribution of climatic suitability.
We ran the MaxEnt software using the default settings, which have been validated in studies involving a variety of species and types of environmental data [50]. To obtain binary models of presence/absence from the continual logistic probabilities generated by MaxEnt, we selected a threshold at which the training sensitivity and specificity values were the same, minimizing the absolute difference between sensitivity and specificity [51]. This threshold approach yields low rates of both false positives and false negatives (both <0.2; [52]).
We restricted the geographical extent of the models and of the background sampling to a region between 0° and 56° degrees in latitude, and 34° and 81° degrees in longitude, which includes most of tropical and subtropical South America east of the Andes and covers the entire distribution of the genus in South America.
To evaluate model performance for each species, we used AUC values (Area Under the Receiver Operating Characteristics Curve) calculated through 10-fold cross-validation [49]. AUC values range from 0.5 for models with no predictive power to 1.0 for models with perfect predictive power [53]. AUC values greater than 0.9 denote “very good” predictive power, values between 0.8 and 0.9 denote “good” predictive power and values between 0.7 and 0.8 indicate “useful” predictive power [53]. Although AUC has known limitations as a measure of model performance [54], it still is the most used metric.
We produced maps of current and future suitable climatic areas for each of the 24 selected species of Melanophryniscus, and projected models of the future in different years (2020 and 2080) to take into consideration the short-, and long-term effects in two different scenarios of climate change.
Potential effects on species and populations and ranking criteria
To estimate the effects of climate change on different species and populations, we used two different approaches. We evaluated the relative changes in their potential distribution areas and the relative changes in the probabilities of occurrence at each known presence location.
To evaluate the changes in the suitable climatic areas for each of the 24 species, we overlaid the current, 2020 and 2080 potential presence maps to check for coincident presence regions. These coincident regions are indicative of persistent presence from the present time to 2080. This overlay operation is algebraically equivalent to P2080 = Ppresent*P2020*P2080; where P can be either 0 (absence) or 1 (presence). When P2080 = 1 for a given location, we then interpreted the predicted presence as persistent presence from the present time to 2080. This is a rigid estimation of potential persistence because it does not allow for recolonization of areas after a local extinction has taken place, even if climatic conditions become suitable again in the future. For example, if absence has been predicted for a given location in 2020, then Persistence2080 = 1*0*1, so that P2080 = 0 (i.e., 0 = predicted absence, or “local extinction” in 2080). The opposite result (and interpretation) would be obtained if two dates only had been considered (present date and 2080). In this case, Persistence2080 = 1*1 = 1 (i.e., 1 = predicted presence, or “long-term persistence from the present to 2080”). These analyses were based on the assumption that none of the studied species would be able to disperse to new potential areas within the modeling time frame, and therefore, any future increase in area was considered as not ecologically possible. We also assumed that no Melanophryniscus species would be able to adapt to new conditions. We made these assumptions due to the lack of information about the adaptive potential of Melanophryniscus species and because they seem to have low dispersal ability [35]. We acknowledge the debatable nature of these assumptions [24], [55], [56], however they can be considered plausible within the relatively short time frame (80 years) and spatial extent of our study, and considering that Melanophryniscus species are similar in size.
To quantify the effects of climate change at each known occurrence point, we used current and future (2020 and 2080) probabilities of occurrence generated by MaxEnt in a 5 km2 area centered on each point, using Idrisi Taiga GIS software [57]. This approach was used to attenuate the potential errors associated to using the values of a single point location (i.e., the unknown error in accuracy of the geographical coordinates of presence data taken from museum records). We assumed each presence location as a distinct population. For each species, we checked whether the probability of occurrence at each recorded presence point was above or below the presence threshold. Populations were then classified into two groups: (a) populations with occurrence probabilities below the presence threshold in 2020 and (b) populations with occurrence probabilities above the presence threshold in 2080.
We also evaluated whether differences in the magnitudes of area losses by 2080 were related to phylogeny (inferred from species groups), to the size of the original distribution range, and to Major Habitat Types (Dataset S5).
To prioritize the species according to the degree of impact caused by climate change, we used a scatterplot between the percent decreases in the distribution areas projected for each species by 2080 and the percent of occurrence points lost by 2080. Thus, species with both high percent loss in area by 2080 and high percent loss in known occurrence points should receive higher priority, while species with high percent decrease only in either area or known occurrence points would receive intermediate priority. Species with low decrease in both parameters would receive the lowest priority.
Results
The models of suitable climatic areas presented high AUC values ranging from 0.96 to 1.00 (Table 1). The presence threshold values varied across species from less lenient thresholds, such as that observed for Melanophryniscus cambaraensis (threshold = 0.725), to more lenient values, such as that of M. dorsalis (threshold = 0.026) (Table 1).
The bioclimatic variables that most frequently presented a high contribution to the climatic suitability models were: Isothermality (the diurnal temperature range divided by the seasonal temperature range) (BIO 3), the Coefficient of Variation of Seasonal Precipitation (BIO 15), or the Precipitation of Driest Quarter (BIO 17). At least one of these was between the two most important bioclimatic variables for almost 80% of species (Table 1).
The current projected suitable areas varied from 720,505 km2, (for Melanophryniscus klappenbachi), to 4,461 km2 (for M. macrogranulosus) (Table 2). The projected range of suitable areas estimated for most species (n = 11) were less than 100,000 km2. In two species, the estimated range of suitable areas were smaller than 10,000 km2, and for five species they were between 10,000 and 50,000 km2 (Table 2).
Potential magnitude of change in the size of species suitable areas
The estimated reduction in the geographic range of suitable conditions was widely variable at the species level and between the different scenarios (Table 2, Fig. S1–S6). Scenario A2a more frequently resulted in large decreases of suitable areas in 2080, with reductions of up to 70% for seven species. In Scenario B2a, a similar decrease in suitable climatic area in 2080 was predicted for only three species (Table 2, Fig. S1–S6), although nine species were predicted to have larger area reductions in comparison to Scenario A2a.
Considering the consensus scenario, 25% of species were predicted to have less than 10% reduction in their projected suitable areas, 40% were estimated to present more than 50% reduction. Particularly large losses of favorable habitat within projected suitable areas were projected for Melanophryniscus spectabilis and M. montevidensis, which were associated with estimated decreases of 90.1% and 100%, respectively (Table 2, Fig. S3 and S5). The projected range of suitable areas for 14 species was estimated to decrease up to 50% by the year 2080.
Potential loss of known populations
The MaxEnt models for the year 2080 indicated that the climatic conditions at a number of known presence sites – here assumed to be different population units or subunits – may no longer be suitable for the persistence of certain species (Table 2). For four species a pronounced reduction was predicted, with more than 40% of known presence sites below the presence threshold already in 2020 (Table 2, Fig. S1-S6). Nevertheless, for most species (60%), all of the currently known occurrence sites remained above the presence thresholds in 2080. In general, those species with larger estimated percent loss in area also presented the largest percent loss in currently known presence sites (r2 = 0.91).
Effects related to short-, and long-term climate change
We found that even in the short term models (2020) and regardless of the scenario considered, some species, like Melanophryniscus montevidensis and M. stelzneri, could already lose the totality of their projected suitable areas (Table 2, Fig. S1–S6). Furthermore, as previously mentioned, known populations of at least ten species (Fig. S1–S6) would fall in areas below the threshold of low climatic suitability. In contrast, three species (M. klappenbachi, M. paraguayensis, and M. cupreuscapularis) showed little or no reduction in their original projected suitable areas over the time period considered (Table 2, Figs. S1, S2, and S4).
Correlations with original distribution range, phylogenetic groups and Major Habitat types
Using a consensus of scenarios A2a and B2a, we found no significant correlation between the estimated area loss (%) by 2080 and the size of the original distribution range (Spearman rank order correlation, rS = −0.314, p = 0.135). There was considerable variation across species in the magnitude of the projected area losses, especially for species with current suitable areas of less than 100,000 km2 (Fig. 2). We also found no significant differences in the estimated percent area lost by 2080 between species occurring along the main Major Habitat types (“Tropical and Subtropical Grasslands, Savannas and Shrublands” vs “Tropical and Subtropical Moist Broadleaf Forests”; t-test, p = 0.192), or between the two larger groups of species in Melanophryniscus (stelzneri and tumifrons, t-test, p = 0.398) (Fig. 3).
Areas were estimated by MaxEnt models using bioclimatic variables; values are percentage reduction projected by 2080 (consensus between the IPCC A2a and B2a scenarios) for each species of Melanophryniscus. 1 - M. montevidensis, 2 - M. spectabilis, 3 - M. stelzneri, 4 - M. sp.2, 5 - M. cambaraensis, 6 - M. sp.3, 7 - M. tumifrons, 8 - M. macrogranulosus, 9 - M.atroluteus, 10 - M. sanmartini, 11 - M. rubriventris, 12 - M. sp.1, 13 - M. simplex, 14 - M. pachyrhynus, 15 - M. fulvoguttatus, 16 - M. moreirae, 17 - M. dorsalis, 18 - M. langonei, 19 - M. krauczuki, 20 - M. estebani, 21 - M. devincenzii, 22 - M. klappenbachi, 23 - M. paraguayensis and 24 - M. cupreuscapularis.
No significant differences were found in either situations (t-tests, TSGSS vs. TSMBF, p>0;19; stelzneri vs tumifrons, p>0.39). TSGSS, Tropical and Subtropical Grasslands, Savannas and Shrublands; TSMBF, Tropical and Subtropical Moist Broadleaf Forests.
Prioritization of species and populations for conservation
Based on the percent decreases in the distribution areas projected for 2080, and the percent of known occurrence points potentially lost by 2080, four Melanophryniscus species, and also their populations, can be set to high priority for conservation and monitoring (Fig. 4): M. montevidensis, (with 100% of its original suitable range and all known occurrence points potentially lost by 2080), followed by M. sp.2, M. cambaraensis and M. tumifrons. Three other species (M. spectabilis, M. stelzneri, and M. sp.3) can be set between high and intermediate priority because they are predicted to lose climatic suitability in more than 60% of their present (∼2000) range, although the predicted loss of currently known occurrence points is below 30%. Nine other species can be ranked as intermediate priority, with predicted losses in area or known occurrence points between ca. 30% and 60%, while eight species can be ranked as low conservation priority in terms of global climatic change.
Conservation priority is high for species with high loss values measured in both % area and % occurrence points. 1 - M. montevidensis, 2 - M. spectabilis, 3 - M. stelzneri, 4 - M. sp.2, 5 - M. cambaraensis, 6 - M. sp.3, 7 - M. tumifrons, 8 - M. macrogranulosus, 9 - M.atroluteus, 10 - M. sanmartini, 11 - M. rubriventris, 12 - M. sp.1, 13 - M. simplex, 14 - M. pachyrhynus, 15 - M. fulvoguttatus, 16 - M. moreirae, 17 - M. dorsalis, 18 - M. langonei, 19 - M. krauczuki, 20 - M. estebani, 21 - M. devincenzii, 22 - M. klappenbachi, 23 - M. paraguayensis and 24 - M. cupreuscapularis.
Discussion
Our results indicate that species of Melanophryniscus may be affected by climate changes in different ways. The projected suitable areas of some species were drastically different from the present, while for other species only minor changes were predicted. As a consequence, conservation, research and monitoring efforts should be prioritized specifically for those species which are expected to be more readily and largely affected by climate change (see Fig. 2 and Table 2). Among the species for which marked reductions in projected suitable areas are expected by 2080, M. montevidensis, M. sp.2, and M. cambaraensis have particularly small range sizes, which might consequently make these species more sensitive to climate change if there are any concomitant pressure from other environmental alterations.
The results for species with less than 10 occurrence points, as Melanophryniscus sp.2, should be treated with caution because predictive power decreases with low sample sizes [58], [59]. Nevertheless, the low number of known occurrences for several Melanophryniscus species is a consequence of their naturally small geographical distributions and not of undersampling their total distribution. Therefore, adding new occurrences would only increase the density of records within the already sampled small area, and would not improve the performance of the models.
The genus Melanophryniscus includes a high number of species endangered at regional, national or global levels [19], [36], [37], [38], [39], [40], which are threatened by several factors apart from global climate change [19]. The fact that these species are already under extinction risk and usually have small distribution areas makes them even more vulnerable to extinction, since relatively small geographic shifts in climatic suitability could affect almost the whole range of a given species. Examples from the present study are the already endangered M. montevidensis, M. cambaraensis, and M. macrogranulosus, for which an area reduction of more than 50% was projected, should be given special attention. On the other hand, the models for at least three endangered species (M. devincenzii, M. dorsalis, and M. langonei) predicted small losses in area of suitable climatic conditions by 2080. This means that having a small range area is not sufficient to predict that a species will be more vulnerable to climatic changes and that the effects of climate change can be very different even among congener species, demanding different degrees of prioritization. For Melanophryniscus, the magnitude of potential change was not related to the size of the present suitable areas, i.e., species with more restricted ranges would not necessarily be more affected by climate change than the ones that are more broadly distributed. This result contrasts with the findings of a previous study [60] in which the range of current projected suitable areas was found to be inversely related to the projected area loss, indicating a disproportionately greater vulnerability for species with restricted distributions. In fact, a weak relation between species range sizes and climatic change is not surprising, since the spatial distribution of climatic changes is not homogeneous across the geographic space. In some regions, the climate does not change as much as in others, even within biomes, or the Major Habitat Types used in our analyses. This intra-regional heterogeneity in climatic change is implicit in Figure 3, where the predicted area reduction within each Major Habitat Type (e.g., TSGSS) varied from no change (0%) to high changes (ca. 100%) in area. Therefore, species with small ranges will only be affected by climate change if their ranges are geographically coincident with the regions suffering major climatic changes, e.g., species 1 to 5 in Fig. 2 (see also Figs. S1 to S6). On the other hand, when there is no such coincidence, species with small ranges would not be affected by climatic changes (e.g., species 18,19, 23, 24 in Fig. 2).
By examining the projected changes in the climatic suitability for known populations of each species of Melanophryniscus, we identified two contrasting groups of populations (i.e., each known presence site was considered to be a population). The first group includes populations for which the area of suitable climatic conditions was not expected to decrease by the year 2080 (see Figs. S1 to S6). The second group includes populations projected to lose areas with suitable climatic conditions already by the year 2020. Monitoring this second group of populations to document trends in population sizes, reproductive events, and local habitat conditions (especially those that can be related to climate, distribution, amount and duration of local rains), would provide empirical data to test the accuracy of our projections. Data on these populations should also prove useful in supporting the development of more effective conservation plans for these species. Additionally, because these populations are expected to be the first Melanophryniscus populations to be affected by climate change, they are the most suited targets for in situ conservation efforts, such as breeding site enhancement and manipulation of the hydroperiod or water levels at breeding sites [61].
We consider in situ studies, particularly those involving long-term monitoring of known populations, to be a necessary approach for testing the accuracy of our model projections. In the absence of better ecological information, the bioclimatic variables that contribute most in generating the projected suitable areas models for each species may be used to select which local environmental conditions we should monitor in the future (Table 1). For instance, BIO 17 (precipitation of the driest quarter) and BIO 15 (coefficient of variation of seasonal precipitation) were important in the models of several species, suggesting that local precipitation patterns should be monitored. This is consistent with the life-history characteristics of Melanophryniscus, since their reproductive events are known to be related to intense rainfall [33], [34], [35]. Additionally, it would be important to investigate what are the thresholds of rainfall that trigger the reproductive events. Therefore, monitoring of local climatic conditions and their relation with Melanophryniscus population dynamics could be informative of thresholds and habitat requirements useful in conservation actions.
The usefulness of models of suitable climatic areas in helping to set conservation strategies for several taxa is undisputed (e.g. [12], [62], [63], [64], [65]). According to our results for Melanophryniscus toads (and probably also for other species with restricted ranges and limited dispersal ability), comparing modeled areas of present and future climatic suitability seems a useful approach for determining which species might suffer earlier reductions in habitat availability. Clearly, these species should be set as priorities for research and conservation, although we recognize that climate change analyses contain uncertainties [66], [67] and that climate change is not the only or even the main factor threatening all amphibian species (see [68], [69], [70], [71]).
In this study we predicted which Melanophryniscus species should be most vulnerable to climate change, based on the projected reduction in their suitable climatic areas. At the intra-specific level, analyzing the average percent reduction in the number of known presence sites (here assumed to be different populations), along with the percent reduction in potential ranges, provides an additional estimate of vulnerability for a given species. This analysis may be more useful to objectively pin-point priority conservation sites and to establish local level research, monitoring and conservation actions. In fact, by combining empirical data (analyses based on occurrence points) with projections generated by distribution models (analyses of projected suitable climatic areas), we here suggest one possible way of increasing the comprehensiveness, reliability and applicability of the assessments of climate change impacts.
Supporting Information
Figure S1.
Modeled distribution maps. Maps for M. atroluteus, M. cambaraensis, M. cupreuscapularis, and M. devincenzii. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s001
(TIF)
Figure S2.
Modeled distribution maps. Maps for M. dorsalis, M. estebani, M. fulvoguttatus, and M. klappenbachi. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s002
(TIF)
Figure S3.
Modeled distribution maps. Maps for M. krauczuki, M. langonei, M. macrogranulosus, and M. montevidensis. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s003
(TIF)
Figure S4.
Modeled distribution maps. Maps for M. moreirae, M. pachyrhynus, M. paraguayensis, and M. rubriventris. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s004
(TIF)
Figure S5.
Modeled distribution maps. Maps for M. sanmartini, M. simplex, M. spectabilis, and M. stelzneri. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s005
(TIF)
Figure S6.
Modeled distribution maps. Maps for M. tumifrons, M. sp.n.1, M. sp.n.2, and M. sp.n.3. The maps show the potential distribution areas in present time (∼2000), the areas potentially lost by 2020, potentially remaining by 2020 and potentially remaining by 2080. The 2080 consensus of remaining areas represents regions that persist as climatically suitable in either scenarios A2a (left) and B2a (right).
https://doi.org/10.1371/journal.pone.0094625.s006
(TIF)
Dataset S1.
Melanophryniscus species. Species of the genus Melanophryniscus included in this study.
https://doi.org/10.1371/journal.pone.0094625.s007
(DOC)
Dataset S2.
Literature used. The published literature used to obtain the presence locations for each study species: Caramaschi and Cruz, 2002; Baldo and Basso, 2004; Kwet et al., 2005; Brusquetti et al., 2007; Weber et al., 2007; Colombo et al., 2007; Langone et al., 2008; Maneyro and Kwet, 2008; Airaldi et al., 2009; Boeris et al., 2010; Bidau et al., 2011.
https://doi.org/10.1371/journal.pone.0094625.s008
(DOC)
Dataset S3.
Scientific collections. List of scientific collections containing Melanophryniscus samples reviewed for record and species validation.
https://doi.org/10.1371/journal.pone.0094625.s009
(DOC)
Dataset S4.
Environmental variable. List of variables used to model the potential distributions range of Melanophryniscus species. Source: project Worldclim versão 1.4 (http://www.worldclim.org).
https://doi.org/10.1371/journal.pone.0094625.s010
(DOC)
Dataset S5.
Melanophryniscus species and its major habitat type and phylogenetic groups. Melanophryniscus species: major habitat types in South America (those including more than 50% of each species distribution), and phylogenetic groups. Major Habitat Types (Olson, 2001): TSGSS = Tropical and Subtropical Grasslands, Savannas and Shrublands, TSMBF = Tropical and Subtropical Moist Broadleaf Forests, TGSS = Temperate Grasslands, Savannas and Shrublands. The taxonomic groups were defined based on Cruz and Caramaschi 2003; Baldo et al., 2012, and Baldo et al., unpubl. data.
https://doi.org/10.1371/journal.pone.0094625.s011
(DOC)
Acknowledgments
We are grateful to H. Hasenack and D. J. Alvares for helping with data processing issues in IDRISI-Taiga and DIVA programs use. We thank P. Colombo, M. Freire and T. Kunz for the exchange of information and fruitful discussions on the species of the genus Melanophryniscus. We also thank the curators of scientific collections, especially C.F.B. Haddad (CFBH); S. Kretzschmar, M. Cánepa, and E. Lavilla (FML); F. Brusquetti and F. Netto (IIBP); G. Carrizo and J. Faivovich (MACN); M. Di-Bernardo, G.M. Funk Pontes, and T. Grant (MCP); M.L.M. Alves (MCN/FZBRS); M. Motte (MHNP); J. Williams and S. Rosset (MLP); J. Langone and D. Arrieta (MNHN); U. Caramaschi, J. Pombal Jr., and C. Cruz (MNRJ), C. Mello, P. Narvaez, and H. Zaher (MZUSP); G. Winck and S. Cechin (ZUFSM); for their help during visits and for lending the specimens under their care. We are also thankful to the anonymous referees for their suggestions on earlier versions of the manuscript. Centro Nacional de Pesquisa e Conservação de Répteis e Anfíbios (RAN/ICMBio) for collecting permits (SISBIO 10341-1 and 23504-1). MBM is supported by CNPq (Research grant 312968/2013-2). DB acknowledge to Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) and ANPCyT for their financial support: PICTs 1524/2011, 1895/2011, 2687/2012, PIP 112201101/00875.
Author Contributions
Conceived and designed the experiments: CZ FGB MBM. Performed the experiments: CZ FGB MBM MA. Analyzed the data: CZ MA. Contributed reagents/materials/analysis tools: CZ FGB MBM MA DB RM. Wrote the paper: CZ FGB MBM DB RM.
References
- 1.
Christensen JH, Hewitson B, Busuioc A, Chen A, Gao X, et al.. (2007) Regional Climate Projections. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, et al.., editors. Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York. pp. 849–940.
- 2.
Allison I, Bindoff NL, Bindschadler RA, Cox PM, de Noblet N, et al.. (2009) The Copenhagen Diagnosis, 2009: Updating the World on the Latest Climate Science. Sydney: The University of New South Wales Climate Change Research Centre (CCRC). 60p.
- 3. Williams JW, Jackson ST, Kutzbach JE (2007) Projected distributions of novel and disappearing climates by 2100 AD. Proc Natl Acad Sci U S A 104: 5738–5742.
- 4. Menzel A, Sparks TH, Estrella N, Koch E, Aasa A, et al. (2006) European phenological response to climate change matches the warming pattern. Glob Chang Biol 12: 1969–1976.
- 5.
Meehl GA, Stocker TF, Collins WD, Friedlingstein P, Gaye AT, et al.. (2007) Global Climate Projections. In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, et al.. editors. Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press. pp. 747–845.
- 6. Scheffer M, Bascompte J, Brock WA, Brovkin V, Carpenter SR, et al. (2009) Early-warning signals for critical transitions. Nature 461: 53–59.
- 7. Parmesan C, Yohe G (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42.
- 8. Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, et al. (2004) Extinction risk from climate change. Nature 427: 145–148.
- 9. Attorre F, Francesconi F, Taleb N, Scholte P, Saed A, et al. (2007) Will dragonblood survive the next period of climate change? Current and future potential distribution of Dracaena cinnabari (Socotra, Yemen). Biol Conserv 138: 430–439.
- 10.
Penman TD, Pike DA, Webb JK, Shine R, Penman CTD (2009) Predicting the impact of climate change on Australia's most endangered snake, Hoplocephalus bungaroides. Divers Distrib 1–10.
- 11. Gibson L, McNeill A, Tores PD, Wayne A, Yates C (2010) Will future climate change threaten a range restricted endemic species, the quokka (Setonix brachyurus), in south west Australia? Biol Conserv 143: 2453–2461.
- 12. Marini MÂ, Barbet-Massin M, Martinez J, Prestes NP, Jiguet F (2010) Applying ecological niche modelling to plan conservation actions for the Red-spectacled Amazon (Amazona pretrei). Biol Conserv 143: 102–112.
- 13. Martínez-Meyer E, Peterson AT, Hargrove WW (2004) Ecological niches as stable distributional constraints on mammal species, with implications for Pleistocene extinctions and climate change projections for biodiversity. Glob Ecol Biogeogr 13: 305–314.
- 14. Tingley MW, Monahan WB, Beissinger SR, Moritz C (2009) Birds track their Grinnellian niche through a century of climate change. Proc Natl Acad Sci U S A 106: 19637–19643.
- 15.
Peterson AT, Tian H, Martínez-Meyer E, Soberón J, Sanchéz-Cordero V, et al.. (2005) Modeling distributional Shifts of Individual Species and Biomes. In: Lovejoy TE, Hannah L, editors. Climate Change and Biodiversity. New Halen & London: Yale University Press. pp. 211–228.
- 16. Malcolm JR, Markham A, Neilson RP, Garaci M (2002) Estimated migration rates under scenarios of global climate change. J Biogeogr 29: 835–849.
- 17. Loarie SR, Duffy PB, Hamilton H, Asner GP, Field CB, et al. (2009) The velocity of climate change. Nature 462: 1052–1055.
- 18. Stuart SN, Chanson JS, Cox NA, Young BE, Rodrigues ASL, et al. (2004) Status and trends of amphibian declines and extinctions worldwide. Science 306: 1783–1786.
- 19.
IUCN (2013) IUCN Red List of Threatened Species. Version 2013.1. Available: http://www.iucnredlist.org. Accessed on 20 August 2013.
- 20. Pounds JA, Bustamante MR, Coloma LA, Consuegra JA, Fogden MPL, et al. (2006) Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439: 161–167.
- 21. Griffiths RA, Sewell D, McCrea RS (2010) Dynamics of a declining amphibian metapopulation: Survival, dispersal and the impact of climate. Biol Conserv 143: 485–491.
- 22. Duarte H, Tejedo M, Katzenberger M, Marangoni F, Baldo D, et al. (2012) Can amphibians take the heat? Vulnerability to climate warming in subtropical and temperate larval amphibian communities. Glob Chang Biol 18: 412–421.
- 23. Foden WB, Butchart SHM, Stuart SN, Vié JC, Akçakaya HR, et al. (2013) Identifying the World's Most Climate Change Vulnerable Species: A Systematic Trait-Based Assessment of all Birds, Amphibians and Corals. PlosOne 8: e65427.
- 24. Elith J, Leathwick J (2009) Species distribution models: Ecological explanation and prediction across space and time. Annu Rev Ecol Evol Syst 40: 677–697.
- 25. Frost DR, Grant T, Faivovich J, Bain R, Haas A, et al. (2006) The amphibian tree of life. Bull Am Mus Nat Hist 297: 1–370.
- 26. Pramuk JB, Robertson T, Sites JW, Noonan BP (2008) Around the world in 10 million years: biogeography of the nearly cosmopolitan true toads (Anura: Bufonidae). Glob Ecol Biogeogr 17: 72–83.
- 27. Van Bocxlaer I, Loader SP, Roelants K, Biju SD, Menegon M, et al. (2010) Gradual adaptation toward a range-expansion phenotype initiated the global radiation of toads. Science 327: 679–682.
- 28.
Frost DR (2011) Amphibian Species of the World: an Online Reference. Version 5.5 (31 January, 2011). Available: http://research.amnh.org/vz/herpetology/amphibia/. Accessed 22 April 2012.
- 29. Caramaschi U, Cruz CAG (2011) A new, possibly threatened species of Melanophryniscus Gallardo, 1961 from the state of Minas Gerais, southeastern Brazil (Amphibia, Anura, Bufonidae). Bol Mus Nac 528: 1–9.
- 30. Baldo D, Borteiro C, Kolenc F, Rosset S, Prigioni C, et al. (2012) The taxonomic status of Melanophryniscus orejasmirandai Prigioni & Langone, 1987 "1986" (Anura: Bufonidae) Zootaxa. 3235: 45–61.
- 31. Peloso PLV, Faivovich J, Grant T, Gasparini JL, Haddad CFB (2012) An extraordinary new species of Melanophryniscus (Anura, Bufonidae) from southeastern Brazil. American Museum Novitates 3762: 1–32.
- 32. Wells KD (1977) The social behaviour of anuran amphibians. Anim Behav 25: 449–455.
- 33. Vaira M (2005) Annual variation of breeding patterns of the toad, Melanophryniscus rubriventris (Vellard, 1947). Amphib-reptil 26: 193–199.
- 34. Goldberg FJ, Quinzio S, Vaira M (2006) Oviposition site selection by the toad Melanophryniscus rubriventris in an unpredictable environment in Argentina. Can J Zool 84: 699–705.
- 35. Santos RR, Grant T (2010) Diel pattern of migration in a poisonous toad from Brazil and the evolution of chemical defenses in diurnal amphibians. Evol Ecol 25: 249–258.
- 36.
Lavilla EO, Ponssa ML, Baldo D, Basso N, Bosso A, et al.. (2000) Categorización de los Anfibios de Argentina. In: Lavilla EO, Richard E, Scrocchi GJ, editors. Tucumán: Asociación Herpetológica Argentina. pp. 11–34.
- 37.
Garcia PCA, Vinciprova G (2003) Anfíbios. In: Fontana CS, Bencke GA, Reis RE, editors. Livro Vermelho da Fauna Ameaçada no Rio Grande do Sul. Porto Alegre: EDIPUCRS. pp. 147–164.
- 38.
Haddad CFB (2008) Uma Análise da Lista Brasileira de Anfíbios Ameaçados de Extinção. In: Machado BM, Drummond GM, Paglia AP, editors. Livro Vermelho da Fauna Brasileira Ameaçada de Extinção Volume II. Brasília: MMA Biodiversidade 19. pp. 287–320.
- 39. Motte M, Nuñez K, Cacciali P, Brusquetti F, Scott N, et al. (2009) Categorización del estado de conservación de los anfibios y reptiles de Paraguay. Cuad Herpetol 23: 5–18.
- 40. Canavero A, Carreira S, Langone JA, Achaval F, Borteiro C, et al. (2010) Conservation status assessment of the amphibians and reptiles of Uruguay. Iheringia Ser Zool 100: 5–12.
- 41.
Specieslink (2012) Centro de Referência em Informação Ambiental. The project Species Link. Electronic Database accessible at http://splink.cria.org.br. Captured on March 2010.
- 42. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. Intern J Clim 25: 1965–1978.
- 43. Nori J, Urbina-Cardona JN, Loyola RD, Lescano JN, Leynaud GC (2011) Climate Change and American Bullfrog Invasion: What Could We Expect in South America? PlosOne 6: 1–8.
- 44. Radic V, Clarke GKC (2011) Evaluation of IPCC Models' Performance in Simulating Late-Twentieth-Century Climatologies and Weather Patterns over North America. J Clim 24: 5257–5274.
- 45. Rahmstorf S, Cazenave A, Church JA, Hansen JE, Keeling RF, et al. (2007) Recent Climate Observations Compared to Projections. Science 316: 709.
- 46. Smith JB, Schneider SH, Oppenheimer M, Yohe GW, Hare W, et al. (2009) Assessing dangerous climate change through an update of the Intergovernmental Panel on Climate Change (IPCC) "reasons for concern". Proc Natl Acad Sci U S A 106: 4133–4137.
- 47.
Hijmans RJ, Guarino L, Jarvis A, O'Brien R, Prem M, et al.. (2005) DIVA-GIS Version 5.2. A geographic information system for the analysis of species distribution data. Available at: http://www.diva-gis.org). Accessed 12 july 2011.
- 48. Elith J, Phillips SJ, Hastie T, Dudík M, Chee YE, et al. (2011) A statistical explanation of MaxEnt for ecologists. Divers Distrib 17: 43–57.
- 49. Phillips S, Anderson R, Schapire R (2006) Maximum entropy modeling of species geographic distributions. Ecol Modell 190: 231–259.
- 50. Phillips SJ, Dudík M (2008) Modeling of species distributions with MaxEnt: new extensions and a comprehensive evaluation. Ecography 31: 161–176.
- 51. Cantor SB, Sun CC, Tortolero-luna G, Richards-kortum R, Follen M (1999) A Comparison of C / B Ratios from Studies Using Receiver Operating Characteristic Curve Analysis. J Clin Epidemiol 52: 885–892.
- 52. Liu C, Berry PM, Dawson TP, Pearson RG (2005) Selecting thresholds of occurrence in the prediction of species distributions. Ecography 28: 385–393.
- 53. Swets K (1988) Measuring the accuracy of diagnostic systems. Science 240: 1285–1293.
- 54.
Franklin J (2009) Mapping Species Distributions, Spacial Inference and Prediction. New York: Cambridge University Press. 320p.
- 55. Araújo MB, Pearson RG (2005) Equilibrium of species' distributions with climate. Ecography 28: 693–695.
- 56. Peterson AT (2011) Ecological niche conservatism: a time-structured review of evidence. J Biogeogr 38: 817–827.
- 57.
Eastman JR (2009) IDRISI Taiga Tutorial. Worcester: Graduate School of Geography, Clark University. 333p.
- 58. Wisz MS, Hijmans RJ, Li J, Peterson AT, Graham CH, et al. (2008) Effects of sample size on the performance of species distribution models. Divers Distrib 14: 763–773.
- 59. Pearson RG, Raxworthy CJ, Nakamura M, Peterson AT (2007) Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. J Biogeogr 34: 102–117.
- 60. Broennimann O, Thuiller W, Hughes G, Midgley GF, Alkemade JMR, et al. (2006) Do geographic distribution, niche property and life form explain plants vulnerability to global change? Glob Chang Biol 12: 1079–1093.
- 61.
Shoo LP, Olson DH, McMenamin SK, Murray KA, Van Sluys M, et al.. (2011) Engineering a future for amphibians under climate change. J Appl Ecol: 1–6.
- 62. Garcia A (2006) Using ecological niche modelling to identify diversity hotspots for the herpetofauna of Pacific lowlands and adjacent interior valleys of Mexico. Biol Conserv 130: 25–46.
- 63. Ficetola GF, Thuiller W, Miaud C (2007) Prediction and validation of the potential global distribution of a problematic alien invasive species - the American bullfrog. Divers Distrib 13: 476–485.
- 64. Pineda E, Lobo JM (2009) Assessing the accuracy of species distribution models to predict amphibian species richness patterns. J Anim Ecol 78: 182–190.
- 65. Brito JC, Acosta AL, Álvares F, Cuzin F (2009) Biogeography and conservation of taxa from remote regions: An application of ecological-niche based models and GIS to North-African canids. Biol Conserv 142: 3020–3029.
- 66. Pearson RG, Dawson TP (2003) Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Glob Ecol Biogeogr 12: 361–371.
- 67. Hampe A (2004) Bioclimatic models: what they detect and what they hide. Glob Ecol Biogeogr 11: 469–471.
- 68. Young BE, Lips KR, Reaser JK, Ibañez R, Salas AW, et al. (2001) Population declines and priorities for amphibian conservation in Latin America. Conserv Biol 15: 1213–1223.
- 69.
Lovejoy TE (2005) Conservation with a Changing Climate. In: Lovejoy TE, Hannah L, editors. New Halen & London: Yale University Press. pp. 325–328.
- 70. Becker CG, Fonseca CR, Haddad CFB, Batista RF, Prado PI (2007) Habitat split and the global decline of amphibians. Science 318: 1775–1777.
- 71. Lips KR, Diffendorfer J, Mendelson JR, Sears MW (2008) Riding the wave: reconciling the roles of disease and climate change in amphibian declines. PLoS Biol 6: 441–454.