WordPress.com...ORIGINAL PAPER Divergent lineages and conserved niches: using ecological niche modeling to examine the evolutionary patterns of the Nile monitor (Varanus niloticus)

1 23

��

��

��

1 23

Your article is protected by copyright andall rights are held exclusively by SpringerInternational Publishing Switzerland. This e-offprint is for personal use only and shall notbe self-archived in electronic repositories. Ifyou wish to self-archive your article, pleaseuse the accepted manuscript version forposting on your own website. You mayfurther deposit the accepted manuscriptversion in any repository, provided it is onlymade publicly available 12 months afterofficial publication or later and providedacknowledgement is given to the originalsource of publication and a link is insertedto the published article on Springer'swebsite. The link must be accompanied bythe following text: "The final publication isavailable at link.springer.com”.

ORIGINAL PAPER

Divergent lineages and conserved niches: using ecologicalniche modeling to examine the evolutionary patternsof the Nile monitor (Varanus niloticus)

Stephanie A. Dowell1 • Evon R. Hekkala1

Received: 4 December 2015 / Accepted: 16 January 2016! Springer International Publishing Switzerland 2016

Abstract Ecological niche modeling is a useful tool that can support phylogeographicanalyses, offering insight into the evolutionary processes that have generated present-daypatterns of biodiversity. Findings of ecological divergence across evolutionary lineages canbe utilized to bolster inferences of parapatric or sympatric modes of speciation, and pro-vide support for species-level classifications. Conversely, conserved ecological nichesacross evolutionary timescales are thought to have facilitated allopatric speciation. Here,we examined the climatic niche of three genetic lineages of the Nile monitor (Varanusniloticus) to better understand the processes involved in generating patterns of geneticvariation, and to potentially clarify their taxonomic status. We built ecological nichemodels using genetically confirmed occurrence points from the three evolutionary lineagesof V. niloticus, occupying the western, northern, and southern regions of Africa. Pairwisecomparisons of climatic niche overlap provided evidence in support of niche conservatismacross all V. niloticus lineages. These findings are consistent with an allopatric mode ofdifferentiation. Furthermore, climatic niche conservatism could have played a role inisolating V. niloticus populations during historic climate oscillations, generating theobserved genetic patterns across Africa.

Keywords Niche conservatism ! Allopatric divergence ! Ecological niche modeling(ENM) ! African biogeography

Electronic supplementary material The online version of this article (doi:10.1007/s10682-016-9818-7)contains supplementary material, which is available to authorized users.

& Stephanie A. [email protected]; [email protected]

1 Department of Biological Sciences, Fordham University, 441 East Fordham Road, Bronx,NY 10458, USA

123

Evol EcolDOI 10.1007/s10682-016-9818-7

Author's personal copy

Introduction

The field of phylogeography aims to identify the spatial distribution of genealogical lin-eages, within and among species, and to clarify the underlying processes that created thesepatterns (Avise 2000). With genetic methods becoming increasingly available, identifyingpatterns of biodiversity is common; however, new questions emerge concerning the evo-lutionary processes responsible for such patterns (Alvarado-Serrano and Knowles 2014).Many phylogeographic studies propose that present-day patterns of variation reflect his-toric climatic oscillations and the subsequent shifts in habitat availability (Flagstad et al.2001; Hekkala et al. 2011; Lorenzen et al. 2012; Moodley and Bruford 2007). As suit-able habitat was reduced to isolated patches of refugia, populations likely diverged inallopatry (Haffer 1969; Lorenzen et al. 2012). Conversely, in heterogeneous environments,ecological differences have been shown to influence genetic structure and divergence(Evans et al. 2009; Keyghobadi et al. 1999; Peterson and Holt 2003; Schliewen et al.2001), even in marine systems (Mendez et al. 2010). Incorporating ecological niche models(ENM) into phylogeographic studies can provide insight into the potential evolutionaryprocesses that have generated present-day patterns of genetic variation.

Examinations of ecological niche conservatism and divergence have shed light onevolutionary mechanisms, as both are thought to play crucial roles in the speciation process(Wiens and Graham 2005). Niche conservatism, the tendency of species to retain aspects ofthe fundamental niche over evolutionary time, could be an important aspect of allopatricspeciation by limiting adaptation to conditions at a geographic barrier (Wiens 2004b;Wiens and Graham 2005). For example, isolated populations may arise when environ-mental change in a portion of a species’ range occurs more rapidly than adaptation to thenew conditions, resulting in a fragmented distribution (Wiens 2004a). Evidence from sisterspecies comparisons has revealed largely equivalent ecological niches across disjunctdistributions, suggesting that niche conservatism has facilitated divergence (Kozak andWiens 2006; Wiens 2004a). Alternatively, adaptation to different environmental condi-tions, niche divergence, is thought to be an important factor in sympatric and parapatricspeciation (Coyne and Orr 2004). Populations occupying non-overlapping niche spacewould be unlikely to exchange migrants and reproduce (Graham et al. 2004; Wiens andGraham 2005). However, geographically separated populations that have diverged viaallopatry can potentially occupy distinct ecological niches by chance due to differinghabitat conditions in their separate distributions (Warren et al. 2010). Therefore, whenmaking inferences of niche conservatism or divergence among disjunct populations, it isimportant to consider the underlying environmental conditions within their geographicalranges (Warren et al. 2010).

Inferences of niche conservatism or divergence can additionally aid in prioritizingconservation strategies. Identifying patterns of biodiversity, and understanding the pro-cesses involved in generating such patterns, are essential aspects of effective conservationpractices (Dubois 2003). However, our ability to successfully recognize and preserveevolutionary units is often hindered by poorly resolved taxonomic classifications (Dubois2003; May 1990). Therefore, biodiversity studies are increasingly taking a multidisci-plinary approach to delimit species boundaries, utilizing ecological information in additionto molecular and morphological characters (Dayrat 2005; Leache et al. 2009; Rissler andApodaca 2007; Welton et al. 2014). Inferences of niche divergence, as assessed withmodeling tools, have frequently been invoked in species delimitation, offering support forspecies-level status (Leache et al. 2009; Raxworthy et al. 2007; Rissler and Apodaca 2007).

Evol Ecol

123


Furthermore, understanding how the climatic niche of a species has changed over evolu-tionary time can be useful when predicting how species will respond to global climatechange (Wiens et al. 2010). For species with highly constrained climatic tolerances, thosewith strong climatic niche conservatism will be at high risk of extinction as climaticconditions shift (Wiens et al. 2010). Conversely, sister species or species groups exhibitingniche divergence may be more evolutionarily labile, and therefore, have a greater chanceof adapting to the changing environmental conditions (Holt 1990; Wiens et al. 2010).

The Nile monitor (Varanus niloticus) represents an ideal case to test for niche con-servatism or divergence. This lizard species has a widespread distribution throughout sub-Saharan Africa and the Nile River drainage basin (Lenz 2004). Exhibiting semi-aquaticbehavior, V. niloticus is predominantly restricted to permanent bodies of water throughoutits range (Lenz 2004); however, it can be found in a wide variety of climatic and vegetationzones, from tropical forests to semi-arid savannahs (Bayless 1997). This species reacheslarge sizes (2 m in length) and is highly mobile, occupying home ranges up to 50 km2

(Lenz 2004). As a large, colorfully pigmented reptile species, V. niloticus is under intenseexploitation pressure for the leather industry, in addition to threats posed by bushmeat andthe pet trade (de Buffrenil 1995; de Buffrenil and Castanet 2000; De Lisle 1996; Jenkinsand Broad 1994; Pernetta 2009). This species is managed under Appendix II of theConvention on International Trade in Endangered Species (CITES; http://www.cites.org);however, additional information on population structure is necessary for effectivemanagement.

A recent phylogeographic analysis by Dowell et al. (2016) identified three distinctgenetic lineages within V. niloticus, occupying the western, northern, and southern portions

0 1,500 3,000750 Km

100

Western

Northern

Southern

Major Rivers

Lineages:

0.03

Western

Northern

Southern

Outgroup

90100

Fig. 1 (Left) Molecular-based phylogenetic tree based on the data and results of Dowell et al. (2016)showing the relationships among the evolutionary lineages of Varanus niloticus. The maximum likelihoodtree was constructed using the concatenated mitochondrial (12s, ND1, ND2, and ND4) and nuclear (RAG-1,KIAA1217, KIAA1549) DNA sequence data, and values at major notes indicate bootstrap support (%). Thescale bar denotes 0.03 substitutions/site. (Right) Map showing the occurrence points of individualsbelonging to each of the three major V. niloticus lineages

Evol Ecol

123


of its range across Africa (Fig. 1). The western lineage was found to exhibit levels ofsequence divergence above those typically observed between sister varanid species, withno evidence of substantive admixture (Dowell et al. 2015, 2016). Additionally, thenorthern and southern groups displayed a moderate level of genetic divergence, though alarge degree of admixture was detected between them (Dowell et al. 2016). Historicspecies descriptions have noted morphological differences between V. niloticus popula-tions, proposing separate species or sub-species classifications (Daudin 1802; Schlegel1844). However, V. niloticus is currently recognized as a single species throughout itsrange.

In this study, we test for climatic niche conservatism versus divergence among the threeV. niloticus evolutionary lineages to (1) gain a deeper understanding of the processesresponsible for their present-day phylogeographic patterns, and (2) attempt to clarify theirtaxonomic status. By projecting the niche models onto historic and future climate sce-narios, we also examine how the distribution of V. niloticus has changed over time andassess how their populations may be affected by climate change.

Materials and methods

Occurrence records

We utilized the dataset of Dowell et al. (2016) to obtain occurrence records for specimensgenetically assigned to the three major lineages. Specific locality information was availablefor 72 of these genetically confirmed V. niloticus individuals, partitioned into the western(N = 16), northern (N = 16), and southern (N = 40) lineages (Fig. 1). These data werecompiled from museum specimens as well as contemporary samples, and collection datesranged from 1878 to 2014 (Supplementary Table S1).

Ecological niche models

To examine the present-day climatic niche of each V. niloticus lineage, we used thetemperature and precipitation variables developed by Hijmans et al. (2005) as well aselevation data (available from http://www.worldclim.org) at 2.5 arc-minute resolution.These variables represent the current climate averaged across the years 1950–2000 andanalyses were carried out using the WGS 1984 projection. The environmental data werereduced to six variables, based on physiological constraints of the organism, contributionto the model performance based on jackknife replicates, and visual inspection of envi-ronmental data quality. The variables included temperature seasonality (BIO4) and tem-perature annual range (BIO7), as well as precipitation (BIO12, BIO13, and BIO16) andelevation. Because of their dependence on permanent sources of water (Lenz 2004), weexpected precipitation to be an important factor in determining the distribution of V.niloticus populations.

Ecological niche models (ENMs) were generated using the machine-learning maximumentropy software, Maxent v3.3.3 (Phillips et al. 2006). This method outperforms othermodeling methods when predicting the current distribution of a given species (Elith et al.2006). We generated ENMs for each genetic lineage independently, using environmentallayers clipped to the African continent and masking regions occupied by the other lineages.Because Maxent treats background points as pseudo-absences (Phillips et al. 2006), regions

Evol Ecol

123


not occupied by any of the lineages (i.e. the Sahara desert) were not masked to allow themodel to incorporate less suitable environmental conditions. The distribution of eachlineage was defined by a minimum convex polygon surrounding all occurrence points witha 100 km buffer to account for dispersal distance. The regularization multiplier, whichadjusts the specificity of the model to the training points, was varied from 1 to 5 at 0.5increments and the resulting models were evaluated based on the area under the receiver-operating characteristic curve (AUC) score. Higher AUC scores (closer to 1.0) indicate agreater predictive ability, while scores near 0.5 have poor predictive ability (no better thanrandom). From this assessment, a regularization multiplier of 1.0 was selected for subse-quent models.

Models were evaluated using a fourfold cross-validation method, where occurrencepoints were randomly divided into four groups. Models were then trained on three groupsand tested on the remaining group. For each fold, we calculated AUC scores as well as theomission rate of the test data based on binary predictions using the minimum trainingpresence (Pearson et al. 2007). Binomial probabilities were calculated to assess whetherthe omission rate was lower than expected compared to a random prediction (Andersonet al. 2002). Due to the small number of occurrence records for two of the lineages, weadditionally evaluated the models with 100 bootstrap replicates in which 50 % of thepresence localities were randomly selected with replacement for training/testing the model.Across all replicates, we calculated the average AUC scores, average omission rate, andaverage binomial probabilities. ENMs for each lineage were projected onto the Africancontinent without clamping to avoid predicting regions with environmental conditions notevaluated by the model. We then created suitability maps based on the 10 percentiletraining presence to account for potential outliers.

Niche conservatism versus divergence

To assess the degree of niche overlap among lineages, we performed pairwise analyses,examining the niche space between western/northern, western/southern, and north-ern/southern lineage pairs. Overlap among niche models was assessed using Schoener’sD (1968) and Hellinger’s I, calculated in ENMTools (Warren et al. 2010). These metricsrange from 0 to 1, indicating no niche overlap and complete niche overlap, respectively(Warren et al. 2008). Both indices are based on the difference in occurrence probabilitiesbetween the species/lineages summed across all cells of the study area (Warren et al.2008). While Schoener’s D may suggest a biological interpretation, reflecting relative useof microhabitats and/or prey items, Hellinger’s I is strictly based on probability distribu-tions over geographic space (Warren et al. 2008).

To determine whether the calculated niche similarity metrics were significant, weperformed the identity test in ENMTools. In this method, the occurrence points fromlineage pairs were pooled, the lineage identities randomized, and then partitioned into twonew populations. For each generated data set, new ENMs from the simulated populationswere produced and niche similarity indices calculated. The actual observed calculations ofSchoener’s D and Hellinger’s I were then compared to the null distribution, based on 100simulations, to test whether they significantly differ. A hypothesis of identical niche spacecan be rejected if the observed niche similarity calculations are significantly lower than thevalues generated from the random data sets (Warren et al. 2010). We plotted the results ofthis test using the ggplot2 package (Wickham 2009) in R v3.1.3 (R Core Team 2014) andassessed the significance with a one-sided Wilcoxon test.

Evol Ecol

123


Because the three V. niloticus lineages occupy distinct geographic distributions, it isalso important to test whether the ENMs are more divergent than would be expected giventhe inherent environmental discrepancies between the various geographic regions.Therefore, we implemented the background test in ENMTools (Warren et al. 2010), whichcompares the actual ENM of one population to an ENM generated from random back-ground points within the other population’s distribution, and vice versa (Warren et al.2010). The actual niche similarity measures can then be compared to the null distributionbased on overlap values from 100 replicates in each direction. If the observed values aresignificantly higher/lower than the null distribution, then niches are more similar/divergentthan expected based on habitat availability of the region (Warren et al. 2010).

Within each lineage distribution, the number of random points generated was equal tothe number of original occurrence points multiplied by 100 (Warren et al. 2010). Theresulting null distributions for each pairwise comparison were plotted using ggplot2(Wickham 2009) and significance was assessed using a two-sided Wilcoxon test and analpha level of 0.05.

Historical and future distributions

ENMs were projected onto climate models of the Mid-Holocene time period, approxi-mately 6000 years bp (years before the present) as well as the Last Glacial Maximum,around 22,000 years bp using the Community Climate System Model, CCSM4 (Gent et al.2011). Past suitability was mapped based on the 10 percentile training presence. Areas thatwere suitable across all three time periods, including present-day, were considered stable.

To examine how the distribution of V. niloticus would be affected by climate change,we projected ENMs onto future climate scenarios for the year 2070 using the CommunityClimate System Model, CCSM4 (Gent et al. 2011). We selected the most severe climateprojection, incorporating the highest concentration of greenhouse gases (RepresentativeConcentration Pathway 8.5) to assess the worst-case scenario for V. niloticus populations.

Table 1 Model performance for Maxent models of the three Varanus niloticus lineages, evaluated withfourfold cross-validation and 100 bootstrap replicates

Western lineage Northern lineage Southern lineage

N 16 16 40

Cross-validation

AUC (range) 0.878–0.917 0.750–0.876 0.797–0.857

Fold 1 omission rate 0 (P = 0.0015) 0.5 (P = 0.22) 0.1 (P\ 0.001)

Fold 2 omission rate 0 (P = 0.0016) 0.25 (P = 0.35) 0.2 (P = 0.0026)

Fold 3 omission rate 0.25 (P = 0.028) 0 (P = 0.14) 0 (P\ 0.001)

Fold 4 omission rate 0.33 (P = 0.098) 0 (P = 0.051) 0.2 (P = 0.0064)

Bootstrap (100 replicates)

AUC (average) 0.900 0.828 0.860

Omission rate (average) 0.12 (P = 0.0029) 0.14 (P = 0.023) 0.11 (P\ 0.001)

The number of occurrence records (N) are provided as well as the test AUC, omission error rate, andP values for each fold (cross-validation method) or averaged across replicates (bootstrap method) based onthe minimum training presence threshold

Evol Ecol

123


Suitability was mapped based on the 10 percentile training presence, and overlaid ontopresent-day suitability to visualize the distributional changes.

Results

The ENMs for all three lineages displayed moderate to strong predictive ability, withaverage AUC values [ 0.81. For the cross-validation evaluation method, the modelsproduced for the western and southern lineages exhibited high predictive ability withsignificant binomial tests across all but one fold of the western lineage (Table 1). Thenorthern lineage ENM displayed lower levels of discrimination and predictive ability, withbinomial tests for only one fold approaching significance (P = 0.051). The few occurrencerecords in the northernmost extent of this lineage’s range compounded with the consid-erable variation in climate throughout northern Africa could account for the reduced modelperformance. In contrast, evaluating the models with the bootstrap method produced sig-nificant binomial probabilities across all lineages (Table 1).

All pairwise ENM comparisons produced high values for Hellinger’s I and Schoener’sD and showed high degrees of geographic overlap (Table 2; Figs. 2, 3, 4). The observedvalues of I and D overlapped with the null distributions from the identity test in all pairwisecomparisons (Supplementary Fig. S1). For the western/northern comparison, the Wilcoxontest resulted in a nonsignificant P value; therefore, we could not reject the hypothesis ofidentical niche space. Comparisons of the western/southern and northern/southern lineagesproduced significant P values for the identity test, leading to rejections of the identicalniche space hypothesis.

For the background test, all observed values of I and D were significantly higher thanthe null distributions, and the null hypothesis could be rejected in all cases (Table 2;Figs. 2, 3, 4). These results provide evidence that the climatic niches of all lineages aremore conserved than expected, based on habitat availability in the regions.

Because all lineage comparisons showed evidence of niche conservatism, the combinedoccurrence data from all lineages across Africa (N = 72) were used to create a V. niloticusENM that we subsequently projected to past time periods. While the projected historicdistributions showed increased fragmentation compared to the present-day distribution ofV. niloticus, no regions were completely isolated in the time periods examined (Fig. 5).

Table 2 Results of niche overlap analyses for pairwise comparisons of Varanus niloticus lineages

Pairwisecomparison

D I Identity testP values

Background testP values

Inference

D I D I

Western/northern 0.735 0.932 0.99 1.0 \ 0.001,\ 0.001

\ 0.001,\ 0.001

Conservative

Western/southern 0.747 0.947 \ 0.001 \ 0.001 \ 0.001,\ 0.001

\ 0.001,\ 0.001

Conservative

Northern/southern 0.752 0.930 \ 0.001 \ 0.001 \ 0.001,\ 0.001

\ 0.001,\ 0.001

Conservative

Identity and background tests were based on 100 replicates. D = Schoener’s D; I = Hellinger’s I. SeeFigs. 2, 3, and 4 and Supplementary Fig. 1 for histograms

Evol Ecol

123


The future climate projections for the year 2070 under the most severe climate changescenario produced similar distributions for all lineages as well as the full species distri-bution (Fig. 6). All lineages showed slight range contractions in the westernmost regions ofthe continent and sporadically throughout central Africa. However, overall the distributionsremained similar across current and future climates.

Discussion

Species delimitation is often bolstered by findings of divergent ecological niche space(Blair et al. 2013; Raxworthy et al. 2007); however, our results strongly support nicheconservatism across all V. niloticus lineages as assessed with climate and elevation data.The heavy reliance on museum occurrence records in this study (with collection dates asearly as the late 1800s) prevented the use of vegetation data, with habitat degradation,agriculture, and general ecological changes making the vegetation composition unlikely tohave remained constant across the past century. However, future studies should be aimed atrecognizing the fine-scale habitat use differences, if any, among the genetically distinctlineages of V. niloticus. Although the collection dates for some of the museum specimens

0 1,500 3,000750 Km

Suitability

Western lineage

Northern lineage

Major Rivers

Not suitable

Western

Northern

Both

0

5

10

0.5 0.6 0.7 0.8 0.9

Hellinger's I

Freq

uenc

y

0

5

10

15

0.3 0.4 0.5 0.6 0.7

Schoener's D

Freq

uenc

y

Fig. 2 (Left) Occurrence points and ecological niche models (ENMs) for the western and northern lineagesof Varanus niloticus. Suitability was based on the 10 percentile training presence. (Right) Histogramsshowing results of the background test performed with 100 replicates. The dashed lines denote the observedvalues of Hellinger’s I (0.932) and Schoener’s D (0.735) with significant P values for both niche overlapmeasures (P\ 0.001)

Evol Ecol

123


fell outside of the time range for the current climate data (averaged across the years1950–2000), the climatic conditions likely remained relatively constant. Therefore, we donot expect that the differing collection dates significantly affected the accuracy of theclimate models. While our findings do not help to resolve the taxonomic designations ofthese genetically and geographically disparate groups, this study provided insight into theevolutionary mechanisms that have influenced the observed genetic structure of V.niloticus.

All three V. niloticus lineages were found to occupy a more restricted geographic rangethan the area predicted to be climatically suitable. This overprediction of ENMs trained onindividual lineages demonstrates that, in general, factors other than the climate variablesexamined are limiting the dispersal and gene flow of the genetically distinct V. niloticuspopulations. However, inspection of the spatial overlap between ENMs revealed that thewestern lineage is climatically more restricted than the northern lineage (Fig. 2) and couldpotentially explain the low level of gene flow reported between geographically adjacentpopulations in West Africa (Dowell et al. 2015). Future studies assessing the fine-scaleecological characteristics in relation to genetic patterns could clarify the role of environ-mental factors in limiting gene flow among V. niloticus populations.

0 1,500 3,000750 Km

Suitability

Western lineage

Southern lineage

Major Rivers

Not suitable

Western

Southern

Both

0

5

10

15

0.3 0.4 0.5 0.6 0.7Schoener's D

0

5

10

15

20

0.6 0.7 0.8 0.9Hellinger's I

Freq

uenc

yFr

eque

ncy

Fig. 3 (Left) Occurrence points and ecological niche models (ENMs) for the western and southern lineagesof Varanus niloticus. Suitability was based on the 10 percentile training presence. (Right) Histogramsshowing results of the background test performed with 100 replicates. The dashed lines denote the observedvalues of Hellinger’s I (0.947) and Schoener’s D (0.747) with significant P values for both niche overlapmeasures (P\ 0.001)

Evol Ecol

123


Assuming the climatic niches have remained stable across evolutionary time, ourfindings of niche conservatism among all three V. niloticus lineages support an allopatricmode of divergence (Wiens and Graham 2005). These results are consistent with the idea

0 1,500 3,000750 Km

Suitability

Northern lineage

Southern lineage

Major Rivers

Not suitable

Northern

Both

Southern

0

5

10

15

20

25

0.4 0.6 0.8Hellinger's I

Freq

uenc

y0

5

10

15

0.2 0.4 0.6 0.8Schoener's D

Freq

uenc

y

Fig. 4 (Left) Occurrence points and ecological niche models (ENMs) for the northern and southern lineagesof Varanus niloticus. Suitability was based on the 10 percentile training presence. (Right) Histogramsshowing results of the background test performed with 100 replicates. The dashed lines denote the observedvalues of Hellinger’s I (0.930) and Schoener’s D (0.752) with significant P values for both niche overlapmeasures (P\ 0.001)

mK005,1 000,6000,30Not suitable Suitable

Current Mid-Holocene Last Glacial Maximum Stability(~22,000 yrs bp)(~6,000 yrs bp)

Fig. 5 Ecological niche models (ENMs) created with occurrence data from all lineages of Varanusniloticus and projected onto past climate models. Suitability was based on the 10 percentile trainingpresence. Areas that were suitable in all three time periods were considered stable

Evol Ecol

123


that populations became isolated in refugia during historic climatic shifts (Haffer 1969).The distributions of the three lineages, and the contact zones among them, are therefore,likely a result of secondary contact. Projecting the ENM onto past climate predictions ofthe Mid-Holocene and Last Glacial Maximum revealed a more fragmented distribution.However, because no region was found to be completely isolated by unsuitable climaticconditions, these results could not completely explain the current phylogeographicpatterns.

Not Suitable

Full Species

Suitability

0 1,500 3,000750 Km

Current

2070

Both

Northern Lineage

Suitability

Not Suitable

Current

2070

Both

Not Suitable

Current

2070

Both

Not Suitable

Current

2070

Both

Western Lineage

Suitability

Southern Lineage

Suitability

Fig. 6 Ecological niche models (ENMs) created with occurrence data from each Varanus niloticus lineage,individually and combined, and projected onto a future climate scenario based on the Community ClimateSystem Model (CCSM4) for the year 2070 (Representative Concentration Pathway 8.5). Suitability underthe current climate was also overlaid on each map. For both time periods, suitability was based on the10 percentile training presence

Evol Ecol

123


Molecular-based dating analyses have suggested that the V. niloticus lineages divergedprior to the time periods presented here. The western lineage diverged from the otherlineages during the Late Miocene (Dowell et al. 2016). This time period marked the onsetof drastic climatic events, transitioning from a warm, humid climate to a cooler, more aridenvironment (Axelrod and Raven 1978; Coetzee 1993; Plana 2004). Additionally, thenorthern and southern lineages are thought to have diverged during the Early Pliocene(Dowell et al. 2016). In contrast to the events of the Miocene, this time period wascharacterized by increased humidity with the accompanying expansion and reconnection ofthe rainforest (Coetzee 1993; Plana 2004). Congruent north–south genetic breaks havebeen observed in numerous savannah-dwelling species and are largely attributed to thedense rainforest belt that extended across the continent, dividing suitable habitat to thenorth and south (Lorenzen et al. 2012). Climatic niche conservatism could have facilitatedthe isolation of V. niloticus populations during these severe climatic oscillations.

Our findings have important conservation implications for V. niloticus populations asglobal climate change becomes an increasing threat to biodiversity (Dawson et al. 2011;Thomas et al. 2004). While evolutionarily labile species can potentially adapt to thechanging climatic conditions, species exhibiting strong climatic niche conservatism musteither disperse to new areas or risk extinction (Holt 1990; Wiens et al. 2010). Although ourresults suggest that V. niloticus will be relatively unaffected by climate change, based onthe Community Climate System Model, other climate scenarios have predicted theexpansion of the Sahara desert southward into the Sahel region of West Africa (De Wit andStankiewicz 2006). This scenario would severely reduce or fragment the currentlyrestricted geographic range of the western V. niloticus lineage unless substantial migrationto other suitable areas occurs. Populations of V. niloticus in this area are under additionalpressure from the leather trade, with intensive harvesting concentrated in Sahelian Africa(de Buffrenil and Castanet 2000; De Lisle 1996; Jenkins and Broad 1994). Some suggestthat V. niloticus populations in this region are currently in danger of extirpation (De Lisle1996). With the combined threats of exploitation and the potential habitat loss due toclimate change, the western V. niloticus lineage should be of high conservation priority.

Our study provides increased support for the role of niche conservatism in shaping theevolutionary patterns within and among species. Stability in niche space across evolu-tionary timescales has been reported in a number species and is thought to facilitateallopatric speciation, generating present-day patterns of biodiversity (Knouft et al. 2006;Kozak and Wiens 2006; Peterson et al. 1999; Wiens 2004a). ENMs can further the field ofphylogeography by exploring species distributions during past climate conditions (Al-varado-Serrano and Knowles 2014; Carnaval and Moritz 2008; Hugall et al. 2002). Whileour results of niche conservatism could not provide further support for the taxonomicclassifications of V. niloticus lineages, ENMs have been useful in species delimitation aswell as identifying cryptic species (Blair et al. 2013; Florio et al. 2012; Raxworthy et al.2007; Rissler and Apodaca 2007; Sites and Marshall 2003). Lastly, identifying speciesgroups that exhibit niche conservatism and estimating species distributions under futureclimate scenarios can focus conservation efforts to mitigate the effects of global climatechange (Gomez-Mendoza and Arriaga 2007; Midgley et al. 2003; Pearson and Dawson2003; Peterson et al. 2002; Thuiller et al. 2005).

Acknowledgments We would like to sincerely thank Mary Blair, Eleanor Sterling, and Camilo Sanin fortheir assistance, guidance, and support throughout this project. Additionally, we would like to thank Viviande Buffrenil, Daniel Portik, Ivan Ineich, Eli Greenbaum, Richard Fergusson, and Louis la Grange forproviding GPS coordinates from contemporary samples. Occurrence records from historic samples were

Evol Ecol

123


provided by the American Museum of Natural History, California Academy of Sciences, Museum ofComparative Zoology (Harvard), Museum national d’Histoire naturelle-RA, Museum of Vertebrate Zool-ogy, Natural History Museum of London, University of Michigan Museum of Zoology, SmithsonianNational Museum of Natural History, Burke Museum of Natural History and Culture, Port ElizabethMuseum, and the South Africa Biodiversity Institute. Lastly, we are grateful for the editorial assistance fromTim Vines and Leslie Rissler at Axios Review, as well as the three anonymous reviewers who contributedvaluable suggestions.

References

Alvarado-Serrano DF, Knowles LL (2014) Ecological niche models in phylogeographic studies: applica-tions, advances and precautions. Mol Ecol Resour 14:233–248

Anderson RP, Gomez-Laverde M, Peterson AT (2002) Geographical distributions of spiny pocket mice inSouth America: insights from predictive models. Glob Ecol Biogeogr 11:131–141

Avise JC (2000) Phylogeography: the history and formation of species. Harvard University Press,Cambridge

Axelrod D, Raven P (1978) Late cretaceous and tertiary vegetation history of Africa. In: Werger M (ed)Biogeography and ecology of southern Africa. Junk, The Hague, pp 77–130

Bayless M (1997) The distribution of African monitor lizards (Sauria: Varanidae). Afr J Ecol 35:374–377Blair M, Sterling E, Dusch M, Raxworthy C, Pearson R (2013) Ecological divergence and speciation

between lemur (Eulemur) sister species in Madagascar. J Evol Biol 26:1790–1801Carnaval AC, Moritz C (2008) Historical climate modelling predicts patterns of current biodiversity in the

Brazilian Atlantic forest. J Biogeogr 35:1187–1201Coetzee J (1993) African flora since the terminal Jurassic. In: Goldblatt P (ed) Biological relationships

between Africa and South America, vol 37. Yale University Press, New Haven, pp 37–61Coyne JA, Orr HA (2004) Speciation. Sinauer Associates Inc., SunderlandDaudin FM (1802) Histoire naturelle, generale et particuliere des reptiles (trans: Ivan Ineich). Paris, FranceDawson TP, Jackson ST, House JI, Prentice IC, Mace GM (2011) Beyond predictions: biodiversity con-

servation in a changing climate. Science 332:53–58Dayrat B (2005) Towards integrative taxonomy. Biol J Linn Soc 85:407–415de Buffrenil V (1995) Monitor hunting. CITES/C&M Int Mag 4:6–20de Buffrenil V, Castanet J (2000) Age estimation by skeletochronology in the Nile monitor (Varanus

niloticus), a highly exploited species. J Herpetol 34:414–424De Lisle HF (1996) The natural history of monitor lizards. Krieger Publishing Co., MalabarDe Wit M, Stankiewicz J (2006) Changes in surface water supply across Africa with predicted climate

change. Science 311:1917–1921Dowell SA, de Buffrenil V, Kolokotronis S-O, Hekkala ER (2015) Fine-scale genetic analysis of the

exploited Nile monitor (Varanus niloticus) in Sahelian Africa. BMC Genet 16:32Dowell SA, Portik DM, de Buffrenil V, Ineich I, Greenbaum E, Kolokotronis S-O, Hekkala E (2016)

Molecular data from contemporary and historical collections reveal a complex story of crypticdiversification in the Varanus (Polydaedalus) niloticus species group. Mol Phylogenet Evol94:591–604. doi:10.1016/j.ympev.2015.10.004

Dubois A (2003) The relationships between taxonomy and conservation biology in the century of extinc-tions. C R Biol 326:9–21

Elith J, Graham CH, Anderson RP et al (2006) Novel methods improve prediction of species’ distributionsfrom occurrence data. Ecography 29:129–151

Evans ME, Smith SA, Flynn RS, Donoghue MJ (2009) Climate, niche evolution, and diversification of the‘‘bird-cage’’ evening primroses (Oenothera, Sections Anogra and Kleinia). Am Nat 173:225–240

Flagstad Ø, Syversten PO, Stenseth NC, Jakobsen KS (2001) Environmental change and rates of evolution:the phylogeographic pattern within the hartebeest complex as related to climatic variation. Proc R SocLond Ser B Biol Sci 268:667–677

Florio A, Ingram C, Rakotondravony H, Louis E, Raxworthy C (2012) Detecting cryptic speciation in thewidespread and morphologically conservative carpet chameleon (Furcifer lateralis) of Madagascar.J Evol Biol 25:1399–1414

Gent PR, Danabasoglu G, Donner LJ et al (2011) The community climate system model version 4. J Clim24:4973–4991

Gomez-Mendoza L, Arriaga L (2007) Modeling the effect of climate change on the distribution of oak andpine species of Mexico. Conserv Biol 21:1545–1555

Evol Ecol

123


Graham CH, Ron SR, Santos JC, Schneider CJ, Moritz C (2004) Integrating phylogenetics and environ-mental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution 58:1781–1793

Haffer J (1969) Speciation in Amazonian forest birds. Science 165:131–137Hekkala E, Shirley MH, Amato G et al (2011) An ancient icon reveals new mysteries: mummy DNA

resurrects a cryptic species within the Nile crocodile. Mol Ecol 20:4199–4215Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate

surfaces for global land areas. Int J Climatol 25:1965–1978Holt RD (1990) The microevolutionary consequences of climate change. Trends Ecol Evol 5:311–315Hugall A, Moritz C, Moussalli A, Stanisic J (2002) Reconciling paleodistribution models and comparative

phylogeography in the Wet Tropics rainforest land snail Gnarosophia bellendenkerensis (Brazier1875). Proc Natl Acad Sci USA 99:6112–6117

Jenkins M, Broad S (1994) International trade in reptile skins: a review and analysis of the main consumermarkets, 1983–1991. Traffic International. Report no. 1858500478

Keyghobadi N, Roland J, Strobeck C (1999) Influence of landscape on the population genetic structure ofthe alpine butterfly Parnassius smintheus (Papilionidae). Mol Ecol 8:1481–1495

Knouft JH, Losos JB, Glor RE, Kolbe JJ (2006) Phylogenetic analysis of the evolution of the niche in lizardsof the Anolis sagrei group. Ecology 87:S29–S38

Kozak KH, Wiens J (2006) Does niche conservatism promote speciation? A case study in North Americansalamanders. Evolution 60:2604–2621

Leache AD, Koo MS, Spencer CL, Papenfuss TJ, Fisher RN, McGuire JA (2009) Quantifying ecological,morphological, and genetic variation to delimit species in the coast horned lizard species complex(Phrynosoma). Proc Natl Acad Sci USA 106:12418–12423

Lenz S (2004) Varanus niloticus. In: Pianka ER, King D, King RA (eds) Varanoid lizards of the world.Indiana University Press, Bloomington, pp 133–138

Lorenzen ED, Heller R, Siegismund HR (2012) Comparative phylogeography of African savannah ungu-lates. Mol Ecol 21:3656–3670

May RM (1990) Taxonomy as destiny. Nature 347:129–130Mendez M, Rosenbaum HC, Subramaniam A, Yackulic C, Bordino P (2010) Isolation by environmental

distance in mobile marine species: molecular ecology of franciscana dolphins at their southern range.Mol Ecol 19:2212–2228

Midgley GF, Hannah L, Millar D, Thuiller W, Booth A (2003) Developing regional and species-levelassessments of climate change impacts on biodiversity in the Cape Floristic Region. Biol Conserv112:87–97

Moodley Y, Bruford MW (2007) Molecular biogeography: towards an integrated framework for conservingpan-African biodiversity. PLoS One 2:e454

Pearson RG, Dawson TP (2003) Predicting the impacts of climate change on the distribution of species: arebioclimate envelope models useful? Global Ecol Biogeogr 12:361–371

Pearson RG, Raxworthy CJ, Nakamura M, Townsend Peterson A (2007) Predicting species distributionsfrom small numbers of occurrence records: a test case using cryptic geckos in Madagascar. J Biogeogr34:102–117

Pernetta AP (2009) Monitoring the trade: using the CITES database to examine the global trade in livemonitor lizards (Varanus spp.). Biawak 3:37–45

Peterson AT, Holt RD (2003) Niche differentiation in Mexican birds: using point occurrences to detectecological innovation. Ecol Lett 6:774–782

Peterson A, Soberon J, Sanchez-Cordero V (1999) Conservatism of ecological niches in evolutionary time.Science 285:1265–1267

Peterson AT, Ortega-Huerta MA, Bartley J, Sanchez-Cordero V, Soberon J, Buddemeier RH, Stockwell DR(2002) Future projections for Mexican faunas under global climate change scenarios. Nature416:626–629

Phillips SJ, Anderson RP, Schapire RE (2006) Maximum entropy modeling of species geographic distri-butions. Ecol Model 190:231–259

Plana V (2004) Mechanisms and tempo of evolution in the African Guineo-Congolian rainforest. PhilosTrans R Soc Lond Ser B Biol Sci 359:1585–1594

Raxworthy CJ, Ingram CM, Rabibisoa N, Pearson RG (2007) Applications of ecological niche modeling forspecies delimitation: a review and empirical evaluation using day geckos (Phelsuma) from Mada-gascar. Syst Biol 56:907–923

Rissler LJ, Apodaca JJ (2007) Adding more ecology into species delimitation: ecological niche models andphylogeography help define cryptic species in the black salamander (Aneides flavipunctatus). Syst Biol56:924–942

Evol Ecol

123


Schlegel H (1844) Abbildungen neuer oder unvollstandig bekannter Amphibien: nach der Natur oder demLeben entworfen. Verlag von Arnz & Comp., Dusseldorf

Schliewen U, Rassmann K, Markmann M, Markert J, Kocher T, Tautz D (2001) Genetic and ecologicaldivergence of a monophyletic cichlid species pair under fully sympatric conditions in Lake Ejagham,Cameroon. Mol Ecol 10:1471–1488

Schoener TW (1968) The Anolis lizards of Bimini: resource partitioning in a complex fauna. Ecology49:704–726

Sites JW, Marshall JC (2003) Delimiting species: a Renaissance issue in systematic biology. Trends EcolEvol 18:462–470

R Core Team (2014) R: a language and environment for statistical computing. R Foundation for StatisticalComputing, Vienna

Thomas CD, Cameron A, Green RE et al (2004) Extinction risk from climate change. Nature 427:145–148Thuiller W, Lavorel S, Araujo MB, Sykes MT, Prentice IC (2005) Climate change threats to plant diversity

in Europe. Proc Natl Acad Sci USA 102:8245–8250Warren DL, Glor RE, Turelli M (2008) Environmental niche equivalency versus conservatism: quantitative

approaches to niche evolution. Evolution 62:2868–2883Warren DL, Glor RE, Turelli M (2010) ENMTools: a toolbox for comparative studies of environmental

niche models. Ecography 33:607–611Welton LJ, Travers SL, Siler CD, Brown RM (2014) Integrative taxonomy and phylogeny-based species

delimitation of Philippine water monitor lizards (Varanus salvator Complex) with descriptions of twonew cryptic species. Zootaxa 3881:201–227

Wickham H (2009) ggplot2: elegant graphics for data analysis. Springer, New YorkWiens JJ (2004a) Speciation and ecology revisited: phylogenetic niche conservatism and the origin of

species. Evolution 58:193–197Wiens JJ (2004b) What is speciation and how should we study it? Am Nat 163:914–923Wiens JJ, Graham CH (2005) Niche conservatism: integrating evolution, ecology, and conservation biology.

Annu Rev Ecol Evol Syst 36:519–539Wiens JJ, Ackerly DD, Allen AP et al (2010) Niche conservatism as an emerging principle in ecology and

conservation biology. Ecol Lett 13:1310–1324

Evol Ecol

123


Documents

WordPress.com...ORIGINAL PAPER Divergent lineages and conserved niches: using ecological niche modeling to examine the evolutionary patterns of the Nile monitor (Varanus niloticus)