Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

Andreas F. Kautt; Gonzalo Machado‐Schiaffino; Julian Torres‐Dowdall; Axel Meyer

doi:10.1002/ece3.2287

Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

Kautt, Andreas F.; Machado‐Schiaffino, Gonzalo; Torres‐Dowdall, Julian; Meyer, Axel 2016-08-01 00:00:00 Demographic inference, habitat isolation, Understanding how speciation can occur without geographic isolation remains individual specialization, phenotypic plasticity, a central objective in evolutionary biology. Generally, some form of disruptive RAD-seq, stable isotopes. selection and assortative mating are necessary for sympatric speciation to occur. Correspondence Disruptive selection can arise from intraspeciﬁc competition for resources. If Axel Meyer, Department of Biology, this competition leads to the differential use of habitats and variation in rele- University of Konstanz, Universit€atsstrasse 10, vant traits is genetically determined, then assortative mating can be an auto- 78457 Konstanz, Germany. matic consequence (i.e., habitat isolation). In this study, we caught Midas Tel: +497531884163; cichlid ﬁsh from the limnetic (middle of the lake) and benthic (shore) habitats Fax: +497531883018; of Crater Lake Asososca Managua to test whether some of the necessary condi- E-mail: [email protected] tions for sympatric speciation due to intraspeciﬁc competition and habitat iso- Funding Information lation are given. Lake As. Managua is very small (<900 m in diameter), European Research Council (Grant/Award extremely young (maximally 1245 years of age), and completely isolated. It is Number: “GenAdap” 293700), Deutsche inhabited by, probably, only a single endemic species of Midas cichlids, Forschungsgemeinschaft (Grant/Award Amphilophus tolteca. We found that ﬁsh from the limnetic habitat were more Number: MA 6144/1-1, TO 914/2-1). elongated than ﬁsh collected from the benthic habitat, as would be predicted from ecomorphological considerations. Stable isotope analyses conﬁrmed that Received: 18 April 2016; Revised: 8 June 2016; Accepted: 9 June 2016 the former also exhibit a more limnetic lifestyle than the latter. Furthermore, split-brood design experiments in the laboratory suggest that phenotypic plas- Ecology and Evolution 2016; 6(15): 5342– ticity is unlikely to explain much of the observed differences in body elongation that we observed in the ﬁeld. Yet, neutral markers (microsatellites) did not reveal any genetic clustering in the population. Interestingly, demographic doi: 10.1002/ece3.2287 inferences based on RAD-seq data suggest that the apparent lack of genetic dif- ferentiation at neutral markers could simply be due to a lack of time, as *These authors contributed equally to this intraspeciﬁc competition may only have begun a few hundred generations ago. work. resources (Martin and Pfennig 2009). The conditions for Introduction sympatric speciation are very restricted (Gavrilets 2005), Empirical studies of the conditions that may allow for or and only if the strength of disruptive selection and the constrain the process of speciation with gene ﬂow and probability of assortative mating are sufﬁciently strong especially its most extreme form, sympatric speciation, may sympatric speciation occur (Udovic 1980). Moreover, are essential to test theoretical predictions (Bolnick 2011; if there are costs associated with choosing a mate, the Martin 2013; Meyer and Kautt 2014; Comeault et al. likelihood of sympatric speciation is strongly reduced 2015). Generally, theory has revealed that ecological sym- (Matessi et al. 2001). On the other hand, if only few loci patric speciation requires some form of disruptive selec- of large effect are underlying ecologically relevant traits tion and assortative mating to overcome the and those conferring assortative mating and both are clo- homogenizing process of gene ﬂow and recombination sely genetically linked (e.g., in close physical proximity or (Felsenstein 1981; Gavrilets 2003, 2004). Disruptive selec- located on an inversion), sympatric speciation is facili- tion can arise from intraspeciﬁc competition, which may tated as their association will be less frequently broken be apparent by individual specialization to certain food down by recombination (Trickett and Butlin 1994; 5342 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Rieseberg 2001). In the most extreme case, traits under earliest stages, holds great promise to identify the relevant selection and those that confer assortative mating are one mechanisms and conditions driving speciation (Via 2009). and the same (Gavrilets 2004). Consequently, if individual Different crater lakes located even within a narrow geo- specialization leads automatically to assortative mating graphic area are often of different age and usually extre- due to the spatial segregation of mating pools (habitat mely remote, isolated bodies of water that are typically isolation) or inﬂuences the propensity of individuals to not connected to other lakes or rivers, making multiple mate with others of a similar phenotype (mate prefer- invasions rather unlikely (but see Martin et al. 2015). ence), population divergence in sympatry is greatly facili- Thus, in contrast to ﬁsh in postglacial lakes in which tated (Gavrilets 2004; Servedio et al. 2011; Smadja and divergence may have often happened due to character dis- Butlin 2011). Importantly, the variation in ecologically placement after a secondary invasion (Schluter 1996), relevant traits and thus individual specialization has to be divergence in crater lake cichlids is often assumed to have genetically determined and not due to phenotypic plastic- happened in sympatry. This should allow for a more ity: While adaptive phenotypic plasticity might increase direct study of the conditions that can lead to speciation the potential for ecological speciation in the long term by with gene ﬂow, not contingent or affected by the initial allowing for the colonization of new environments and amount of population divergence and reproductive isola- population persistence, it will usually impede population tion that was already present at the time of secondary divergence in sympatry (Bolnick 2011; Thibert-Plante and contact. Hendry 2011). Similarly to ﬁsh in postglacial lakes, on the other hand, Thus, some of the necessary (but not sufﬁcient) condi- the evolutionary outcomes of crater lake cichlid popula- tions for sympatric speciation due to intraspeciﬁc compe- tions vary. Haplochromine cichlids in Uganda, for exam- tition for resources and habitat isolation are as follows: ple, have repeatedly evolved an overall more limnetic (1) individuals differ in ecomorphological traits, (2) their body shape without diversifying after the colonization of ecomorphological characteristics match, and individuals crater lakes (Machado-Schiafﬁno et al. 2015). In a small differentially use, distinct habitats, and (3) variation in crater lake in Tanzania, cichlids are diverging along the ecologically relevant traits is genetically determined and littoral zone and the deeper benthic areas instead of the not due to phenotypic plasticity. At the genomic level, benthic–limnetic axis (Malinsky et al. 2015). differentiation in this process is initially expected to be Midas cichlids (belonging to the Amphilophus sp. spe- only reﬂected at the few loci under disruptive selection, cies complex) that have independently colonized several whereas genome-wide differentiation will be virtually crater lakes in Nicaragua from the same source popula- absent and only build up with time after gene ﬂow has tion of the great lakes Nicaragua and Managua, but prob- ceased (Wu 2001; Nosil et al. 2009a; Feder et al. 2012). ably at different time points, show a variable pattern Competition for resources is thought to be the main (Elmer et al. 2010b): They have independently evolved driver of divergence between the ground-associated (ben- into one limnetic and several benthic species in the two thic) and open-water (limnetic) habitats in freshwater Crater Lakes Apoyo and Xiloa (Barluenga et al. 2006; ﬁsh, which has occurred in a variety of ﬁsh taxa in tem- Kautt et al. 2012; Elmer et al. 2014), whereas in other perate as well as tropical environments (Robinson and crater lakes, they have not diversiﬁed along the benthic– Wilson 1994; Hulsey et al. 2013; Elmer et al. 2014). Lim- limnetic axis (Elmer et al. 2010b). We note that the lim- netic ecomorphs exhibit characteristic elongated/fusiform netic niche is completely absent in the relatively shallow body shapes compared to the rather high-bodied benthic (mean depth around 8–12 m) and turbid source lakes ecomorphs (Webb 1984, 1988). Although ﬁsh have diver- (Barlow 1976; Elmer et al. 2010b). One potential explana- siﬁed along the benthic–limnetic axis in a considerable tion is that the variability in body elongation is a key fac- number of cases, still more commonly ﬁsh have not tor that inﬂuences the potential for intrapopulation diversiﬁed and a single generalist population occupies a competition and thus frequency-dependent disruptive lake (Bolnick 2011). Identifying the conditions that have selection. Within the repeated adaptive radiations of facilitated or hindered population divergence and specia- Midas cichlids (where 13 species have been described so tion remains thus a key objective in enhancing our ability far; Recknagel et al. 2013b), this variability seems to be to predict evolutionary processes. correlated with the mean depth of the respective crater Crater lakes in Africa and Nicaragua inhabited by cich- lake (Recknagel et al. 2014). lid ﬁshes represent a unique setting to investigate the pro- In this regard, Crater Lake Asososca Managua – the cess of sympatric speciation in a replicated manner and at focal lake of this study – has a mean depth (54.3 m) sim- different stages of the divergence continuum (Elmer et al. ilar to that of Crater Lake Xiloa (60 m). Further, the pop- 2010b; Kautt et al. 2012; Malinsky et al. 2015). Studying ulation in Lake As. Managua exhibits a large variation in different stages of population divergence, including the body shape elongation, and individual body shapes are ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5343 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. correlated with their lower pharyngeal jaw shapes and cichlids form seasonally monogamous pairs that breed at feeding ecology (Kusche et al. 2014). Thus, one of the the shore, but that pair formation happens probably in conditions (condition 1 outlined above) for divergence the respective habitat. Thus, habitat isolation would act as along the benthic–limnetic axis seems given. Yet, Crater an effective isolating barrier during the time of pair for- Lake As. Managua is thought to harbor only a single pop- mation and not breeding (Gavrilets et al. 2007). ulation of Midas cichlids (Barluenga and Meyer 2010; Another factor that had so far not been explicitly tested Kusche et al. 2014). Importantly, however, ﬁsh had previ- in Midas Cichlids is the role of phenotypic plasticity in ously only been collected at the shore and never in the body shape elongation. While a previous QTL mapping middle of the lake, where limnetic ﬁsh are thought to for- study in closely related limnetic and benthic species of age. Without a representative sample from both habitats, Midas cichlids in Crater Lake Apoyo already demon- population genetic methods are arguably underpowered strated a strong genetic component in the determination for testing population divergence. Asososca Managua is of body elongation (Franchini et al. 2014), in this study also the youngest of the crater lakes (max. 1245 years old; we investigated the contribution of plasticity to the phe- Pardo et al. 2008) known to harbor an endemic species notypic variation in body elongation observed in Midas of the adaptive radiation of Midas cichlids (Elmer et al. cichlids. In other words, we tested whether differences in 2010b), Amphilophus tolteca (Recknagel et al. 2013b), and the ecologically relevant trait of body elongation are pre- the population may thus be at a much earlier stage of sumably mostly genetically determined (condition 3) or divergence than the benthic–limnetic species pairs in Cra- can be attributed to phenotypic plasticity. This question ter Lakes Apoyo and Xiloa. In other words, the propen- was investigated with split-brood design experiments con- sity to diverge along the benthic–limnetic axis in Midas ducted in the laboratory in which one half of the ﬁsh cichlids seems to be related to the depth of the crater were reared in tanks with a constant water ﬂow and the lake, yet the extent of divergence along this axis might other half in control tanks without water ﬂow. In addi- depend on the age of the resident population (time since tion to the focal species from As. Managua, we also sub- colonization). jected broods of three other Midas cichlid species (two Here, we set out to investigate whether the second and limnetic species from different crater lakes and one ben- third conditions for incipient sympatric speciation as out- thic species from the source population of the crater lined above are present and whether population diver- lakes) to the experiment. gence may currently be happening in the extremely young Finally, we used a previously in-house generated RAD- population of A. tolteca in Crater Lake As. Managua. To seq data set from ﬁsh captured at the shore to reconstruct do so, we caught ﬁsh from the open-water zone in the the demographic history of the population in Crater Lake middle of the lake (limnetic habitat) in addition to the As. Managua using coalescent simulations and the site shore (benthic habitat) and used an integrative approach frequency spectrum. This allowed us to put our results of to test whether there is morphological (body elongation), the putative stage of population divergence in perspective ecological (long-term diet in form of stable isotope signa- to the actual time of colonization and size of the founder tures), and genetic (microsatellites) divergence between population. ﬁsh captured in the two different habitats. If the variation in body shape is the result of specialization along the ben- Materials and Methods thic–limnetic axis, we would predict ﬁsh that preferen- tially use the limnetic habitat to exhibit on average a Sampling more elongated body shape and limnetic lifestyle (i.e., a higher proportion of carbon from a limnetic source in Fish were captured with gill nets or angling in Lake Aso- their diet) compared to individuals from the benthic sosca Managua in 2013. Most of the ﬁsh (n = 250) were habitat (condition 2). So far, this prediction has never captured at the shore close to the aqueduct main station been tested in Midas cichlids, as sampling of ﬁsh usually (Fig. 1A). In addition, the protected environment of Cra- occurred exclusively close to the shore (i.e., the benthic ter Lake As. Managua (surrounded by a fence with only habitat) and during the breeding season. Differential habi- restricted access) allowed us to catch a sufﬁcient number tat use, even in a very small lake like As. Managua (ca. of ﬁsh from the middle of the lake (n = 27) by attaching 900 m in diameter), is a key factor that may increase the gill nets to a buoy; without a ﬁxed anchor point, gill nets likelihood of population divergence via habitat isolation get tangled up and leaving nets or a buoy undisturbed in (Servedio et al. 2011). In fact, habitat isolation is a neces- the middle of a lake is not feasible in other crater lakes in sary assumption in a model of sympatric speciation that Nicaragua. For ﬁsh from the middle of the lake and a was speciﬁcally tailored to Midas cichlids in Crater Lake subset (randomly chosen) of those from the shore Apoyo (Gavrilets et al. 2007). We note that Midas (n = 27), standardized photographs were taken from the 5344 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. d A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? (A) (B) Nicaragua L. Asososca Managua Amphilophus tolteca Figure 1. Crater Lake Asososca Managua is (C) located in the center of Nicaragua’s capitol, Limnetic Benthic the city of Managua, and protected from its surroundings. (A) Fish were collected in the Buoy limnetic (middle of the lake) and benthic Limnetic (shore) zones of the lake. (B) Representative specimen of the focal species of this study, Amphilophus tolteca, endemic to As. Managua. (C) Relative frequency (kernel Aqueduct density estimate) of body height index (i.e., a station 200 m measure of overall body elongation) of ﬁsh 0.40 0.42 0.44 0.46 0.48 0.50 collected from the middle of the lake (blue) Body height index (BHI) and the shore (red). lateral view including a size standard. For these samples, remeasured the BHI for a subset of ﬁsh from the shore also tissue samples were taken and preserved in pure (n = 27) – later used for stable isotope and population ethanol for stable isotope and population genetic analyses. genetic analyses – from individual photographs. These An overview about the samples and the morphological, analyzed specimens from the shore represent a random stable isotope, and microsatellite data is provided in subsample of the 250 collected individuals. This random Table S1, Supporting information. Samples for RAD selection and the two different measurement modes are sequencing used for demographic inference (n = 49) were unlikely to have biased our results, as the subset was collected in earlier ﬁeld expeditions in 2007 and 2010 at indistinguishable in terms of the BHI from the whole set the shore and processed in the same way. Sampling was of ﬁsh from the shore (subset: mean = 0.437 0.013 SD; approved and performed according to the regulations of whole set: mean = 0.438 0.014 SD; Welch two-sample the local authorities, the Ministerio de Ambiente y t-test, t = 0.443, P = 0.660). Recursos Naturales, Nicaragua (MARENA). Stable isotopes Measurements of body elongation We compared carbon and nitrogen stable isotope signa- Overall body elongation in ﬁsh is a useful univariate mea- ture of individuals captured in the limnetic (n = 25) and sure to capture the main axis of body shape variation that the benthic zones (n = 25). Stable isotope analyses were distinguishes limnetic and benthic individuals (Kusche conducted on ethanol-preserved muscle tissue extracted et al. 2014; Recknagel et al. 2014). The relative extent of from dorsal musculature. Tissues were dried at 55°C until body elongation was measured in terms of the body there was no more change in mass (ca. 48 h). Muscle height index (BHI). The body height index is deﬁned as samples were subsequently ground up and those between the ratio of body height (distance between insertion of ca. 0.7 and 0.9 mg were placed in tin capsules for analy- the pelvic ﬁns and the most anterior point of the dorsal ses. Gas chromatography–combustion–isotope ratio mass ﬁn) to standard length (distance between the snout and spectrometry was performed at the Isotopes Laboratory of the most posterior point of the caudal peduncle). Higher the Limnological Institute of the University of Konstanz. values of the BHI denote thus more high-bodied (ben- Values of d C & were corrected for lipid content thic) individuals, whereas lower values of the BHI denote (Kiljunen et al. 2006). more elongated (limnetic) individuals. Body height and standard length were initially mea- Phenotypic plasticity sured in the ﬁeld for a large number of individuals cap- tured at the shore (n = 250) with calipers. For ﬁsh from To test whether the observed variation in body shapes the middle of the lake (n = 26; one individual was miss- could be due to phenotypic plasticity, we performed split- ing its pelvic ﬁns and its body height could thus not be brood design breeding experiments in the laboratory. Sin- measured reliably), the BHI was measured from individ- gle broods were divided into two groups and raised at ual photographs. To rule out that the two different mea- approximately equal densities in identical GFK tanks with suring modes (calipers, photographs) had any effect, we rounded corners (ca. 2000 L volume, 160 cm diameter) ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5345 Benthic Density 0 5 10 15 20 25 30 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. either with a constant circular water ﬂow or in control sodium chloride extraction and ethanol precipitation tanks without water ﬂow (Fig. S1, Supporting informa- (Bruford et al. 1998). A total of 52 samples (DNA extrac- tion). The ﬂow regimes in these tanks were designed to tion failed for two samples) from the middle of the lake simulate the increased swimming demand related to for- (n = 26) and the shore (n = 26) were successfully geno- aging in the limnetic habitat compared to the benthic typed at 13 microsatellites loci: Abur28, Abur45, Abur82, habitat. In addition to A. tolteca – the focal species ende- Abur151 (Sanetra et al. 2009); burt-kit (Salzburger et al. mic to Crater Lake As. Managua – we included Amphilo- 2007); M1M, M2, M7, M12 (Noack et al. 2000); TmoM7 phus citrinellus from the great lakes, which has acted as (Zardoya et al. 1996); UNH011, UNH012, UNH013 the source population of all Midas cichlids endemic to (Kellogg et al. 1995). All loci were PCR-ampliﬁed with the crater lakes and is thought to resemble the ancestral ﬂuorescent reverse or forward primers (HEX, FAM, and state of the whole species complex. Moreover, we NED dyes), and fragment lengths were determined with included the two limnetic species of Midas cichlids, the internal size standard Genescan-500 ROX (Applied Amphilophus zaliosus and Amphilophus sagittae, endemic Biosystems, Waltham, MA) on an ABI 3130 Automated to Crater Lakes Apoyo and Xiloa, respectively. Broods of Sequencer (Applied Biosystems, Waltham, MA) with the the four species stemmed from stock ﬁsh kept in the ani- GeneMapper v4.0 (Applied Biosystems, Waltham, MA) mal research facility of the University of Konstanz. At the software. Scoring errors, large allele dropout, and null start of the experiment, ﬁsh of A. sagittae were approxi- alleles were checked employing the program MICRO- mately 5 months (mean weight 2.9 g 1.2 g SD), A. za- CHECKER (Van Oosterhout et al. 2004). The most sup- liosus ca. 7 months (6.8 g 1.7 g SD), A. citrinellus ca. ported number of genetic clusters was determined using 8 months (8.6 g 1.8 g SD), and A. tolteca ca. the model-based Bayesian approach of STRUCTURE v2.3 12 months old (35.4 g 17.2 g SD). Samples sizes for (Pritchard et al. 2000). A burn-in period of 100,000 steps control and treatment groups of the four species were 41 followed by 500,000 Markov chain Monte Carlo iterations and 42 for A. sagittae, 16 and 31 for A. zaliosus, 24 and was sufﬁcient to ensure convergence. Five independent 24 for A. citrinellus, and 16 and 17 for A. tolteca, respec- runs each assuming between one and three genetic clus- tively. In control tanks, water inﬂow from the ﬁlter was ters (K = 1–3) were performed using the admixture directed from the top to the center of the tanks (pump: model (each individual draws some fraction of its genome Eheim type 2260, 65 W power), whereas in treatment from each of the K populations) with correlated allele tanks, the hose conducting the inﬂow was aligned with frequencies. Genetic differentiation by means of the one of the side walls to create a constant circular water ﬁxation index F was estimated with ARLEQUIN ST ﬂow. This current was further enhanced by two additional v.3.5.1.3 (Excofﬁer et al. 2005; Excofﬁer and Lischer water pumps aligned with the ﬁlter hose on the same side 2010) using 20,000 permutations to determine statistical of the tanks. In control tanks, the two additional pumps signiﬁcance. were attached to opposite sides of the tanks and directed at the center canceling out any water current. Water cur- RAD sequencing rent in treatment tanks was measured with an Aquadopp HR-Proﬁler (Nortek) and estimated to be on average ca. Population genomic data were generated following a dou- 14 cm/s across the water column and the whole tank ble-digest RAD sequencing approach (Peterson et al. (measured in 2-min intervals in each of the four corners). 2012) with minor modiﬁcations (Recknagel et al. 2013a). The experiment ran for 6 months, after which all ﬁsh The genomic library (pool of 50 individually barcoded were photographed. Standard length and body height samples) was single-end sequenced for 101 cycles using were taken from the photographs to calculate the BHI. an Illumina HiSeq 2000 platform at the genomics core Due to logistical reasons, ﬁsh from As. Managua could facility of TUFTS University (Boston, MA). After quality only be included 2 months after the start of the experi- inspection (there was no quality drop-off at the end of ment resulting in a total treatment time of 4 instead of the reads, and thus, no trimming was performed), indi- 6 months for this species. Fish were fed without any calo- vidually barcoded reads were demultiplexed using the ric restrictions and handled according to permit number STACKS v.1.29 software pipeline, ﬁltering out low-quality T-13/13. All statistical analyses were carried out in R (R reads (ﬂags: -w 0.1 -s 25 -r -c –q) (Catchen et al. 2011, Development Core Team 2014). 2013). Reads were then mapped to a reference genome assembly of an individual of A. citrinellus from great lake Nicaragua (Elmer et al. 2014) with BWA v.0.7.12 (Li and Microsatellites Durbin 2009). Custom bash scripts were used to remove Total genomic DNA was extracted from ca. 1 mm mus- reads mapping to several positions in the genome, con- cle tissue using a proteinase K digestion followed by taining soft-clipped positions, or showing a mapping 5346 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? quality of less than 25. Genotyping was carried out with than ﬁsh caught at the shore (littoral part of the benthic STACKS using a minimum of ﬁve reads for each locus, zone), the former were on average more elongated than an upper bound of 0.05 for the error rate and a 5% sig- the latter (Fig. 1C): The average BHI of ﬁsh from the niﬁcance level cutoff. The correction module (rxstacks) middle of the lake was 0.430 0.013 SD compared to an was used to correct individual genotype calls and remove average BHI of 0.438 0.014 SD for ﬁsh from the shore loci being confounded in more than 25% of individuals (Welch two sample t-test, t = 3.304, df = 31.08, or showing an excess of haplotypes within populations. P = 0.002). This difference is unlikely to be a result of Individual genotype calls with a log-likelihood of less than allometric effects (e.g., smaller ﬁsh are more elongated) as 10 were ﬁltered out and did not contribute to any sub- ﬁsh in the two groups did not signiﬁcantly differ in abso- sequent analyses. On average, 67,980 13,504 SD loci lute standard length (Welch two-sample t-test, t = 0.591, with a mean coverage of 11.7 1.7 SD reads were df = 35.363, P = 0.558) or in absolute body height obtained per individual. The last two positions of our (Welch two-sample t-test, t = 0.258, df = 33.264, reads exhibited an excess number of, probably artiﬁcial, P = 0.798). polymorphisms, and those loci with such polymorphisms were hence excluded (i.e., blacklisted). Furthermore, loci Stable isotopes deviating from Hardy–Weinberg equilibrium (HWE) (5% signiﬁcance level) or containing more than three SNPs To investigate whether ﬁsh captured in the limnetic zone were excluded from further analyses. HWE exact tests actually exploited this habitat compared to ﬁsh captured (Wigginton et al. 2005) were performed in Plink v.1.19- at the shore that presumably feed predominantly on ben- beta (Purcell et al. 2007). Only one SNP per RAD-tag thic prey, we performed stable isotope analyses on both locus was used for all analyses to reduce the effect of non- groups. Indeed, ﬁsh captured in the middle of the lake independence (linkage) among markers. had signiﬁcantly different carbon and nitrogen isotopic The demographic history was inferred performing coa- signatures compared to those collected at the shore lescent simulations in FASTSIMCOAL2 (Excofﬁer et al. (MANOVA Pillai 0.273, approximate F = 8.808, 2,47 2013). Loci presumably located in coding regions were P = 0.001). This suggests that the long-term diet of these identiﬁed via a BLASTN search against a compilation of two groups differs both in the source of carbon (univari- transcriptomic data from various species and tissues of ate ANOVA, F = 11.949, P = 0.001) as well as the 1,48 Midas cichlids (Elmer et al. 2010a; Henning et al. 2013; trophic level as indicated by the nitrogen isotope compo- Manousaki et al. 2013) and excluded. The minor joint sition (univariate ANOVA, F = 8.230, P = 0.006). Fish 1,48 site frequency spectrum (MSFS) was created as described collected in the limnetic zone of the lake were depleted in 13 13 in Kavembe et al. (2016). Brieﬂy, sample sizes were pro- C (mean d C = 33.34 1.28 SD) compared to those jected downwards to 25 individuals (50 alleles) using from the benthic zone (mean d C = 32.05 1.37 SD) dadi’s projection function (Gutenkunst et al. 2009). The ﬁnal two-dimensional SFS was built from an effective Limnetic 16.6 sequence length of 3.78 Mb and contained 10,075 poly- Benthic morphic sites. Inferred parameters were converted into 16.2 demographic units using a substitution rate of 15.8 7.5 9 10 per site and generation (Guo et al. 2013). For each demographic model, between 50 and 125 indepen- 15.4 dent runs were performed, consisting each of 50 rounds 15.0 of parameter estimation via the ECM algorithm with a length of 100,000–250,000 coalescent simulations each 14.6 (increasing by 5000 steps each round). Parametric boot- 14.2 strapping (n = 50) was performed to obtain 95% conﬁdence intervals (Excofﬁer et al. 2013). –36 –35 –34 –33 –32 –31 –30 –29 δ13C‰ Limnetic Benthic Results Origin of carbon Figure 2. Fish collected from the limnetic zone (blue) exhibit on Body shape differences average higher values of d N than ﬁsh from the benthic zone (red), In concordance with our prediction that ﬁsh collected in suggesting a slightly higher trophic position, and lower values of the open-water area of the lake (limnetic zone) would d C, implying that their source of carbon is of a more limnetic origin. exhibit on average a more fusiform/elongated body shape Large dots represent the mean SD. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5347 Trophic level Lower Higher δ15N‰ Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. (Fig. 2). Values of d C allow to draw inferences about water zone could lead to more elongated body shapes due the relative contribution of different primary producers as to an increased swimming demand (Johansson and a source of carbon of lacustrine animals because plank- Andersson 2009). Thus, in an attempt to test whether the tonic primary producers are depleted in C compared to differences in body shape elongation could be at least benthic primary producers (France 1995a,b; Hecky and partly explained by phenotypic plasticity, that is, whether Hesslein 1995). Fractionation between consumers and enduring swimming would lead to more elongated body their food is negligible, resulting in d C values being shapes, we performed experiments in a split-brood design. conserved throughout the food chain (DeNiro and Using single broods, in addition to the focal species of Epstein 1981). The difference in d C values between pri- this study, A. tolteca, we included A. citrinellus from the mary producers is thus consistently passed on to primary great lakes (the source population of the crater lakes) and and secondary consumers (Zanden and Rasmussen 1999; the two described limnetic species of Midas cichlids, Beaudoin et al. 2001; Post 2002). Our data suggest that A. zaliosus and A. sagittae, from Crater Lakes Apoyo and the diet of ﬁsh captured in the limnetic zone of the lake Xiloa, respectively. After four to six months of treatment had a higher proportion of carbon of a limnetic origin (four for A. tolteca and six for the other three species), than the diet of ﬁsh sampled in the benthic zone. we found that neither treatment nor the interaction of Furthermore, ﬁsh captured in the limnetic zone were species and treatment had a signiﬁcant effect on body 15 15 signiﬁcantly enriched in N (mean d N = 15.71 0.38 elongation (two-way ANOVA, F = 1.126, P = 0.290; SD) compared to those captured in the benthic zone F = 1.784, P = 0.152), while there were signiﬁcant differ- (mean d N = 15.35 0.50 SD) (Fig. 2). Consumers ences among the four species (F = 205.880, 15 16 generally become enriched in N compared to their food P < 2 9 10 ). The average body height indices for the source (DeNiro and Epstein 1981; Minagawa and Wada control and treatment groups after the experiment were 1984; Post 2002). Hence, d N values of consumer’s tis- 0.464 0.017 SD and 0.464 0.014 SD for A. tolteca, sues serve as an indicator of average trophic position 0.476 0.009 SD and 0.477 0.014 SD for A. citrinellus, (Zanden and Rasmussen 1999; Post 2002). The difference 0.424 0.012 SD and 0.421 0.009 SD for A. zaliosus, in the mean value of d N between ﬁsh captured in the and 0.436 0.011 SD and 0.442 0.011 SD for A. sagit- middle and the shore of the lake was relatively small with tae, respectively (Fig. 3). The relatively high BHI values of the former being on average ca. 0.5 & higher than the individuals from A. tolteca in our plasticity experiment latter. The difference between the two groups could be due to two factors. On the one hand, following the stan- dard interpretation of d N values, ﬁsh from the limnetic Control zone of the lake might occupy a slightly higher trophic Treatment position than those from the benthic zone as they con- sume on average more secondary consumers. Alterna- tively, the difference could be due to the fact that the primary producers supporting the food chains of benthic and limnetic ﬁsh differ in their d N values, as planktonic primary producers are slightly enriched in N compared to benthic primary producers (Zanden and Rasmussen 1999). Given the observed difference in d C values between the two groups – suggesting a different origin of carbon – this is also a plausible explanation. Either way, overall our data suggest a rather similar position in the trophic chain, yet evidence for clear differences in the diet emerges from the stable isotope data between ﬁsh from the limnetic zone of the lake and those from the benthic zone. Phenotypic plasticity Figure 3. Body height index of control (shaded in gray) and In concordance with our predictions, ﬁsh from the ben- treatment (white) groups of all four species included in the thic and limnetic zones differed signiﬁcantly in body phenotypic plasticity experiment. Only species differ in their body elongation and long-term diet. A possible explanation for height index (body elongation), whereas treatment did not have a this result is phenotypic plasticity. Foraging in the open- signiﬁcant effect. 5348 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. tolteca L. As. Managua A. citrinellus source lakes A. zaliosus L. Apoyo A. sagittae L. Xiloá Body height index (BHI) 0.42 0.44 0.46 0.48 0.50 A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? compared to wild-caught ﬁsh are probably due to the fact (900 m in diameter) and isolated crater lake. We simu- that we used single broods and the parents of the lated data in a coalescent framework according to 13 dif- A. tolteca brood were by chance relatively high-bodied. ferent demographic models (visualized in Fig. S2, The differences among species reﬂect the differences Supporting information) and evaluated the ﬁt of the sim- found in nature (Elmer et al. 2010b) and are probably ulations against the empirical data summarized in the not merely inﬂuenced by the slightly different age of the MSFS (Table S2, Supporting information) (Excofﬁer et al. four groups. Thus, plasticity induced by increased swim- 2013). As a source population, we used A. citrinellus from ming demands, as expected for ﬁsh inhabiting the lim- Lake Managua (n = 50, see Kautt et al. 2016). netic zone, is unlikely to contribute signiﬁcantly to the The most strongly supported model includes a “bottle- large degree of inter- and intraspeciﬁc variation in body growth” scenario (i.e., a population reduction followed by elongation observed in Midas cichlids. exponential growth) in the source population, exponential growth in the population of Crater Lake As. Managua after its colonization, and a secondary admixture event as Genetic differentiation well as continuous migration from the source lake into The differences in body shapes and long-term diet sug- the crater lake (Fig. 5). According to the maximum-likeli- gested that ﬁsh exhibit individual differences in their hood point estimates of the most supported model, the habitat preference, with some ﬁsh predominantly exploit- source population was reduced from 20,439 (95% CI: ing the limnetic open-water habitat, while others exploit 19,289–21,523) individuals to only 1556 (1047–2282) the benthic shore-associated habitat. We were interested individuals in a bottleneck event 2080 (1498–2885) gener- in whether there is any genetic divergence at neutral ations ago, recovering to a current population size of markers between the groups. Based on 13 microsatellite 260,429 (0–712,004) individuals. Note that conﬁdence markers, we did not ﬁnd any signs of genetic differentia- intervals for population sizes are relatively broad as even tion or population structuring: The overall F -value slight differences in the estimated exponential growth ST between the two groups was 0.005 (20,000 permutations, rates of the parametric bootstrap replicates can translate P = 0.720). Similarly, a STRUCTURE analysis did not to large deviations in the population sizes. Crater lake As. reveal more than one genetic cluster. Assuming a priori, Managua was colonized 797 (95% CI: 516–1284) genera- two clusters (K = 2) resulted in admixture proportions tions ago by only 32 (0–71) individuals. Since then, it has around 50% for all individuals (Fig. 4). 20 439 Past Colonization history 2080 generations Given the apparent absence of genetic structuring, but the 1556 fact that some of the necessary conditions for sympatric speciation are present in the population of Lake As. Man- 797 generations agua, we hypothesized that the exceedingly young age of 32 the crater lake population has not been sufﬁcient to build-up genetic differentiation at neutral markers. Aso- 507 generations 32.3% sosca Managua is geologically the youngest crater lake in Nicaragua known to house Midas cichlids (max. age ca. 1245 years; Pardo et al. 2008), yet the exact timing of col- 19 460 onization and the size of its founder population and its –5 8.95 x 10 260,429 current size had not been inferred before. To this end, we Present used RAD-seq data to infer the demographic history of Source Crater Lake the endemic Midas cichlid population of this very small (great Lake Managua) As. Managua 1.00 Figure 5. Schematic depiction of the most supported demographic 0.80 model and the associated maximum-likelihood point estimates of 0.60 demographic parameters. Numbers in the model refer to number of 0.40 diploid individuals, time points refer to number of generations, and 0.20 the migration rate refers to the probability for an allele to migrate 0.00 Limnetic Benthic into another deme (in forward time). Note that the model is not drawn in scale, but merely indicates differences in timing and Figure 4. Structure analysis based on 13 microsatellite markers population sizes. In addition, population growth was modeled to be assuming K = 2 clusters. exponential and not linear as shown here. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5349 Ancestry Coalescence Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. been growing to a current population size of 19,460 and, importantly, habitat. The match between morphol- (5336–43,039) individuals. In an admixture event that ogy and stable isotope signature is in agreement with our happened 507 (384–652) generations ago, the crater lake previous work (Kusche et al. 2014); more elongated ﬁsh 13 15 population received 32.3% (18.4–50.1%) of its gene pool were depleted in C and enriched in N compared to from the source population (great lake Managua). In less elongated ﬁsh. Yet, we did not recover the whole addition, since its colonization migration from the source range of the previously observed variation in d C and lake (L. Managua) into Crater Lake As. Managua has d N values (Kusche et al. 2014). Our goal was to com- happened with a probability of 8.95 9 10 pare the time-integrated diet of ﬁsh utilizing the limnetic 5 4 (5.40 9 10 –1.14 9 10 ), that is, ca. 9 of 100,000 alle- habitat versus those utilizing the benthic habitat, rather les, per generation. Evidence for an admixture event was than to describe the variation in diet associated with mor- unexpected, yet support for admixture events into Crater phology. Therefore, we did not choose the most extreme Lakes Apoyo and Xiloa was found recently (Kautt et al. high-bodied ﬁsh from the shore, but a randomly selected 2016). Whether the admixture event into Crater Lake As. subsample of all the ﬁsh captured at the shore. Thus, Managua is real and has facilitated divergence remains to most of the ﬁsh from the shore had an intermediate BHI be validated and tested. (subset mean = 0.437). Potentially, we could have recov- In conclusion, Crater Lake As. Managua seems to have ered a higher range by biasing our sample to include been colonized by a small founder population (only more ﬁsh with an extremely high BHI. However, this around 32 individuals) and very recently (ca. 800 genera- relation was already clearly established (Kusche et al. tions ago). Assuming a generation time of 1–2 years (Bar- 2014). It remained unclear though, if the observed pattern luenga and Meyer 2010), the colonization seems to have was associated with differences in habitat use per se.By happened shortly after the formation of the crater lake sampling Midas cichlids for the ﬁrst time from the open- itself (around 1245 years ago). water zone, we provide evidence that the depletion in C and enrichment in N of elongated ﬁsh compared to less elongated ﬁsh is most likely due to the differential use of Discussion the limnetic and benthic habitats within the lake. Surely, Based on the observed correspondence of habitat use with limnetic and benthic habitats in a lake are not completely morphology and diet, we can infer that Midas cichlids in discrete and treating them as two distinct habitats is cer- Crater Lake Asososca Managua specialize along the ben- tainly an oversimpliﬁcation, yet this distinction is a com- thic–limnetic axis: Fish from the limnetic zone of the lake mon and reasonable assumption (Bolnick 2011). are on average more elongated than ﬁsh from the benthic While we did not perform a mark–recapture study to zone and their diet reﬂects a more limnetic lifestyle and test habitat ﬁdelity per se (e.g., Bolnick et al. 2009), the vice versa. Our laboratory split-brood design experiments stable isotope signatures suggested that ﬁsh fed on average suggest that the differences in body elongation are proba- consistently, over long periods of time, to a different bly not to a large extent attributable to phenotypic plas- extent on diets that would be typical of the benthic and ticity, but are probably mostly genetically determined. limnetic habitats. Therefore, we conclude that individuals Yet, despite differential habitat use and a presumably preferentially, although probably not exclusively, use strong genetic basis, no population divergence is appar- either the benthic or the limnetic habitat. Whether the ent: Population genetic analyses show support for only match between eco-morphological traits is due to geneti- one genetic cluster. According to our demographic infer- cally based habitat preference or matching habitat choice ences, the population of Midas cichlids from Lake As. (sensu Edelaar et al. 2008) remains to be tested, but Managua is very young (ca. 800 generations ago) and has importantly, both processes increase the chance of habitat been colonized by a very small founder population (about isolation and thus speciation (Edelaar et al. 2008; Bolnick 32 individuals). This suggests that intraspeciﬁc competi- et al. 2009; Ravigne et al. 2009). tion for resources has probably only begun a few hundred generations ago, which could explain the apparent lack of Differences in body elongation are probably population divergence uncovered so far at the genomic mostly genetically determined rather than level. plastic A major question concerning the morphological differ- Differential habitat use of the benthic and ences between ﬁsh from the limnetic and benthic zones of limnetic habitats the lake is whether they could be due to phenotypic plas- Morphological measurements and stable isotope signa- ticity. If phenotypic plasticity was the main factor tures show a match of body elongation, long-term diet explaining the differences, population divergence and 5350 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? ultimately speciation may be unlikely (Edelaar et al. 2008; trigger this stimulus. Thus, we conclude that phenotypic Bolnick 2011; Thibert-Plante and Hendry 2011). In our plasticity does not seem to play a major role in explaining phenotypic plasticity experiment, a plastic response was the observed morphological differences. not induced in any of the four tested species: neither This conclusion is further supported by a study on the A. citrinellus from the great lakes (the source population genetic basis of the benthic–limnetic morphological diver- of all crater lakes) that resembles the ancestral state, nor gence of sympatric Midas cichlid species in Crater Lake the strongly elongated limnetic species of Midas species Apoyo that found that the divergent body shapes are A. zaliosus (Crater Lake Apoyo) and A. sagittae (Crater maintained in the laboratory and furthermore identiﬁed Lake Xiloa), nor the variable focal species of this study QTLs that explain some of the morphological differences A. tolteca exhibited a plastic response in this treatment (Franchini et al. 2014). The signiﬁcant differences in body (Fig. 3). This result stands in contrast to the fact that shape that we found among the four species in this study cichlids exhibit phenotypic plasticity in a number of dif- (Fig. 3) lend further support to the earlier ﬁndings that ferent traits (Meyer 1987; Kerschbaumer et al. 2011; the differences between species are maintained in captivity Machado-Schiafﬁno et al. 2014). We note, however, that even after one to two generations and are most likely to a we only investigated whether treatment had an effect on large extent genetically determined. the elongation of the main body axis as this was the main trait we were interested in. It is thus possible that an Lack of population differentiation possibly undetected plastic response was induced in our experi- due to very recent origin of intraspeciﬁc ment (e.g., in other aspects of morphology, behavior, or competition physiology). One potential caveat of this experiment is that only Population divergence along the benthic–limnetic axis is one brood per species was tested. Thus, the results cannot common in freshwater ﬁsh (Robinson and Wilson 1994) be readily generalized to infer the degree of plasticity that and is the basis of sympatric speciation in at least two may exist within the entire species. Phenotypic plasticity radiations of Crater Lake Midas cichlids (Barluenga et al. itself may be variable within and among species 2006; Elmer et al. 2014). Despite this and the fact that (Machado-Schiafﬁno et al. 2014). However, we believe ﬁsh in Lake As. Managua differently use the limnetic and that the fact that none of the four species showed any benthic habitats, our population genetic microsatellite pronounced plastic response supports our conclusion. It data suggest that ﬁsh from the middle of the lake are not is also possible that the treatment was applied too late in genetically differentiated from those captured at the shore. their ontogeny as earlier developmental stages might be Note that we are here referring to genetic differentiation more susceptible to exhibit a plastic response than later at neutral markers and not to relatively restricted highly stages or that the treatment was not strong enough. Yet, diverged regions that might differentiate the ecotypes the strength of our treatment was comparable to other (e.g., Malinsky et al. 2015). This lack of genetic structur- studies using a constant water ﬂow in an attempt to ing is in agreement with previous investigations based on induce a plastic response in ﬁsh (Peres-Neto and Magnan microsatellite markers that did not ﬁnd any evidence for 2004; Franssen et al. 2013). more than one genetic cluster in Crater Lake As. Mana- Alternatively, the treatment may have been biologically gua (Barluenga and Meyer 2010; Kusche et al. 2014). It is unrealistic in mimicking the natural conditions. Gener- important to note, though, that these studies differed ally, limnetic ﬁsh are thought to exhibit a more fusiform from our approach in that they exclusively used individu- body shape as an adaptation for increased swimming als captured at the shore. demand (Webb 1984, 1988), yet this is not driven by a Because samples for RAD-seq were collected exclusively constant water current, but rather related to the mode of from the shore, we could not use this data to test for foraging in the open water. Enclosure experiments in the genetic differentiation between ﬁsh from the limnetic and wild restricting individuals to the shore or the open-water benthic zones; this part of the project was started before habitat for foraging would be a more accurate way of we sampled ﬁsh for the ﬁrst time from the middle of the testing a role of phenotypic plasticity (Robinson and Par- lake. Nonetheless, as the two groups do not seem to be sons 2002), but such experiments have so far not been differentiated at neutral markers (yet), our demographic feasible in Nicaragua for logistical reasons. Nonetheless, inferences should not be affected by using only ﬁsh from whether driven by a different mode of foraging, predator the shore and our estimates of the colonization time and avoidance, or any other potentially unknown reason, the population size therefore unbiased. According to our relevant biological stimulus leading to a more elongated demographic inferences, Crater Lake As. Managua has body shape is most likely the resulting increased swim- been colonized only around 800 (516–1284, 95% CI) gen- ming demand. We think our treatment did effectively erations ago and by a very small founder population ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5351 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. consisting of only around 32 (0–71, 95% CI) individuals. allowed for enough time to build up genetic differentia- The estimate of the current population size of around tion throughout the genome. Alternatively, the population 19,500 individuals seems biologically reasonable; Midas may be stalled in its divergence, and the speciation pro- cichlids are by far the most abundant ﬁsh in L. As. Mana- cess may never be completed (Matessi et al. 2001; Nosil gua (pers. observation) and are today likely at carrying et al. 2009b). capacity in the relatively small crater lake (900 m in diameter; surface area of 0.74 km ). Yet, right after colo- Midas cichlids in Crater Lake As. Managua: a nization of the new underexploited environment of the case for incipient sympatric speciation? crater lake, the small founder population will almost cer- tainly not have been limited by resources. Whether population divergence will proceed and ulti- Generally, selection pressures in the beginning will mately lead to speciation depends foremost on the most likely have been directed at better adapting the strengths of disruptive selection and assortative mating founder population as a whole to the crater lake environ- (Gavrilets 2005; Bolnick 2011). Determining the strength ment in general: Crater lakes are very deep and their of selection acting on the benthic–limnetic divergence has water is usually very clear in contrast to the great lakes so far proofed not feasible in Midas cichlids due to the (the source) that are relatively shallow (mean depth of difﬁculty of performing experiments in the ﬁeld (e.g. 8–12 m) and turbid (Barlow 1976; Elmer et al. 2010b). In placing enclosures in the middle of a crater lake) or real- this regard, it is interesting that A. tolteca has evolved a istically resembling the open-water niche of a crater lake distinct morphology from the source population (Reck- (up to 200 m deep) in the laboratory. Yet, the fact that nagel et al. 2013b) in only 800 generations. It is possible sympatric speciation along the benthic–limnetic axis has that the founder effect that we have identiﬁed here has happened in two other crater lakes (Elmer et al. 2014) facilitated the rapid divergence of the crater lake popula- suggests that selection pressures along the benthic–lim- tion as a whole in allopatry (e.g., Kolbe et al. 2012), yet netic axis in Nicaraguan crater lakes have been at least in population divergence after the colonization of a new some cases sufﬁcient in driving sympatric speciation in environment due to selection can commence extremely Midas cichlids. However, the speciﬁcs matter and every fast (Lescak et al. 2015). lake environment and every population’s demographic In any case, only with time will the population have history will result in different conditions that may or may become large enough for intraspeciﬁc competition for not be conducive to sympatric speciation (Bolnick 2011; resources to elicit frequency-dependent disruptive selec- Martin 2013). tion. Resource limitation due to a high population den- While our results suggest that ﬁsh in Crater Lake As. sity is a necessary condition for stable disruptive selection Managua differentially use the benthic and limnetic habi- (Bolnick 2011). Hence, it seems possible that the pro- tats, it is currently unclear if this would readily translate cesses of disruptive selection and assortative mating have to reproductive isolation by habitat isolation like in phy- been at work in A. tolteca, but that genetic differentiation tophagous insects, for example (Rice 1987; Feder 1998; has not built up at neutral markers across the genome Via 1999). Midas cichlids form seasonally monogamous yet. Neutral genetic differentiation may not be expected pairs that breed at the shore (Barlow 1992), and hence, in the earliest stages of divergence (Elmer et al. 2010c; the spatial segregation breaks down during the time of Colborne et al. 2016). It seems likely that only few breeding (Baylis 1976). Yet, if pair formation happened in regions in the genome, that is, small genomic islands, that the respective habitat before pairs move to the shore to harbor the genetic basis for the observed differences in breed, differential habitat use would effectively result in morphology and trophic ecology are differentiated assortative mating by habitat (Gavrilets et al. 2007). between the benthic and limnetic ecomorphs in Crater Behavioral experiments have shown that sympatric ben- Lake As. Managua, as has been found recently between thic and limnetic species from Crater Lakes Apoyo and littoral and benthic ecomorphs of crater lake cichlids in Xiloa mate assortatively even under laboratory conditions Tanzania (Malinsky et al. 2015) or carrion and hooded (i.e., in the absence of different habitats), but that pair crows in Europe (Poelstra et al. 2014). Note that our formation happens before the pairs establish territories RAD-seq data set did not explicitly include individuals for breeding (Baylis 1976; Kautt et al., unpublished data). from the two habitats and we could thus not perform Hence, active mate preference seems to be a strong mech- outlier tests for signatures of selection. anism leading to assortative mating in these species, but Altogether, it seems that the population in Lake As. this does not negate the possibility that habitat isolation Managua is at the earliest stages of population divergence, is still contributing to reproductive isolation or has played and we propose that the very young age of the population a role in the initial divergence of limnetic and benthic and the even later onset of disruptive selection have not species in these two crater lakes. Different reproductive 5352 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? barriers often act in concert, and their contribution to and limnetic habitats. Whether these conditions are sufﬁ- overall reproductive isolation can change over time (Nosil cient for sympatric speciation to occur remains to be and Schluter 2011; Nosil 2012). Whether, and to what tested (e.g., Bolnick 2011; Martin 2013). Here, we did extent, assortative mating due to habitat isolation or mate not ﬁnd evidence for neutral genetic divergence in the preference exists in A. tolteca remains to be tested, but population. This could either mean that the strengths of our study shows that the conditions for assortative mat- disruptive selection and assortative mating are not strong ing by habitat isolation can exist even in a very small cra- enough and that the population is therefore stalled in its ter lake such as As. Managua. divergence (Matessi et al. 2001) or that we are dealing with a very recent population divergence (Elmer et al. 2010c; Colborne et al. 2016). While ecological experi- Conclusions ments are needed to test the former explanation, our Divergence along the benthic–limnetic axis is a common demographic inferences suggest that it is plausible that theme in the diversiﬁcation of freshwater ﬁshes (Schluter the apparent lack of genetic divergence is due to a lack of and McPhail 1992; Robinson and Wilson 1994; Hulsey time. In conclusion, our study shows how knowledge et al. 2013; Colborne et al. 2016), and in at least two cra- about the demographic history can inform on studies of ter lakes in Nicaragua, Midas cichlid ﬁsh have speciated speciation and that some of the necessary conditions for along this axis in sympatry (Elmer et al. 2014; Kautt et al. sympatric speciation do occur in nature and can occur 2016). Theory predicts that disruptive selection due to even in such a small environment as Crater Lake intraspeciﬁc competition for resources and assortative Asososca Managua. mating due to habitat isolation are necessary conditions for this process (Gavrilets et al. 2007). However, this pre- Acknowledgments diction has never been tested before in Midas cichlids. In complement to previous studies that have focused on the Sample collection permits were granted by MARENA, benthic and limnetic species in Crater Lakes Apoyo and Nicaragua. We are grateful to Empresa Nicaraguense de Xiloa (Barluenga et al. 2006; Kautt et al. 2012; Elmer Acueductos y Alcantarillados (ENACAL) and particularly et al. 2014), in this study we investigated a much earlier Ingeniero Martin Brenes for support of this study. The stage of the divergence continuum: population divergence coalescence analyses were partly performed on the compu- along the benthic–limnetic axis in a species of Midas tational resource bwUniCluster funded by the Ministry of cichlids, A. tolteca (Recknagel et al. 2013b), endemic to Science, Research and Arts and the Universities of the the extremely young and small Crater Lake As. Managua. State of Baden-Wurttemberg, € within the framework pro- More speciﬁcally, we studied whether some of the neces- gram bwHPC and partly on the High Performance Com- sary conditions for sympatric speciation due to intraspeci- puting (HPC) cluster of the University of Konstanz. We ﬁc competition for resources and habitat isolation are thank the Limnological Institute of the University of Kon- given in Lake As. Managua. stanz for help with the stable isotope analyses, in particu- In agreement with the prediction of a match between lar Elizabeth Yohannes, and Hilmar Hofmann for help phenotype and habitat, we found that individuals caught with measuring the water current in the phenotypic plas- in the limnetic habitat are more elongated than ﬁsh col- ticity experiment. Moreover, we want to thank two anony- lected in the benthic habitat. Stable isotope analyses – by mous reviewers whose comments helped to improve this integrating diet over longer time spans – further con- manuscript. A. F. Kautt was supported by the Landes- ﬁrmed that ﬁsh from the limnetic habitat also exhibit a graduiertenforderung € (LGFG) of the State of Baden- more limnetic lifestyle. Thus, we conclude that individu- Wurttemberg € and the International Max Planck Research als differentially use the two habitats. Together with pre- School (IMPRS) for Organismal Biology. G. Machado- vious evidence (Franchini et al. 2014), our experiments Schiafﬁno was supported by the Alexander von Humboldt conducted in the laboratory suggest that the differences Foundation and a grant from the Deutsche Forschungsge- we found in the most relevant ecological trait, body meinschaft (MA 6144/1-1). J. Torres-Dowdall was sup- shape elongation, are unlikely to be due to phenotypic ported by an EU FP7 Marie Curie Zukunftskolleg plasticity and are probably strongly genetically deter- Incoming Fellowship Program of the University of Kon- mined. Altogether, our data therefore support the notion stanz (grant number 291784) and a grant from the that some of the necessary conditions for sympatric spe- Deutsche Forschungsgemeinschaft (TO 914/2-1). The ciation are present in Midas cichlids in Crater Lake As. study was funded by support of the University of Kon- Managua: Individuals vary in genetically determined eco- stanz and a European Research Council advanced grant morphological traits and differentially use the benthic (ERC “GenAdap” 293700) to A. Meyer. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5353 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. divergence in a species characterized by a trophic Data Accessibility polymorphism. J. Evol. Biol. 29:633–644. Morphometric measurements, stable isotope data, and Comeault, A. A., S. M. Flaxman, R. Riesch, E. Curran, V. microsatellite genotypes are provided in Table S1, Sup- Soria-Carrasco, Z. Gompert, et al. 2015. Selection on a porting information. The variant call format (VCF) of the genetic polymorphism counteracts ecological speciation in a RAD-seq data is available as Data set S1, Supporting stick insect. Curr. Biol. 25:1975–1981. information. DeNiro, M. J., and S. Epstein. 1981. Inﬂuence of diet on the distribution of nitrogen isotopes in animals. Geochim. Cosmochim. Acta 45:341–351. Conﬂict of Interest Edelaar, P., A. M. Siepielski, and J. Clobert. 2008. Matching habitat choice causes directed gene ﬂow: a neglected None declared. dimension in evolution and ecology. Evolution 62:2462–2472. Elmer, K. R., S. Fan, H. M. GunterJ. C. Jones, S. Boekhoff, S. References Kuraku, et al. 2010a. Rapid evolution and selection inferred Barlow, G. W. 1976. The Midas cichlid in Nicaragua. Pp. 333– from the transcriptomes of sympatric Crater lake cichlid 358 in T. B. Thorson, ed. Investigations of the ichthyology ﬁshes. Mol. Ecol. 19:197–211. of Nicaraguan lakes. Univ. of Nebraska Press, Lincoln. Elmer, K. R., H. Kusche, T. K. Lehtonen, and A. Meyer. Barlow, G. W. 1992. Is mating different in monogamous species? 2010b. Local variation and parallel evolution: morphological The Midas cichlid ﬁsh as a case study. Am. Zool. 32:91–99. and genetic diversity across a species complex of neotropical Barluenga, M., and A. Meyer. 2010. Phylogeography, crater lake cichlid ﬁshes. Philos. Trans. R. Soc. Lond. B Biol. colonization and population history of the Midas cichlid Sci. 365:1763–1782. species complex (Amphilophus spp.) in the Nicaraguan Elmer, K. R., T. K. Lehtonen, A. F. Kautt, C. Harrod, and A. crater lakes. BMC Evol. Biol. 10:326. Meyer. 2010c. Rapid sympatric ecological differentiation of Barluenga, M., K. N. Stolting, W. Salzburger, M. Muschick, crater lake cichlid ﬁshes within historic times. BMC Biol. and A. Meyer. 2006. Sympatric speciation in Nicaraguan 8:60. Crater lake cichlid ﬁsh. Nature 439:719–723. Elmer, K. R., S. H. Fan, H. Kusche, M. L. Spreitzer, A. F. Baylis, J. R. 1976. Quantitative study of long-term courtship: Kautt, P. Franchini, et al. 2014. Parallel evolution of 1. Ethological Isolation between sympatric populations of Nicaraguan crater lake cichlid ﬁshes via non-parallel routes. the Midas cichlid, Cichlasoma citrinellum, and the Arrow Nat. Commun. 5:5168. Cichlid, C. zaliosum. Behaviour 59:59–69. Excofﬁer, L., and H. E. L. Lischer. 2010. Arlequin suite ver Beaudoin, C. P., E. E. Prepas, W. M. Tonn, L. I. Wassenaar, 3.5: a new series of programs to perform population and B. G. Kotak. 2001. A stable carbon and nitrogen isotope genetics analyses under Linux and Windows. Mol. Ecol. study of lake food webs in Canada’s Boreal Plain. Freshw. Resour. 10:564–567. Biol. 46:465–477. Excofﬁer, L., G. Laval, and S. Schneider. 2005. Bolnick, D. I. 2011. Sympatric speciation in threepine Arlequin (version 3.0): an integrated software package for stickleback: why not? Int. J. Ecol. 2011:15. population genetics data analysis. Evol. Bioinform. Online Bolnick, D. I., L. K. Snowberg, C. Patenia, W. E. Stutz, T. 1:47–50. Ingram, and O. L. Lau. 2009. Phenotype-dependent native Excofﬁer, L., I. Dupanloup, E. Huerta-Sanchez, V. C. Sousa, habitat preference facilitates divergence between Parapatric and M. Foll. 2013. Robust demographic inference from Lake and stream stickleback. Evolution 63:2004–2016. genomic and SNP data. PLoS Genet. 9:e1003905. Bruford, M. W., O. Hanotte, J. F. Y. Brookﬁeld, and T. Burke. Feder, J. L. 1998. The apple maggot ﬂy, Rhagoletis pomonella. 1998. Multilocus and single locus DNA ﬁngerprinting. Pp. Pp. 130–144 in D. J. Howard and S. H. Berlocher, eds. 287–336 in A. R. Hoelzel, ed. Molecular genetics analysis of Endless forms: species and speciation. Oxford Univ. Press, populations: a practical approach. Oxford Univ. Press, Oxford, U.K. Oxford, U.K. Feder, J. L., S. P. Egan, and P. Nosil. 2012. The genomics of Catchen, J. M., A. Amores, P. A. Hohenlohe, W. Cresko, and speciation-with-gene-ﬂow. Trends Genet. 28:342–350. J. H. Postlethwait. 2011. Stacks: building and genotyping Felsenstein, J. 1981. Skepticism towards santa rosalia, or why loci de novo from short-read sequences. G3: Genes - are there so few kinds of animals. Evolution 35:124–138. Genomes - Genetics 1:171–182. France, R. L. 1995a. Differentiation between littoral and Catchen, J., P. A. Hohenlohe, S. Bassham, A. Amores, and W. pelagic food webs in lakes using stable carbon isotopes. A. Cresko. 2013. Stacks: an analysis tool set for population Limnol. Oceanogr. 40:1310–1313. genomics. Mol. Ecol. 22:3124–3140. France, R. L. 1995b. C-13 enrichment in benthic compared to Colborne, S. F., S. R. Garner, F. K. Longstaffe, and B. D. Neff. planktonic algae – foodweb implications. Mar. Ecol. Prog. 2016. Assortative mating but no evidence of genetic Ser. 124:307–312. 5354 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Franchini, P., C. Fruciano, M. L. Spreitzer, J. C. Jones, K. R. Kerschbaumer, M., L. Postl, M. Koch, T. Wiedl, and C. Elmer, F. Henning, et al. 2014. Genomic architecture of Sturmbauer. 2011. Morphological distinctness despite large- ecologically divergent body shape in a pair of sympatric scale phenotypic plasticity-analysis of wild and pond-bred crater lake cichlid ﬁshes. Mol. Ecol. 23:1828–1845. juveniles of allopatric populations of Tropheus moorii. Franssen, N. R., L. K. Stewart, and J. F. Schaefer. 2013. Naturwissenschaften 98:125–134. Morphological divergence and ﬂow-induced phenotypic Kiljunen, M., J. Grey, T. Sinisalo, C. Harrod, H. Immonen, plasticity in a native ﬁsh from anthropogenically altered and R. I. Jones. 2006. A revised model for lipid-normalizing stream habitats. Ecol. Evol. 3:4648–4657. delta C-13 values from aquatic organisms, with implications Gavrilets, S. 2003. Perspective: models of speciation: what have for isotope mixing models. J. Appl. Ecol. 43:1213–1222. we learned in 40 years? Evolution 57:2197–2215. Kolbe, J. J., M. Leal, T. W. Schoener, D. A. Spiller, and J. B. Gavrilets, S. 2004. Fitness landscapes and the origin of species. Losos. 2012. Founder effects persist despite adaptive Princeton Univ. Press, Princeton, NJ. differentiation: a ﬁeld experiment with lizards. Science Gavrilets, S. 2005. “Adaptive speciation” – it is not that easy: a 335:1086–1089. reply to Doebeli et al.. Evolution 59:696–699. Kusche, H., H. Recknagel, K. R. Elmer, and A. Meyer. 2014. Gavrilets, S., A. Vose, M. Barluenga, W. Salzburger, and A. Crater lake cichlids individually specialize along the benthic- Meyer. 2007. Case studies and mathematical models of limnetic axis. Ecol. Evol. 4:1127–1139. ecological speciation. 1. Cichlids in a crater lake. Mol. Ecol. Lescak, E. A., S. L. Bassham, J. Catchen, O. Gelmond, M. L. 16:2893–2909. Sherbick, F. A. von Hippel, et al. 2015. Evolution of Guo, B. C., F. J. J. Chain, E. Bornberg-Bauer, E. H. Leder, and stickleback in 50 years on earthquake-uplifted islands. Proc. J. Merila.2013. Genomic divergence between nine- and Natl Acad. Sci. USA 112:E7204–E7212. three-spined sticklebacks. BMC Genom. 14:756. Li, H., and R. Durbin. 2009. Fast and accurate short read Gutenkunst, R. N., R. D. Hernandez, S. H. Williamson, and C. alignment with Burrows-Wheeler transform. Bioinformatics D. Bustamante. 2009. Inferring the joint demographic 25:1754–1760. history of multiple populations from multidimensional SNP Machado-Schiafﬁno, G., F. Henning, and A. Meyer. 2014. frequency data. PLoS Genet. 5:e1000695. Species-speciﬁc differences in adaptive phenotypic plasticity Hecky, R. E., and R. H. Hesslein. 1995. Contributions of in an ecologically relevant trophic trait: hypertrophic lips in benthic algae to lake food webs as revealed by stable isotope Midas cichlid ﬁshes. Evolution 68:2086–2091. analysis. J. N. Am. Benthol. Soc. 14:631–653. Machado-Schiafﬁno, G., A. F. Kautt, H. Kusche, and A. Meyer. Henning, F., J. C. Jones, P. Franchini, and A. Meyer. 2013. 2015. Parallel evolution in Ugandan crater lakes: repeated Transcriptomics of morphological color change in evolution of limnetic body shapes in haplochromine cichlid polychromatic Midas cichlids. BMC Genom. 14:171. ﬁsh. BMC Evol. Biol. 15:9. Hulsey, C. D., R. J. Roberts, Y. H. E. Loh, M. F. Rupp, and J. Malinsky, M., R. J. Challis, A. M. Tyers, S. Schiffels, Y. Terai, T. Streelman. 2013. Lake Malawi cichlid evolution along a B. P. Ngatunga, et al. 2015. Genomic islands of speciation benthic/limnetic axis. Ecol. Evol. 3:2262–2272. separate cichlid ecomorphs in an East African crater lake. Johansson, F., and J. Andersson. 2009. Scared ﬁsh get lazy, and Science 350:1493–1498. lazy ﬁsh get fat. J. Anim. Ecol. 78:772–777. Manousaki, T., P. M. Hull, H. Kusche, G. Machado-Schiafﬁno, Kautt, A. F., K. R. Elmer, and A. Meyer. 2012. Genomic signatures P. Franchini, C. Harrod, et al. 2013. Parsing parallel of divergent selection and speciation patterns in a natural evolution: ecological divergence and differential experiment’, the young parallel radiations of Nicaraguan crater gene expression in the adaptive radiations of thick- lake cichlid ﬁshes. Mol. Ecol. 21:4770–4786. lipped Midas cichlid ﬁshes from Nicaragua. Mol. Ecol. Kautt, A. F., G. Machado-Schiafﬁno, and A. Meyer. 2016. 22:650–669. Multispecies outcomes of sympatric speciation after Martin, C. H. 2013. Strong assortative mating by diet, color, admixture with the source population in two radiations of size, and morphology but limited progress toward sympatric nicaraguan crater lake cichlids. PLoS Genet. 12:e1006157. speciation in a classic example: Cameroon crater lake doi:10.1371/journal.pgen.1006157 cichlids. Evolution 67:2114–2123. Kavembe, G. D., A. F. Kautt, G. Machado-Schiafﬁno, and A. Martin, R. A., and D. W. Pfennig. 2009. Disruptive Meyer. 2016. Eco-morphological differentiation in Lake selection in natural populations: the roles of ecological Magadi tilapia, an extremophile cichlid ﬁsh living in hot, specialization and resource competition. Am. Nat. alkaline and hypersaline lakes in East Africa. Mol. Ecol. 174:268–281. 25:1610–1625. Martin, C. H., J. S. Cutler, J. P. Friel, C. D. Touokong, G. Kellogg, K. A., J. A. Markert, J. R. Stauffer, and T. D. Kocher. Coop, and P. C. Wainwright. 2015. Complex histories of 1995. Microsatellite variation demonstrates multiple repeated gene ﬂow in Cameroon crater lake cichlids cast paternity in Lekking Cichlid ﬁshes from Lake Malawi, doubt on one of the clearest examples of sympatric Africa. Proc. R. Soc. B 260:79–84. speciation. Evolution 69:1406–1422. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5355 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. Matessi, C., A. Gimelfarb, and S. Gavrilets. 2001. Long-term whole-genome association and population-based linkage buildup of reproductive isolation promoted by disruptive analyses. Am. J. Hum. Genet. 81:559–575. selection: how far does it go? Selection 2:41–64. R Development Core Team. 2014. R: a Language and Meyer, A. 1987. Phenotypic plasticity and heterochrony in Environment for Statistical Computing. Available at http:// Cichlasoma managuense (Pisces, Cichlidae) and their www.R-project.org. (accessed 20 March 2016). implications for speciation in cichlid ﬁshes. Evolution Ravigne, V., U. Dieckmann, and I. Olivieri. 2009. Live where 41:1357–1369. you thrive: joint evolution of habitat choice and local Meyer, A., and A. F. Kautt. 2014. Sympatric speciation. Oxford adaptation facilitates specialization and promotes diversity. bibliographies in evolutionary biology. DOI:10.1093/obo/ Am. Nat. 174:E141–E169. 9780199941728-0053. Recknagel, H., K. R. Elmer, and A. Meyer. 2013a. A hybrid Minagawa, M., and E. Wada. 1984. Stepwise enrichment of genetic linkage map of two ecologically and morphologically N-15 along food-chains – further evidence and the relation divergent Midas cichlid ﬁshes (Amphilophus spp.) obtained between delta-N-15 and animal age. Geochim. Cosmochim. by massively parallel DNA sequencing (ddRADSeq). G3: Genes - Genomes - Genetics 3:65–74. Acta 48:1135–1140. Noack, K., A. B. Wilson, and A. Meyer. 2000. Broad Recknagel, H., H. Kusche, K. R. Elmer, and A. Meyer. 2013b. taxonomic applicability of microsatellites developed for the Two new endemic species in the Midas cichlid species complex from Nicaraguan crater lakes: Amphilophus tolteca highly polymorphic neotropical cichlid, Amphilophus citrinellum. Anim. Genet. 31:151–152. and Amphilophus viridis (Perciformes, Cichlidae). Aqua Nosil, P. 2012. Ecological speciation. Oxford Univ. Press, New 19:207–224. York. Recknagel, H., K. R. Elmer, and A. Meyer. 2014. Crater lake habitat predicts morphological diversity in adaptive Nosil, P., and D. Schluter. 2011. The genes underlying the process of speciation. Trends Ecol. Evol. 26:160–167. radiations of cichlid Fishes. Evolution 68:2145–2155. Nosil, P., D. J. Funk, and D. Ortiz-Barrientos. 2009a. Rice, W. R. 1987. Speciation via habitat specialization: the evolution of reproductive isolation as a correlated character. Divergent selection and heterogeneous genomic divergence. Mol. Ecol. 18:375–402. Evol. Ecol. 1:301–314. Nosil, P., L. J. Harmon, and O. Seehausen. 2009b. Ecological Rieseberg, L. H. 2001. Chromosomal rearrangements and explanations for (incomplete) speciation. Trends Ecol. Evol. speciation. Trends Ecol. Evol. 16:351–358. 24:145–156. Robinson, B. W., and K. J. Parsons. 2002. Changing times, Pardo, N., D. R. Avellan, J. L. Macias, T. Scolamacchia, and spaces, and faces: tests and implications of adaptive D. Rodriguez. 2008. The ~1245 yr BP Asososca maar: new morphological plasticity in the ﬁshes of northern postglacial lakes. Can. J. Fish Aquat. Sci. 59:1819–1833. advances on recent volcanic stratigraphy of Managua (Nicaragua) and hazard implications. J. Volcanol. Geoth. Robinson, B. W., and D. S. Wilson. 1994. Character release Res. 176: and displacement in ﬁshes – a neglected literature. Am. Nat. 493–512. 144:596–627. Peres-Neto, P. R., and P. Magnan. 2004. The inﬂuence of Salzburger, W., I. Braasch, and A. Meyer. 2007. Adaptive swimming demand on phenotypic plasticity and sequence evolution in a color gene involved in the morphological integration: a comparison of two formation of the characteristic egg-dummies of male haplochromine cichlid ﬁshes. BMC Biol. 5:51. polymorphic charr species. Oecologia 140:36–45. Peterson, B. K., J. N. Weber, E. H. Kay, H. S. Fisher, and H. Sanetra, M., F. Henning, S. Fukamachi, and A. Meyer. 2009. E. Hoekstra. 2012. Double digest RADseq: an inexpensive A microsatellite-based genetic linkage map of the cichlid method for de novo SNP discovery and genotyping in ﬁsh, Astatotilapia burtoni (Teleostei): a comparison of model and non-model species. PLoS One 7:e37135. genomic architectures among rapidly speciating cichlids. Poelstra, J. W., N. Vijay, C. M. Bossu, H. Lantz, B. Ryll, I. Genetics 182:387–397. Mueller, et al. 2014. The genomic landscape underlying Schluter, D. 1996. Ecological speciation in postglacial ﬁshes. phenotypic integrity in the face of gene ﬂow in crows. Philos. Trans. R. Soc. Lond. B Biol. Sci. 351:807–814. Science 344:1410–1414. Schluter, D., and J. D. McPhail. 1992. Ecological character Post, D. M. 2002. Using stable isotopes to estimate trophic displacement and speciation in sticklebacks. Am. Nat. position: models, methods, and assumptions. Ecology 140:85–108. 83:703–718. Servedio, M. R., G. S. Van Doorn, M. Kopp, A. M. Frame, Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference and P. Nosil. 2011. Magic traits in speciation: ‘magic’ but of population structure using multilocus genotype data. not rare? Trends Ecol. Evol. 26:389–397. Genetics 155:945–959. Smadja, C. M., and R. K. Butlin. 2011. A framework for Purcell, S., B. Neale, K. Todd-Brown, L. Thomas, M. A. R. comparing processes of speciation in the presence of gene Ferreira, D. Bender, et al. 2007. PLINK: a tool set for ﬂow. Mol. Ecol. 20:5123–5140. 5356 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Thibert-Plante, X., and A. P. Hendry. 2011. The consequences Wu, C. I. 2001. The genic view of the process of speciation. J. of phenotypic plasticity for ecological speciation. J. Evol. Evol. Biol. 14:851–865. Biol. 24:326–342. Zanden, M. J. V., and J. B. Rasmussen. 1999. Primary Trickett, A. J., and R. K. Butlin. 1994. Recombination consumer delta C-13 and delta N-15 and the trophic suppressors and the evolution of new species. Heredity position of aquatic consumers. Ecology 80:1395–1404. 73:339–345. Zardoya, R., D. M. Vollmer, Craddock, J. T. Streelman, S. Udovic, D. 1980. Frequency-dependent selection, disruptive Karl, and A. Meyer. 1996. Evolutionary conservation of selection, and the evolution of reproductive isolation. Am. microsatellite ﬂanking regions and their use in resolving the Nat. 116:621–641. phylogeny of cichlid ﬁshes (Pisces: Perciformes). Proc. R. Van Oosterhout, C., W. F. Hutchinson, D. P. M. Willis, and Soc. B 263:1589–1598. P. Shipley. 2004. Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol. Supporting Information Ecol. Notes 4:535–538. Via, S. 1999. Reproductive isolation between sympatric races Additional Supporting Information may be found online of pea aphids. I. Gene ﬂow restriction and habitat choice. in the supporting information tab for this article: Evolution 53:1446–1457. Figure S1. Experimental setup of the phenotypic plasticity Via, S. 2009. Natural selection in action during speciation. experiment. Proc. Natl Acad. Sci. USA 106:9939–9946. Figure S2. Schematic illustration of all 13 tested demo- Webb, P. W. 1984. Body form, locomotion and foraging in graphic models. aquatic vertebrates. Am. Zool. 24:107–120. Table S1. Morphometric measurements, stable isotope Webb, P. W. 1988. Simple physical principles and vertebrate data, and microsatellite genotypes. aquatic locomotion. Am. Zool. 28:709–725. Table S2. Support for tested demographic models. Wigginton, J. E., D. J. Cutler, and G. R. Abecasis. 2005. A note Data set S1. RADseq genotype data in variant call for- on exact tests of Hardy-Weinberg equilibrium. Am. J. Hum. mat. Genet. 76:887–893. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5357 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecology and Evolution Wiley http://www.deepdyve.com/lp/wiley/incipient-sympatric-speciation-in-midas-cichlid-fish-from-the-youngest-HHcas4SD84

Loading next page...

References (112)

(1995)
Differentiation between littoral and pelagic food webs in lakes using stable carbon isotopes
P. Nosil, D. Funk, D. Ortiz-Barrientos (2009)
Divergent selection and heterogeneous genomic divergence
Molecular Ecology, 18
W. Salzburger, I. Braasch, A. Meyer (2007)
Adaptive sequence evolution in a color gene involved in the formation of the characteristic egg-dummies of male haplochromine cichlid fishes
BMC Biology, 5
S. Purcell, B. Neale, K. Todd-Brown, L. Thomas, M. Ferreira, David Bender, J. Maller, P. Sklar, P. Bakker, M. Daly, P. Sham (2007)
PLINK: a tool set for whole-genome association and population-based linkage analyses.
American journal of human genetics, 81 3
D. Schluter, J. Mcphail (1992)
Ecological Character Displacement and Speciation in Sticklebacks
The American Naturalist, 140
T. Burke, M. Bruford, O. Hanotte, J. Brookfield (1998)
Multilocus and single-locus DNA fingerprinting
K. Elmer, Topi Lehtonen, Andreas Kautt, C. Harrod, A. Meyer (2012)
Rapid sympatric ecological differentiation of crater lake cichlid fishes within historic times
BMC Biology, 10
M. Malinsky, Richard Challis, Alexandra Tyers, S. Schiffels, Y. Terai, B. Ngatunga, E. Miska, R. Durbin, M. Genner, G. Turner (2015)
Genomic islands of speciation separate cichlid ecomorphs in an East African crater lake
Science, 350
Natthakan Boongoen, Tossapon Garrett (2004)
BIOINFORMATICS ORIGINAL PAPER
C. Martin, J. Cutler, J. Friel, Cyrille Touokong, G. Coop, P. Wainwright (2015)
Complex histories of repeated gene flow in Cameroon crater lake cichlids cast doubt on one of the clearest examples of sympatric speciation
Evolution, 69
G. Kavembe, Andreas Kautt, G. Machado-Schiaffino, A. Meyer (2016)
Eco‐morphological differentiation in Lake Magadi tilapia, an extremophile cichlid fish living in hot, alkaline and hypersaline lakes in East Africa
Molecular Ecology, 25
Per Capita, E. Dawson, Myfan Jordan (1995)
About the authors
Machine Vision and Applications, 1
H. Li, R. Durbin (2009)
Fast and accurate short read alignment with Burrows‐Wheeler transform
, 25
M. Zanden, J. Rasmussen (1999)
PRIMARY CONSUMER δ13C AND δ15N AND THE TROPHIC POSITION OF AQUATIC CONSUMERS
Ecology, 80
M. Servedio, G. Doorn, M. Kopp, Alicia Frame, P. Nosil (2011)
Magic traits in speciation: 'magic' but not rare?
Trends in ecology & evolution, 26 8
G. Machado-Schiaffino, Andreas Kautt, Henrik Kusche, A. Meyer (2015)
Parallel evolution in Ugandan crater lakes: repeated evolution of limnetic body shapes in haplochromine cichlid fish
BMC Evolutionary Biology, 15
P. Nosil, L. Harmon, O. Seehausen (2009)
Ecological explanations for (incomplete) speciation.
Trends in ecology & evolution, 24 3
G. Barlow (1992)
Is Mating Different in Monogamous Species? The Midas Cichlid Fish as a Case Study
Integrative and Comparative Biology, 32
M. Kerschbaumer, Lisbeth Postl, Martin Koch, T. Wiedl, C. Sturmbauer (2010)
Morphological distinctness despite large-scale phenotypic plasticity—analysis of wild and pond-bred juveniles of allopatric populations of Tropheus moorii
Die Naturwissenschaften, 98
L. Excoffier, H. Lischer (2010)
Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows
Molecular Ecology Resources, 10
J. Feder, S. Egan, P. Nosil (2012)
The genomics of speciation-with-gene-flow.
Trends in genetics : TIG, 28 7
RL France (2006)
Carbon-13 enrichment in benthic compared to planktonic algae : foodweb implications
Baocheng Guo, Frédéric Chain, E. Bornberg-Bauer, E. Leder, J. Merilä (2013)
Genomic divergence between nine- and three-spined sticklebacks
BMC Genomics, 14
G. Machado-Schiaffino, F. Henning, A. Meyer (2014)
SPECIES‐SPECIFIC DIFFERENCES IN ADAPTIVE PHENOTYPIC PLASTICITY IN AN ECOLOGICALLY RELEVANT TROPHIC TRAIT: HYPERTROPHIC LIPS IN MIDAS CICHLID FISHES
Evolution, 68
P. Nosil, D. Schluter
The Genes Underlying the Process of Speciation
B. Robinson, D. Wilson (1994)
Character Release and Displacement in Fishes: A Neglected Literature
The American Naturalist, 144
K. Elmer, Shaohua Fan, Henrik Kusche, Maria Spreitzer, Andreas Kautt, P. Franchini, A. Meyer (2014)
Parallel evolution of Nicaraguan crater lake cichlid fishes via non-parallel routes
Nature Communications, 5
J. Baylis (1976)
A Quantitative Study of Long-Term Courtship: I. Ethological Isolation Between Sympatric Populations of the Midas Cichlid, Cichlasoma Citrinellum, and the Arrow Cichlid, C. Zaliosum
Behaviour, 59
A. Meyer (1987)
PHENOTYPIC PLASTICITY AND HETEROCHRONY IN CICHLASOMA MANAGUENSE (PISCES, CICHLIDAE) AND THEIR IMPLICATIONS FOR SPECIATION IN CICHLID FISHES
Evolution, 41
F. Henning, Julia Jones, P. Franchini, A. Meyer (2013)
Transcriptomics of morphological color change in polychromatic Midas cichlids
BMC Genomics, 14
N. Patrik (2012)
What is ecological speciation
N. Pardo, D. Avellán, J. Macías, T. Scolamacchia, D. Rodríguez (2008)
The ∼1245 yr BP Asososca maar: New advances on recent volcanic stratigraphy of Managua (Nicaragua) and hazard implications
Journal of Volcanology and Geothermal Research, 176
D. Bolnick, L. Snowberg, Claire Patenia, William Stutz, T. Ingram, O. Lau (2009)
Phenotype-Dependent Native Habitat Preference Facilitates Divergence between Parapatric Lake and Stream Stickleback
, 63
M. Sanetra, F. Henning, S. Fukamachi, A. Meyer (2009)
A Microsatellite-Based Genetic Linkage Map of the Cichlid Fish, Astatotilapia burtoni (Teleostei): A Comparison of Genomic Architectures Among Rapidly Speciating Cichlids
Genetics, 182
J. Kolbe, M. Leal, T. Schoener, David Spiller, J. Losos (2012)
Founder Effects Persist Despite Adaptive Differentiation: A Field Experiment with Lizards
Science, 335
F. Johansson, J. Andersson (2009)
Scared fish get lazy, and lazy fish get fat.
The Journal of animal ecology, 78 4
Henrik Kusche, H. Recknagel, K. Elmer, A. Meyer (2014)
Crater lake cichlids individually specialize along the benthic–limnetic axis
Ecology and Evolution, 4
R. Hecky, R. Hesslein (1995)
Contributions of Benthic Algae to Lake Food Webs as Revealed by Stable Isotope Analysis
Journal of the North American Benthological Society, 14
Adam Trickett, R. Butlin (1994)
Recombination suppressors and the evolution of new species
Heredity, 73
G. W. Barlow (1976)
Investigations of the ichthyology of Nicaraguan lakes
S. Gavrilets, A. Vose, M. Barluenga, W. Salzburger, A. Meyer (2007)
Case studies and mathematical models of ecological speciation. 1. Cichlids in a crater lake
Molecular Ecology, 16
K. Kellogg, Jeffrey Markert, J. Stauffer, T. Kocher (1995)
Microsatellite variation demonstrates multiple paternity in lekking cichlid fishes from Lake Malawi, Africa
Proceedings of the Royal Society of London. Series B: Biological Sciences, 260
G. Barlow (1976)
The Midas Cichlid In Nicaragua
P. Webb (1988)
Simple Physical Principles and Vertebrate Aquatic Locomotion
Integrative and Comparative Biology, 28
J. L. Feder (1998)
Endless forms: species and speciation
P. Peres‐Neto, P. Magnan (2004)
The influence of swimming demand on phenotypic plasticity and morphological integration: a comparison of two polymorphic charr species
Oecologia, 140
P. Nosil (2012)
Ecological speciation
M. Panova, Johan Hollander, K. Johannesson (2006)
Site‐specific genetic divergence in parallel hybrid zones suggests nonallopatric evolution of reproductive barriers
Molecular Ecology, 15
R. Team (2014)
R: A language and environment for statistical computing.
MSOR connections, 1
M. W. Bruford, O. Hanotte, J. F. Y. Brookfield, T. Burke (1998)
Molecular genetics analysis of populations: a practical approach
S. Gavrilets (2003)
PERSPECTIVE: MODELS OF SPECIATION: WHAT HAVE WE LEARNED IN 40 YEARS?
, 57
R. Zardoya, Dana Vollmer, C. Craddock, J. Streelman, Steve Karl, A. Meyer (1996)
Evolutionary conservation of microsatellite flanking regions and their use in resolving the phylogeny of cichlid fishes (Pisces: Perciformes)
Proceedings of the Royal Society of London. Series B: Biological Sciences, 263
Brant Peterson, Jesse Weber, Emily Kay, H. Fisher, H. Hoekstra (2012)
Double Digest RADseq: An Inexpensive Method for De Novo SNP Discovery and Genotyping in Model and Non-Model Species
PLoS ONE, 7
C. Matessi, A. Gimelfarb, S. Gavrilets (2002)
Long-term Buildup of Reproductive Isolation Promoted by Disruptive Selection: How Far Does it Go?
Selection, 2
S. Via (2009)
Natural selection in action during speciation
Proceedings of the National Academy of Sciences, 106
R. Gutenkunst, Ryan Hernandez, S. Williamson, C. Bustamante (2009)
Inferring the Joint Demographic History of Multiple Populations from Multidimensional SNP Frequency Data
PLoS Genetics, 5
Cheyenne Tait (2018)
Mechanisms of Ecological Divergence and Speciation: Genetic and Neurophysiological Bases for Fruit Volatile Discrimination Behavior in the Apple Maggot Fly ( Rhagoletis pomonella )
J. Wigginton, D. Cutler, G. Abecasis (2005)
A note on exact tests of Hardy-Weinberg equilibrium.
American journal of human genetics, 76 5
C. Hulsey, Richard Roberts, Yong-Hwee Loh, M. Rupp, J. Streelman (2013)
Lake Malawi cichlid evolution along a benthic/limnetic axis
Ecology and Evolution, 3
Andreas Kautt, G. Machado-Schiaffino, A. Meyer (2016)
Multispecies Outcomes of Sympatric Speciation after Admixture with the Source Population in Two Radiations of Nicaraguan Crater Lake Cichlids
PLoS Genetics, 12
Incipient Sympatric Speciation in Lake As
K. Noack, A. Wilson, A. Meyer (2000)
Broad taxonomic applicability of microsatellites developed for the highly polymorphic neotropical cichlid, Amphilophus citrinellum.
Animal genetics, 31 2
M. Deniro, S. Epstein (1978)
Influence of Diet On the Distribtion of Nitrogen Isotopes in Animals
J. Pritchard, Matthew Stephens, P. Donnelly (2000)
Inference of population structure using multilocus genotype data.
Genetics, 155 2
A. Comeault, S. Flaxman, R. Riesch, E. Curran, V. Soria-Carrasco, Z. Gompert, T. Farkas, Moritz Muschick, T. Parchman, T. Schwander, J. Slate, P. Nosil (2015)
Selection on a Genetic Polymorphism Counteracts Ecological Speciation in a Stick Insect
Current Biology, 25
N. Barton (2004)
Fitness Landscapes and the Origin of Species
H. Recknagel, K. Elmer, A. Meyer (2013)
A Hybrid Genetic Linkage Map of Two Ecologically and Morphologically Divergent Midas Cichlid Fishes (Amphilophus spp.) Obtained by Massively Parallel DNA Sequencing (ddRADSeq)
G3: Genes|Genomes|Genetics, 3
W. Rice (1987)
Speciation via habitat specialization: the evolution of reproductive isolation as a correlated character
Evolutionary Ecology, 1
P. Webb (1984)
Body Form, Locomotion and Foraging in Aquatic Vertebrates
Integrative and Comparative Biology, 24
J. Poelstra, Nagarjun Vijay, C. Bossu, H. Lantz, H. Lantz, B. Ryll, I. Müller, I. Müller, V. Baglione, P. Unneberg, M. Wikelski, M. Wikelski, M. Grabherr, J. Wolf (2014)
The genomic landscape underlying phenotypic integrity in the face of gene flow in crows
Science, 344
D. Udovic (1980)
Frequency-Dependent Selection, Disruptive Selection, and the Evolution of Reproductive Isolation
The American Naturalist, 116
C. Oosterhout, William Hutchinson, Derek Wills, P. Shipley (2004)
Micro-Checker: Software for identifying and correcting genotyping errors in microsatellite data
Molecular Ecology Notes, 4
M. Kiljunen, J. Grey, T. Sinisalo, C. Harrod, Hemmo Immonen, Roger Jones (2006)
A revised model for lipid‐normalizing δ13C values from aquatic organisms, with implications for isotope mixing models
Journal of Applied Ecology, 43
J. Catchen, Paul Hohenlohe, S. Bassham, A. Amores, W. Cresko (2013)
Stacks: an analysis tool set for population genomics
Molecular Ecology, 22
L. Excoffier, I. Dupanloup, Emilia Huerta-Sánchez, V. Sousa, M. Foll (2013)
Robust Demographic Inference from Genomic and SNP Data
PLoS Genetics, 9
S. Gavrilets (2005)
“ADAPTIVE SPECIATION”—IT IS NOT THAT EASY: REPLY TO DOEBELI ET AL.
Evolution, 59
D. Bolnick (2011)
Sympatric Speciation in Threespine Stickleback: Why Not?
International Journal of Ecology, 2011
Ryan Martin, D. Pfennig (2009)
Disruptive Selection in Natural Populations: The Roles of Ecological Specialization and Resource Competition
The American Naturalist, 174
M. Minagawa, E. Wada (1984)
Stepwise enrichment of N‐15 along food‐chains – further evidence and the relation between delta‐N‐15 and animal age
, 48
Emily Lescak, S. Bassham, J. Catchen, O. Gelmond, Mary Sherbick, F. Hippel, W. Cresko (2015)
Evolution of stickleback in 50 years on earthquake-uplifted islands
Proceedings of the National Academy of Sciences, 112
X. Thibert-Plante, A. Hendry (2011)
The consequences of phenotypic plasticity for ecological speciation
Journal of Evolutionary Biology, 24
K. Elmer, Shaohua Fan, Helen Gunter, Julia Jones, S. Boekhoff, Shigehiro Kuraku, Axel Meyer (2010)
Rapid evolution and selection inferred from the transcriptomes of sympatric crater lake cichlid fishes
Molecular Ecology, 19
Chung-I Wu (2001)
The genic view of the process of speciation
Journal of Evolutionary Biology, 14
K. Elmer, Henrik Kusche, Topi Lehtonen, A. Meyer (2010)
Local variation and parallel evolution: morphological and genetic diversity across a species complex of neotropical crater lake cichlid fishes
Philosophical Transactions of the Royal Society B: Biological Sciences, 365
P. Edelaar, A. Siepielski, J. Clobert (2008)
Matching Habitat Choice Causes Directed Gene Flow: A Neglected Dimension in Evolution and Ecology
, 62
S. Gavrilets (2005)
“ADAPTIVE SPECIATION”—IT IS NOT THAT EASY: A REPLY TO DOEBELI ET AL
, 59
D. Post (2002)
USING STABLE ISOTOPES TO ESTIMATE TROPHIC POSITION: MODELS, METHODS, AND ASSUMPTIONS
Ecology, 83
Andreas Kautt, K. Elmer, A. Meyer (2012)
Genomic signatures of divergent selection and speciation patterns in a ‘natural experiment’, the young parallel radiations of Nicaraguan crater lake cichlid fishes
Molecular Ecology, 21
L. Rieseberg (2001)
Chromosomal rearrangements and speciation.
Trends in ecology & evolution, 16 7
B. Robinson, K. Parsons (2002)
Changing times, spaces, and faces: tests and implications of adaptive morphological plasticity in the fishes of northern postglacial lakes
Canadian Journal of Fisheries and Aquatic Sciences, 59
J. Catchen, A. Amores, Paul Hohenlohe, W. Cresko, J. Postlethwait (2011)
Stacks: Building and Genotyping Loci De Novo From Short-Read Sequences
G3: Genes|Genomes|Genetics, 1
J. Coyne (2007)
Sympatric speciation
Current Biology, 17
P. Franchini, Carmelo Fruciano, M. Spreitzer, Julia Jones, K. Elmer, F. Henning, A. Meyer (2014)
Genomic architecture of ecologically divergent body shape in a pair of sympatric crater lake cichlid fishes
Molecular Ecology, 23
D. Schluter (1996)
Ecological Speciation in Postglacial Fishes
Philosophical Transactions of the Royal Society B, 351
S. Via (1999)
REPRODUCTIVE ISOLATION BETWEEN SYMPATRIC RACES OF PEA APHIDS. I. GENE FLOW RESTRICTION AND HABITAT CHOICE
Evolution, 53
C. Smadja, R. Butlin (2011)
A framework for comparing processes of speciation in the presence of gene flow
Molecular Ecology, 20
H. Recknagel, Henrik Kusche, K. Elmer, A. Meyer (2013)
Two new endemic species in the Midas cichlid species complex from Nicaraguan crater lakes : Amphilophus tolteca and Amphilophus viridis (Perciformes, Cichlidae)
, 19
C. Martin (2013)
STRONG ASSORTATIVE MATING BY DIET, COLOR, SIZE, AND MORPHOLOGY BUT LIMITED PROGRESS TOWARD SYMPATRIC SPECIATION IN A CLASSIC EXAMPLE: CAMEROON CRATER LAKE CICHLIDS
Evolution, 67
J. Felsenstein (1981)
SKEPTICISM TOWARDS SANTA ROSALIA, OR WHY ARE THERE SO FEW KINDS OF ANIMALS?
Evolution, 35
S. Colborne, S. Garner, F. Longstaffe, B. Neff (2016)
Assortative mating but no evidence of genetic divergence in a species characterized by a trophic polymorphism
Journal of Evolutionary Biology, 29
T. Manousaki, P. Hull, Henrik Kusche, G. Machado-Schiaffino, P. Franchini, C. Harrod, K. Elmer, A. Meyer (2013)
Parsing parallel evolution: ecological divergence and differential gene expression in the adaptive radiations of thick‐lipped Midas cichlid fishes from Nicaragua
Molecular Ecology, 22
M. Barluenga, K. Stölting, W. Salzburger, Moritz Muschick, A. Meyer (2006)
Sympatric speciation in Nicaraguan crater lake cichlid fish
Nature, 439
Catherine Beaudoin, E. Prepas, W. Tonn, L. Wassenaar, B. Kotak (2001)
A stable carbon and nitrogen isotope study of lake food webs in Canada's Boreal Plain
Freshwater Biology, 46
M. Minagawa, E. Wada (1984)
Stepwise enrichment of 15N along food chains: Further evidence and the relation between δ15N and animal age
Geochimica et Cosmochimica Acta, 48
M. Barluenga, A. Meyer (2010)
Phylogeography, colonization and population history of the Midas cichlid species complex (Amphilophus spp.) in the Nicaraguan crater lakes
BMC Evolutionary Biology, 10
Tanja Gernhard, Daniel Ford, Rutger Vos, Mike Steel (2006)
Estimating the Relative Order of Speciation or Coalescence Events on a Given Phylogeny
Evolutionary Bioinformatics Online, 2
N. Franssen, L. Stewart, J. Schaefer (2013)
Morphological divergence and flow-induced phenotypic plasticity in a native fish from anthropogenically altered stream habitats
Ecology and Evolution, 3
V. Ravigné, U. Dieckmann, I. Olivieri (2009)
Live Where You Thrive: Joint Evolution of Habitat Choice and Local Adaptation Facilitates Specialization and Promotes Diversity
The American Naturalist, 174
(2017)
REPRODUCTIVE ISOLATION BETWEEN SYMPATRIC RACES OF PEA APHIDS . 1
H. Recknagel, K. Elmer, A. Meyer (2014)
CRATER LAKE HABITAT PREDICTS MORPHOLOGICAL DIVERSITY IN ADAPTIVE RADIATIONS OF CICHLID FISHES
Evolution, 68
L. Excoffier, G. Laval, S. Schneider (2005)
Arlequin (version 3.0): an integrated software package for population genetics data analysis
, 1
A. Meyer, A. F. Kautt (2014)
Sympatric speciation. Oxford bibliographies in evolutionary biology

Publisher: Wiley
Copyright: © 2016 Published by John Wiley & Sons Ltd.
ISSN: 2045-7758
eISSN: 2045-7758
DOI: 10.1002/ece3.2287
pmid: 27551387
Publisher site: See Article on Publisher Site

Abstract

Demographic inference, habitat isolation, Understanding how speciation can occur without geographic isolation remains individual specialization, phenotypic plasticity, a central objective in evolutionary biology. Generally, some form of disruptive RAD-seq, stable isotopes. selection and assortative mating are necessary for sympatric speciation to occur. Correspondence Disruptive selection can arise from intraspeciﬁc competition for resources. If Axel Meyer, Department of Biology, this competition leads to the differential use of habitats and variation in rele- University of Konstanz, Universit€atsstrasse 10, vant traits is genetically determined, then assortative mating can be an auto- 78457 Konstanz, Germany. matic consequence (i.e., habitat isolation). In this study, we caught Midas Tel: +497531884163; cichlid ﬁsh from the limnetic (middle of the lake) and benthic (shore) habitats Fax: +497531883018; of Crater Lake Asososca Managua to test whether some of the necessary condi- E-mail: [email protected] tions for sympatric speciation due to intraspeciﬁc competition and habitat iso- Funding Information lation are given. Lake As. Managua is very small (<900 m in diameter), European Research Council (Grant/Award extremely young (maximally 1245 years of age), and completely isolated. It is Number: “GenAdap” 293700), Deutsche inhabited by, probably, only a single endemic species of Midas cichlids, Forschungsgemeinschaft (Grant/Award Amphilophus tolteca. We found that ﬁsh from the limnetic habitat were more Number: MA 6144/1-1, TO 914/2-1). elongated than ﬁsh collected from the benthic habitat, as would be predicted from ecomorphological considerations. Stable isotope analyses conﬁrmed that Received: 18 April 2016; Revised: 8 June 2016; Accepted: 9 June 2016 the former also exhibit a more limnetic lifestyle than the latter. Furthermore, split-brood design experiments in the laboratory suggest that phenotypic plas- Ecology and Evolution 2016; 6(15): 5342– ticity is unlikely to explain much of the observed differences in body elongation that we observed in the ﬁeld. Yet, neutral markers (microsatellites) did not reveal any genetic clustering in the population. Interestingly, demographic doi: 10.1002/ece3.2287 inferences based on RAD-seq data suggest that the apparent lack of genetic dif- ferentiation at neutral markers could simply be due to a lack of time, as *These authors contributed equally to this intraspeciﬁc competition may only have begun a few hundred generations ago. work. resources (Martin and Pfennig 2009). The conditions for Introduction sympatric speciation are very restricted (Gavrilets 2005), Empirical studies of the conditions that may allow for or and only if the strength of disruptive selection and the constrain the process of speciation with gene ﬂow and probability of assortative mating are sufﬁciently strong especially its most extreme form, sympatric speciation, may sympatric speciation occur (Udovic 1980). Moreover, are essential to test theoretical predictions (Bolnick 2011; if there are costs associated with choosing a mate, the Martin 2013; Meyer and Kautt 2014; Comeault et al. likelihood of sympatric speciation is strongly reduced 2015). Generally, theory has revealed that ecological sym- (Matessi et al. 2001). On the other hand, if only few loci patric speciation requires some form of disruptive selec- of large effect are underlying ecologically relevant traits tion and assortative mating to overcome the and those conferring assortative mating and both are clo- homogenizing process of gene ﬂow and recombination sely genetically linked (e.g., in close physical proximity or (Felsenstein 1981; Gavrilets 2003, 2004). Disruptive selec- located on an inversion), sympatric speciation is facili- tion can arise from intraspeciﬁc competition, which may tated as their association will be less frequently broken be apparent by individual specialization to certain food down by recombination (Trickett and Butlin 1994; 5342 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Rieseberg 2001). In the most extreme case, traits under earliest stages, holds great promise to identify the relevant selection and those that confer assortative mating are one mechanisms and conditions driving speciation (Via 2009). and the same (Gavrilets 2004). Consequently, if individual Different crater lakes located even within a narrow geo- specialization leads automatically to assortative mating graphic area are often of different age and usually extre- due to the spatial segregation of mating pools (habitat mely remote, isolated bodies of water that are typically isolation) or inﬂuences the propensity of individuals to not connected to other lakes or rivers, making multiple mate with others of a similar phenotype (mate prefer- invasions rather unlikely (but see Martin et al. 2015). ence), population divergence in sympatry is greatly facili- Thus, in contrast to ﬁsh in postglacial lakes in which tated (Gavrilets 2004; Servedio et al. 2011; Smadja and divergence may have often happened due to character dis- Butlin 2011). Importantly, the variation in ecologically placement after a secondary invasion (Schluter 1996), relevant traits and thus individual specialization has to be divergence in crater lake cichlids is often assumed to have genetically determined and not due to phenotypic plastic- happened in sympatry. This should allow for a more ity: While adaptive phenotypic plasticity might increase direct study of the conditions that can lead to speciation the potential for ecological speciation in the long term by with gene ﬂow, not contingent or affected by the initial allowing for the colonization of new environments and amount of population divergence and reproductive isola- population persistence, it will usually impede population tion that was already present at the time of secondary divergence in sympatry (Bolnick 2011; Thibert-Plante and contact. Hendry 2011). Similarly to ﬁsh in postglacial lakes, on the other hand, Thus, some of the necessary (but not sufﬁcient) condi- the evolutionary outcomes of crater lake cichlid popula- tions for sympatric speciation due to intraspeciﬁc compe- tions vary. Haplochromine cichlids in Uganda, for exam- tition for resources and habitat isolation are as follows: ple, have repeatedly evolved an overall more limnetic (1) individuals differ in ecomorphological traits, (2) their body shape without diversifying after the colonization of ecomorphological characteristics match, and individuals crater lakes (Machado-Schiafﬁno et al. 2015). In a small differentially use, distinct habitats, and (3) variation in crater lake in Tanzania, cichlids are diverging along the ecologically relevant traits is genetically determined and littoral zone and the deeper benthic areas instead of the not due to phenotypic plasticity. At the genomic level, benthic–limnetic axis (Malinsky et al. 2015). differentiation in this process is initially expected to be Midas cichlids (belonging to the Amphilophus sp. spe- only reﬂected at the few loci under disruptive selection, cies complex) that have independently colonized several whereas genome-wide differentiation will be virtually crater lakes in Nicaragua from the same source popula- absent and only build up with time after gene ﬂow has tion of the great lakes Nicaragua and Managua, but prob- ceased (Wu 2001; Nosil et al. 2009a; Feder et al. 2012). ably at different time points, show a variable pattern Competition for resources is thought to be the main (Elmer et al. 2010b): They have independently evolved driver of divergence between the ground-associated (ben- into one limnetic and several benthic species in the two thic) and open-water (limnetic) habitats in freshwater Crater Lakes Apoyo and Xiloa (Barluenga et al. 2006; ﬁsh, which has occurred in a variety of ﬁsh taxa in tem- Kautt et al. 2012; Elmer et al. 2014), whereas in other perate as well as tropical environments (Robinson and crater lakes, they have not diversiﬁed along the benthic– Wilson 1994; Hulsey et al. 2013; Elmer et al. 2014). Lim- limnetic axis (Elmer et al. 2010b). We note that the lim- netic ecomorphs exhibit characteristic elongated/fusiform netic niche is completely absent in the relatively shallow body shapes compared to the rather high-bodied benthic (mean depth around 8–12 m) and turbid source lakes ecomorphs (Webb 1984, 1988). Although ﬁsh have diver- (Barlow 1976; Elmer et al. 2010b). One potential explana- siﬁed along the benthic–limnetic axis in a considerable tion is that the variability in body elongation is a key fac- number of cases, still more commonly ﬁsh have not tor that inﬂuences the potential for intrapopulation diversiﬁed and a single generalist population occupies a competition and thus frequency-dependent disruptive lake (Bolnick 2011). Identifying the conditions that have selection. Within the repeated adaptive radiations of facilitated or hindered population divergence and specia- Midas cichlids (where 13 species have been described so tion remains thus a key objective in enhancing our ability far; Recknagel et al. 2013b), this variability seems to be to predict evolutionary processes. correlated with the mean depth of the respective crater Crater lakes in Africa and Nicaragua inhabited by cich- lake (Recknagel et al. 2014). lid ﬁshes represent a unique setting to investigate the pro- In this regard, Crater Lake Asososca Managua – the cess of sympatric speciation in a replicated manner and at focal lake of this study – has a mean depth (54.3 m) sim- different stages of the divergence continuum (Elmer et al. ilar to that of Crater Lake Xiloa (60 m). Further, the pop- 2010b; Kautt et al. 2012; Malinsky et al. 2015). Studying ulation in Lake As. Managua exhibits a large variation in different stages of population divergence, including the body shape elongation, and individual body shapes are ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5343 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. correlated with their lower pharyngeal jaw shapes and cichlids form seasonally monogamous pairs that breed at feeding ecology (Kusche et al. 2014). Thus, one of the the shore, but that pair formation happens probably in conditions (condition 1 outlined above) for divergence the respective habitat. Thus, habitat isolation would act as along the benthic–limnetic axis seems given. Yet, Crater an effective isolating barrier during the time of pair for- Lake As. Managua is thought to harbor only a single pop- mation and not breeding (Gavrilets et al. 2007). ulation of Midas cichlids (Barluenga and Meyer 2010; Another factor that had so far not been explicitly tested Kusche et al. 2014). Importantly, however, ﬁsh had previ- in Midas Cichlids is the role of phenotypic plasticity in ously only been collected at the shore and never in the body shape elongation. While a previous QTL mapping middle of the lake, where limnetic ﬁsh are thought to for- study in closely related limnetic and benthic species of age. Without a representative sample from both habitats, Midas cichlids in Crater Lake Apoyo already demon- population genetic methods are arguably underpowered strated a strong genetic component in the determination for testing population divergence. Asososca Managua is of body elongation (Franchini et al. 2014), in this study also the youngest of the crater lakes (max. 1245 years old; we investigated the contribution of plasticity to the phe- Pardo et al. 2008) known to harbor an endemic species notypic variation in body elongation observed in Midas of the adaptive radiation of Midas cichlids (Elmer et al. cichlids. In other words, we tested whether differences in 2010b), Amphilophus tolteca (Recknagel et al. 2013b), and the ecologically relevant trait of body elongation are pre- the population may thus be at a much earlier stage of sumably mostly genetically determined (condition 3) or divergence than the benthic–limnetic species pairs in Cra- can be attributed to phenotypic plasticity. This question ter Lakes Apoyo and Xiloa. In other words, the propen- was investigated with split-brood design experiments con- sity to diverge along the benthic–limnetic axis in Midas ducted in the laboratory in which one half of the ﬁsh cichlids seems to be related to the depth of the crater were reared in tanks with a constant water ﬂow and the lake, yet the extent of divergence along this axis might other half in control tanks without water ﬂow. In addi- depend on the age of the resident population (time since tion to the focal species from As. Managua, we also sub- colonization). jected broods of three other Midas cichlid species (two Here, we set out to investigate whether the second and limnetic species from different crater lakes and one ben- third conditions for incipient sympatric speciation as out- thic species from the source population of the crater lined above are present and whether population diver- lakes) to the experiment. gence may currently be happening in the extremely young Finally, we used a previously in-house generated RAD- population of A. tolteca in Crater Lake As. Managua. To seq data set from ﬁsh captured at the shore to reconstruct do so, we caught ﬁsh from the open-water zone in the the demographic history of the population in Crater Lake middle of the lake (limnetic habitat) in addition to the As. Managua using coalescent simulations and the site shore (benthic habitat) and used an integrative approach frequency spectrum. This allowed us to put our results of to test whether there is morphological (body elongation), the putative stage of population divergence in perspective ecological (long-term diet in form of stable isotope signa- to the actual time of colonization and size of the founder tures), and genetic (microsatellites) divergence between population. ﬁsh captured in the two different habitats. If the variation in body shape is the result of specialization along the ben- Materials and Methods thic–limnetic axis, we would predict ﬁsh that preferen- tially use the limnetic habitat to exhibit on average a Sampling more elongated body shape and limnetic lifestyle (i.e., a higher proportion of carbon from a limnetic source in Fish were captured with gill nets or angling in Lake Aso- their diet) compared to individuals from the benthic sosca Managua in 2013. Most of the ﬁsh (n = 250) were habitat (condition 2). So far, this prediction has never captured at the shore close to the aqueduct main station been tested in Midas cichlids, as sampling of ﬁsh usually (Fig. 1A). In addition, the protected environment of Cra- occurred exclusively close to the shore (i.e., the benthic ter Lake As. Managua (surrounded by a fence with only habitat) and during the breeding season. Differential habi- restricted access) allowed us to catch a sufﬁcient number tat use, even in a very small lake like As. Managua (ca. of ﬁsh from the middle of the lake (n = 27) by attaching 900 m in diameter), is a key factor that may increase the gill nets to a buoy; without a ﬁxed anchor point, gill nets likelihood of population divergence via habitat isolation get tangled up and leaving nets or a buoy undisturbed in (Servedio et al. 2011). In fact, habitat isolation is a neces- the middle of a lake is not feasible in other crater lakes in sary assumption in a model of sympatric speciation that Nicaragua. For ﬁsh from the middle of the lake and a was speciﬁcally tailored to Midas cichlids in Crater Lake subset (randomly chosen) of those from the shore Apoyo (Gavrilets et al. 2007). We note that Midas (n = 27), standardized photographs were taken from the 5344 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. d A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? (A) (B) Nicaragua L. Asososca Managua Amphilophus tolteca Figure 1. Crater Lake Asososca Managua is (C) located in the center of Nicaragua’s capitol, Limnetic Benthic the city of Managua, and protected from its surroundings. (A) Fish were collected in the Buoy limnetic (middle of the lake) and benthic Limnetic (shore) zones of the lake. (B) Representative specimen of the focal species of this study, Amphilophus tolteca, endemic to As. Managua. (C) Relative frequency (kernel Aqueduct density estimate) of body height index (i.e., a station 200 m measure of overall body elongation) of ﬁsh 0.40 0.42 0.44 0.46 0.48 0.50 collected from the middle of the lake (blue) Body height index (BHI) and the shore (red). lateral view including a size standard. For these samples, remeasured the BHI for a subset of ﬁsh from the shore also tissue samples were taken and preserved in pure (n = 27) – later used for stable isotope and population ethanol for stable isotope and population genetic analyses. genetic analyses – from individual photographs. These An overview about the samples and the morphological, analyzed specimens from the shore represent a random stable isotope, and microsatellite data is provided in subsample of the 250 collected individuals. This random Table S1, Supporting information. Samples for RAD selection and the two different measurement modes are sequencing used for demographic inference (n = 49) were unlikely to have biased our results, as the subset was collected in earlier ﬁeld expeditions in 2007 and 2010 at indistinguishable in terms of the BHI from the whole set the shore and processed in the same way. Sampling was of ﬁsh from the shore (subset: mean = 0.437 0.013 SD; approved and performed according to the regulations of whole set: mean = 0.438 0.014 SD; Welch two-sample the local authorities, the Ministerio de Ambiente y t-test, t = 0.443, P = 0.660). Recursos Naturales, Nicaragua (MARENA). Stable isotopes Measurements of body elongation We compared carbon and nitrogen stable isotope signa- Overall body elongation in ﬁsh is a useful univariate mea- ture of individuals captured in the limnetic (n = 25) and sure to capture the main axis of body shape variation that the benthic zones (n = 25). Stable isotope analyses were distinguishes limnetic and benthic individuals (Kusche conducted on ethanol-preserved muscle tissue extracted et al. 2014; Recknagel et al. 2014). The relative extent of from dorsal musculature. Tissues were dried at 55°C until body elongation was measured in terms of the body there was no more change in mass (ca. 48 h). Muscle height index (BHI). The body height index is deﬁned as samples were subsequently ground up and those between the ratio of body height (distance between insertion of ca. 0.7 and 0.9 mg were placed in tin capsules for analy- the pelvic ﬁns and the most anterior point of the dorsal ses. Gas chromatography–combustion–isotope ratio mass ﬁn) to standard length (distance between the snout and spectrometry was performed at the Isotopes Laboratory of the most posterior point of the caudal peduncle). Higher the Limnological Institute of the University of Konstanz. values of the BHI denote thus more high-bodied (ben- Values of d C & were corrected for lipid content thic) individuals, whereas lower values of the BHI denote (Kiljunen et al. 2006). more elongated (limnetic) individuals. Body height and standard length were initially mea- Phenotypic plasticity sured in the ﬁeld for a large number of individuals cap- tured at the shore (n = 250) with calipers. For ﬁsh from To test whether the observed variation in body shapes the middle of the lake (n = 26; one individual was miss- could be due to phenotypic plasticity, we performed split- ing its pelvic ﬁns and its body height could thus not be brood design breeding experiments in the laboratory. Sin- measured reliably), the BHI was measured from individ- gle broods were divided into two groups and raised at ual photographs. To rule out that the two different mea- approximately equal densities in identical GFK tanks with suring modes (calipers, photographs) had any effect, we rounded corners (ca. 2000 L volume, 160 cm diameter) ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5345 Benthic Density 0 5 10 15 20 25 30 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. either with a constant circular water ﬂow or in control sodium chloride extraction and ethanol precipitation tanks without water ﬂow (Fig. S1, Supporting informa- (Bruford et al. 1998). A total of 52 samples (DNA extrac- tion). The ﬂow regimes in these tanks were designed to tion failed for two samples) from the middle of the lake simulate the increased swimming demand related to for- (n = 26) and the shore (n = 26) were successfully geno- aging in the limnetic habitat compared to the benthic typed at 13 microsatellites loci: Abur28, Abur45, Abur82, habitat. In addition to A. tolteca – the focal species ende- Abur151 (Sanetra et al. 2009); burt-kit (Salzburger et al. mic to Crater Lake As. Managua – we included Amphilo- 2007); M1M, M2, M7, M12 (Noack et al. 2000); TmoM7 phus citrinellus from the great lakes, which has acted as (Zardoya et al. 1996); UNH011, UNH012, UNH013 the source population of all Midas cichlids endemic to (Kellogg et al. 1995). All loci were PCR-ampliﬁed with the crater lakes and is thought to resemble the ancestral ﬂuorescent reverse or forward primers (HEX, FAM, and state of the whole species complex. Moreover, we NED dyes), and fragment lengths were determined with included the two limnetic species of Midas cichlids, the internal size standard Genescan-500 ROX (Applied Amphilophus zaliosus and Amphilophus sagittae, endemic Biosystems, Waltham, MA) on an ABI 3130 Automated to Crater Lakes Apoyo and Xiloa, respectively. Broods of Sequencer (Applied Biosystems, Waltham, MA) with the the four species stemmed from stock ﬁsh kept in the ani- GeneMapper v4.0 (Applied Biosystems, Waltham, MA) mal research facility of the University of Konstanz. At the software. Scoring errors, large allele dropout, and null start of the experiment, ﬁsh of A. sagittae were approxi- alleles were checked employing the program MICRO- mately 5 months (mean weight 2.9 g 1.2 g SD), A. za- CHECKER (Van Oosterhout et al. 2004). The most sup- liosus ca. 7 months (6.8 g 1.7 g SD), A. citrinellus ca. ported number of genetic clusters was determined using 8 months (8.6 g 1.8 g SD), and A. tolteca ca. the model-based Bayesian approach of STRUCTURE v2.3 12 months old (35.4 g 17.2 g SD). Samples sizes for (Pritchard et al. 2000). A burn-in period of 100,000 steps control and treatment groups of the four species were 41 followed by 500,000 Markov chain Monte Carlo iterations and 42 for A. sagittae, 16 and 31 for A. zaliosus, 24 and was sufﬁcient to ensure convergence. Five independent 24 for A. citrinellus, and 16 and 17 for A. tolteca, respec- runs each assuming between one and three genetic clus- tively. In control tanks, water inﬂow from the ﬁlter was ters (K = 1–3) were performed using the admixture directed from the top to the center of the tanks (pump: model (each individual draws some fraction of its genome Eheim type 2260, 65 W power), whereas in treatment from each of the K populations) with correlated allele tanks, the hose conducting the inﬂow was aligned with frequencies. Genetic differentiation by means of the one of the side walls to create a constant circular water ﬁxation index F was estimated with ARLEQUIN ST ﬂow. This current was further enhanced by two additional v.3.5.1.3 (Excofﬁer et al. 2005; Excofﬁer and Lischer water pumps aligned with the ﬁlter hose on the same side 2010) using 20,000 permutations to determine statistical of the tanks. In control tanks, the two additional pumps signiﬁcance. were attached to opposite sides of the tanks and directed at the center canceling out any water current. Water cur- RAD sequencing rent in treatment tanks was measured with an Aquadopp HR-Proﬁler (Nortek) and estimated to be on average ca. Population genomic data were generated following a dou- 14 cm/s across the water column and the whole tank ble-digest RAD sequencing approach (Peterson et al. (measured in 2-min intervals in each of the four corners). 2012) with minor modiﬁcations (Recknagel et al. 2013a). The experiment ran for 6 months, after which all ﬁsh The genomic library (pool of 50 individually barcoded were photographed. Standard length and body height samples) was single-end sequenced for 101 cycles using were taken from the photographs to calculate the BHI. an Illumina HiSeq 2000 platform at the genomics core Due to logistical reasons, ﬁsh from As. Managua could facility of TUFTS University (Boston, MA). After quality only be included 2 months after the start of the experi- inspection (there was no quality drop-off at the end of ment resulting in a total treatment time of 4 instead of the reads, and thus, no trimming was performed), indi- 6 months for this species. Fish were fed without any calo- vidually barcoded reads were demultiplexed using the ric restrictions and handled according to permit number STACKS v.1.29 software pipeline, ﬁltering out low-quality T-13/13. All statistical analyses were carried out in R (R reads (ﬂags: -w 0.1 -s 25 -r -c –q) (Catchen et al. 2011, Development Core Team 2014). 2013). Reads were then mapped to a reference genome assembly of an individual of A. citrinellus from great lake Nicaragua (Elmer et al. 2014) with BWA v.0.7.12 (Li and Microsatellites Durbin 2009). Custom bash scripts were used to remove Total genomic DNA was extracted from ca. 1 mm mus- reads mapping to several positions in the genome, con- cle tissue using a proteinase K digestion followed by taining soft-clipped positions, or showing a mapping 5346 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? quality of less than 25. Genotyping was carried out with than ﬁsh caught at the shore (littoral part of the benthic STACKS using a minimum of ﬁve reads for each locus, zone), the former were on average more elongated than an upper bound of 0.05 for the error rate and a 5% sig- the latter (Fig. 1C): The average BHI of ﬁsh from the niﬁcance level cutoff. The correction module (rxstacks) middle of the lake was 0.430 0.013 SD compared to an was used to correct individual genotype calls and remove average BHI of 0.438 0.014 SD for ﬁsh from the shore loci being confounded in more than 25% of individuals (Welch two sample t-test, t = 3.304, df = 31.08, or showing an excess of haplotypes within populations. P = 0.002). This difference is unlikely to be a result of Individual genotype calls with a log-likelihood of less than allometric effects (e.g., smaller ﬁsh are more elongated) as 10 were ﬁltered out and did not contribute to any sub- ﬁsh in the two groups did not signiﬁcantly differ in abso- sequent analyses. On average, 67,980 13,504 SD loci lute standard length (Welch two-sample t-test, t = 0.591, with a mean coverage of 11.7 1.7 SD reads were df = 35.363, P = 0.558) or in absolute body height obtained per individual. The last two positions of our (Welch two-sample t-test, t = 0.258, df = 33.264, reads exhibited an excess number of, probably artiﬁcial, P = 0.798). polymorphisms, and those loci with such polymorphisms were hence excluded (i.e., blacklisted). Furthermore, loci Stable isotopes deviating from Hardy–Weinberg equilibrium (HWE) (5% signiﬁcance level) or containing more than three SNPs To investigate whether ﬁsh captured in the limnetic zone were excluded from further analyses. HWE exact tests actually exploited this habitat compared to ﬁsh captured (Wigginton et al. 2005) were performed in Plink v.1.19- at the shore that presumably feed predominantly on ben- beta (Purcell et al. 2007). Only one SNP per RAD-tag thic prey, we performed stable isotope analyses on both locus was used for all analyses to reduce the effect of non- groups. Indeed, ﬁsh captured in the middle of the lake independence (linkage) among markers. had signiﬁcantly different carbon and nitrogen isotopic The demographic history was inferred performing coa- signatures compared to those collected at the shore lescent simulations in FASTSIMCOAL2 (Excofﬁer et al. (MANOVA Pillai 0.273, approximate F = 8.808, 2,47 2013). Loci presumably located in coding regions were P = 0.001). This suggests that the long-term diet of these identiﬁed via a BLASTN search against a compilation of two groups differs both in the source of carbon (univari- transcriptomic data from various species and tissues of ate ANOVA, F = 11.949, P = 0.001) as well as the 1,48 Midas cichlids (Elmer et al. 2010a; Henning et al. 2013; trophic level as indicated by the nitrogen isotope compo- Manousaki et al. 2013) and excluded. The minor joint sition (univariate ANOVA, F = 8.230, P = 0.006). Fish 1,48 site frequency spectrum (MSFS) was created as described collected in the limnetic zone of the lake were depleted in 13 13 in Kavembe et al. (2016). Brieﬂy, sample sizes were pro- C (mean d C = 33.34 1.28 SD) compared to those jected downwards to 25 individuals (50 alleles) using from the benthic zone (mean d C = 32.05 1.37 SD) dadi’s projection function (Gutenkunst et al. 2009). The ﬁnal two-dimensional SFS was built from an effective Limnetic 16.6 sequence length of 3.78 Mb and contained 10,075 poly- Benthic morphic sites. Inferred parameters were converted into 16.2 demographic units using a substitution rate of 15.8 7.5 9 10 per site and generation (Guo et al. 2013). For each demographic model, between 50 and 125 indepen- 15.4 dent runs were performed, consisting each of 50 rounds 15.0 of parameter estimation via the ECM algorithm with a length of 100,000–250,000 coalescent simulations each 14.6 (increasing by 5000 steps each round). Parametric boot- 14.2 strapping (n = 50) was performed to obtain 95% conﬁdence intervals (Excofﬁer et al. 2013). –36 –35 –34 –33 –32 –31 –30 –29 δ13C‰ Limnetic Benthic Results Origin of carbon Figure 2. Fish collected from the limnetic zone (blue) exhibit on Body shape differences average higher values of d N than ﬁsh from the benthic zone (red), In concordance with our prediction that ﬁsh collected in suggesting a slightly higher trophic position, and lower values of the open-water area of the lake (limnetic zone) would d C, implying that their source of carbon is of a more limnetic origin. exhibit on average a more fusiform/elongated body shape Large dots represent the mean SD. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5347 Trophic level Lower Higher δ15N‰ Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. (Fig. 2). Values of d C allow to draw inferences about water zone could lead to more elongated body shapes due the relative contribution of different primary producers as to an increased swimming demand (Johansson and a source of carbon of lacustrine animals because plank- Andersson 2009). Thus, in an attempt to test whether the tonic primary producers are depleted in C compared to differences in body shape elongation could be at least benthic primary producers (France 1995a,b; Hecky and partly explained by phenotypic plasticity, that is, whether Hesslein 1995). Fractionation between consumers and enduring swimming would lead to more elongated body their food is negligible, resulting in d C values being shapes, we performed experiments in a split-brood design. conserved throughout the food chain (DeNiro and Using single broods, in addition to the focal species of Epstein 1981). The difference in d C values between pri- this study, A. tolteca, we included A. citrinellus from the mary producers is thus consistently passed on to primary great lakes (the source population of the crater lakes) and and secondary consumers (Zanden and Rasmussen 1999; the two described limnetic species of Midas cichlids, Beaudoin et al. 2001; Post 2002). Our data suggest that A. zaliosus and A. sagittae, from Crater Lakes Apoyo and the diet of ﬁsh captured in the limnetic zone of the lake Xiloa, respectively. After four to six months of treatment had a higher proportion of carbon of a limnetic origin (four for A. tolteca and six for the other three species), than the diet of ﬁsh sampled in the benthic zone. we found that neither treatment nor the interaction of Furthermore, ﬁsh captured in the limnetic zone were species and treatment had a signiﬁcant effect on body 15 15 signiﬁcantly enriched in N (mean d N = 15.71 0.38 elongation (two-way ANOVA, F = 1.126, P = 0.290; SD) compared to those captured in the benthic zone F = 1.784, P = 0.152), while there were signiﬁcant differ- (mean d N = 15.35 0.50 SD) (Fig. 2). Consumers ences among the four species (F = 205.880, 15 16 generally become enriched in N compared to their food P < 2 9 10 ). The average body height indices for the source (DeNiro and Epstein 1981; Minagawa and Wada control and treatment groups after the experiment were 1984; Post 2002). Hence, d N values of consumer’s tis- 0.464 0.017 SD and 0.464 0.014 SD for A. tolteca, sues serve as an indicator of average trophic position 0.476 0.009 SD and 0.477 0.014 SD for A. citrinellus, (Zanden and Rasmussen 1999; Post 2002). The difference 0.424 0.012 SD and 0.421 0.009 SD for A. zaliosus, in the mean value of d N between ﬁsh captured in the and 0.436 0.011 SD and 0.442 0.011 SD for A. sagit- middle and the shore of the lake was relatively small with tae, respectively (Fig. 3). The relatively high BHI values of the former being on average ca. 0.5 & higher than the individuals from A. tolteca in our plasticity experiment latter. The difference between the two groups could be due to two factors. On the one hand, following the stan- dard interpretation of d N values, ﬁsh from the limnetic Control zone of the lake might occupy a slightly higher trophic Treatment position than those from the benthic zone as they con- sume on average more secondary consumers. Alterna- tively, the difference could be due to the fact that the primary producers supporting the food chains of benthic and limnetic ﬁsh differ in their d N values, as planktonic primary producers are slightly enriched in N compared to benthic primary producers (Zanden and Rasmussen 1999). Given the observed difference in d C values between the two groups – suggesting a different origin of carbon – this is also a plausible explanation. Either way, overall our data suggest a rather similar position in the trophic chain, yet evidence for clear differences in the diet emerges from the stable isotope data between ﬁsh from the limnetic zone of the lake and those from the benthic zone. Phenotypic plasticity Figure 3. Body height index of control (shaded in gray) and In concordance with our predictions, ﬁsh from the ben- treatment (white) groups of all four species included in the thic and limnetic zones differed signiﬁcantly in body phenotypic plasticity experiment. Only species differ in their body elongation and long-term diet. A possible explanation for height index (body elongation), whereas treatment did not have a this result is phenotypic plasticity. Foraging in the open- signiﬁcant effect. 5348 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. tolteca L. As. Managua A. citrinellus source lakes A. zaliosus L. Apoyo A. sagittae L. Xiloá Body height index (BHI) 0.42 0.44 0.46 0.48 0.50 A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? compared to wild-caught ﬁsh are probably due to the fact (900 m in diameter) and isolated crater lake. We simu- that we used single broods and the parents of the lated data in a coalescent framework according to 13 dif- A. tolteca brood were by chance relatively high-bodied. ferent demographic models (visualized in Fig. S2, The differences among species reﬂect the differences Supporting information) and evaluated the ﬁt of the sim- found in nature (Elmer et al. 2010b) and are probably ulations against the empirical data summarized in the not merely inﬂuenced by the slightly different age of the MSFS (Table S2, Supporting information) (Excofﬁer et al. four groups. Thus, plasticity induced by increased swim- 2013). As a source population, we used A. citrinellus from ming demands, as expected for ﬁsh inhabiting the lim- Lake Managua (n = 50, see Kautt et al. 2016). netic zone, is unlikely to contribute signiﬁcantly to the The most strongly supported model includes a “bottle- large degree of inter- and intraspeciﬁc variation in body growth” scenario (i.e., a population reduction followed by elongation observed in Midas cichlids. exponential growth) in the source population, exponential growth in the population of Crater Lake As. Managua after its colonization, and a secondary admixture event as Genetic differentiation well as continuous migration from the source lake into The differences in body shapes and long-term diet sug- the crater lake (Fig. 5). According to the maximum-likeli- gested that ﬁsh exhibit individual differences in their hood point estimates of the most supported model, the habitat preference, with some ﬁsh predominantly exploit- source population was reduced from 20,439 (95% CI: ing the limnetic open-water habitat, while others exploit 19,289–21,523) individuals to only 1556 (1047–2282) the benthic shore-associated habitat. We were interested individuals in a bottleneck event 2080 (1498–2885) gener- in whether there is any genetic divergence at neutral ations ago, recovering to a current population size of markers between the groups. Based on 13 microsatellite 260,429 (0–712,004) individuals. Note that conﬁdence markers, we did not ﬁnd any signs of genetic differentia- intervals for population sizes are relatively broad as even tion or population structuring: The overall F -value slight differences in the estimated exponential growth ST between the two groups was 0.005 (20,000 permutations, rates of the parametric bootstrap replicates can translate P = 0.720). Similarly, a STRUCTURE analysis did not to large deviations in the population sizes. Crater lake As. reveal more than one genetic cluster. Assuming a priori, Managua was colonized 797 (95% CI: 516–1284) genera- two clusters (K = 2) resulted in admixture proportions tions ago by only 32 (0–71) individuals. Since then, it has around 50% for all individuals (Fig. 4). 20 439 Past Colonization history 2080 generations Given the apparent absence of genetic structuring, but the 1556 fact that some of the necessary conditions for sympatric speciation are present in the population of Lake As. Man- 797 generations agua, we hypothesized that the exceedingly young age of 32 the crater lake population has not been sufﬁcient to build-up genetic differentiation at neutral markers. Aso- 507 generations 32.3% sosca Managua is geologically the youngest crater lake in Nicaragua known to house Midas cichlids (max. age ca. 1245 years; Pardo et al. 2008), yet the exact timing of col- 19 460 onization and the size of its founder population and its –5 8.95 x 10 260,429 current size had not been inferred before. To this end, we Present used RAD-seq data to infer the demographic history of Source Crater Lake the endemic Midas cichlid population of this very small (great Lake Managua) As. Managua 1.00 Figure 5. Schematic depiction of the most supported demographic 0.80 model and the associated maximum-likelihood point estimates of 0.60 demographic parameters. Numbers in the model refer to number of 0.40 diploid individuals, time points refer to number of generations, and 0.20 the migration rate refers to the probability for an allele to migrate 0.00 Limnetic Benthic into another deme (in forward time). Note that the model is not drawn in scale, but merely indicates differences in timing and Figure 4. Structure analysis based on 13 microsatellite markers population sizes. In addition, population growth was modeled to be assuming K = 2 clusters. exponential and not linear as shown here. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5349 Ancestry Coalescence Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. been growing to a current population size of 19,460 and, importantly, habitat. The match between morphol- (5336–43,039) individuals. In an admixture event that ogy and stable isotope signature is in agreement with our happened 507 (384–652) generations ago, the crater lake previous work (Kusche et al. 2014); more elongated ﬁsh 13 15 population received 32.3% (18.4–50.1%) of its gene pool were depleted in C and enriched in N compared to from the source population (great lake Managua). In less elongated ﬁsh. Yet, we did not recover the whole addition, since its colonization migration from the source range of the previously observed variation in d C and lake (L. Managua) into Crater Lake As. Managua has d N values (Kusche et al. 2014). Our goal was to com- happened with a probability of 8.95 9 10 pare the time-integrated diet of ﬁsh utilizing the limnetic 5 4 (5.40 9 10 –1.14 9 10 ), that is, ca. 9 of 100,000 alle- habitat versus those utilizing the benthic habitat, rather les, per generation. Evidence for an admixture event was than to describe the variation in diet associated with mor- unexpected, yet support for admixture events into Crater phology. Therefore, we did not choose the most extreme Lakes Apoyo and Xiloa was found recently (Kautt et al. high-bodied ﬁsh from the shore, but a randomly selected 2016). Whether the admixture event into Crater Lake As. subsample of all the ﬁsh captured at the shore. Thus, Managua is real and has facilitated divergence remains to most of the ﬁsh from the shore had an intermediate BHI be validated and tested. (subset mean = 0.437). Potentially, we could have recov- In conclusion, Crater Lake As. Managua seems to have ered a higher range by biasing our sample to include been colonized by a small founder population (only more ﬁsh with an extremely high BHI. However, this around 32 individuals) and very recently (ca. 800 genera- relation was already clearly established (Kusche et al. tions ago). Assuming a generation time of 1–2 years (Bar- 2014). It remained unclear though, if the observed pattern luenga and Meyer 2010), the colonization seems to have was associated with differences in habitat use per se.By happened shortly after the formation of the crater lake sampling Midas cichlids for the ﬁrst time from the open- itself (around 1245 years ago). water zone, we provide evidence that the depletion in C and enrichment in N of elongated ﬁsh compared to less elongated ﬁsh is most likely due to the differential use of Discussion the limnetic and benthic habitats within the lake. Surely, Based on the observed correspondence of habitat use with limnetic and benthic habitats in a lake are not completely morphology and diet, we can infer that Midas cichlids in discrete and treating them as two distinct habitats is cer- Crater Lake Asososca Managua specialize along the ben- tainly an oversimpliﬁcation, yet this distinction is a com- thic–limnetic axis: Fish from the limnetic zone of the lake mon and reasonable assumption (Bolnick 2011). are on average more elongated than ﬁsh from the benthic While we did not perform a mark–recapture study to zone and their diet reﬂects a more limnetic lifestyle and test habitat ﬁdelity per se (e.g., Bolnick et al. 2009), the vice versa. Our laboratory split-brood design experiments stable isotope signatures suggested that ﬁsh fed on average suggest that the differences in body elongation are proba- consistently, over long periods of time, to a different bly not to a large extent attributable to phenotypic plas- extent on diets that would be typical of the benthic and ticity, but are probably mostly genetically determined. limnetic habitats. Therefore, we conclude that individuals Yet, despite differential habitat use and a presumably preferentially, although probably not exclusively, use strong genetic basis, no population divergence is appar- either the benthic or the limnetic habitat. Whether the ent: Population genetic analyses show support for only match between eco-morphological traits is due to geneti- one genetic cluster. According to our demographic infer- cally based habitat preference or matching habitat choice ences, the population of Midas cichlids from Lake As. (sensu Edelaar et al. 2008) remains to be tested, but Managua is very young (ca. 800 generations ago) and has importantly, both processes increase the chance of habitat been colonized by a very small founder population (about isolation and thus speciation (Edelaar et al. 2008; Bolnick 32 individuals). This suggests that intraspeciﬁc competi- et al. 2009; Ravigne et al. 2009). tion for resources has probably only begun a few hundred generations ago, which could explain the apparent lack of Differences in body elongation are probably population divergence uncovered so far at the genomic mostly genetically determined rather than level. plastic A major question concerning the morphological differ- Differential habitat use of the benthic and ences between ﬁsh from the limnetic and benthic zones of limnetic habitats the lake is whether they could be due to phenotypic plas- Morphological measurements and stable isotope signa- ticity. If phenotypic plasticity was the main factor tures show a match of body elongation, long-term diet explaining the differences, population divergence and 5350 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? ultimately speciation may be unlikely (Edelaar et al. 2008; trigger this stimulus. Thus, we conclude that phenotypic Bolnick 2011; Thibert-Plante and Hendry 2011). In our plasticity does not seem to play a major role in explaining phenotypic plasticity experiment, a plastic response was the observed morphological differences. not induced in any of the four tested species: neither This conclusion is further supported by a study on the A. citrinellus from the great lakes (the source population genetic basis of the benthic–limnetic morphological diver- of all crater lakes) that resembles the ancestral state, nor gence of sympatric Midas cichlid species in Crater Lake the strongly elongated limnetic species of Midas species Apoyo that found that the divergent body shapes are A. zaliosus (Crater Lake Apoyo) and A. sagittae (Crater maintained in the laboratory and furthermore identiﬁed Lake Xiloa), nor the variable focal species of this study QTLs that explain some of the morphological differences A. tolteca exhibited a plastic response in this treatment (Franchini et al. 2014). The signiﬁcant differences in body (Fig. 3). This result stands in contrast to the fact that shape that we found among the four species in this study cichlids exhibit phenotypic plasticity in a number of dif- (Fig. 3) lend further support to the earlier ﬁndings that ferent traits (Meyer 1987; Kerschbaumer et al. 2011; the differences between species are maintained in captivity Machado-Schiafﬁno et al. 2014). We note, however, that even after one to two generations and are most likely to a we only investigated whether treatment had an effect on large extent genetically determined. the elongation of the main body axis as this was the main trait we were interested in. It is thus possible that an Lack of population differentiation possibly undetected plastic response was induced in our experi- due to very recent origin of intraspeciﬁc ment (e.g., in other aspects of morphology, behavior, or competition physiology). One potential caveat of this experiment is that only Population divergence along the benthic–limnetic axis is one brood per species was tested. Thus, the results cannot common in freshwater ﬁsh (Robinson and Wilson 1994) be readily generalized to infer the degree of plasticity that and is the basis of sympatric speciation in at least two may exist within the entire species. Phenotypic plasticity radiations of Crater Lake Midas cichlids (Barluenga et al. itself may be variable within and among species 2006; Elmer et al. 2014). Despite this and the fact that (Machado-Schiafﬁno et al. 2014). However, we believe ﬁsh in Lake As. Managua differently use the limnetic and that the fact that none of the four species showed any benthic habitats, our population genetic microsatellite pronounced plastic response supports our conclusion. It data suggest that ﬁsh from the middle of the lake are not is also possible that the treatment was applied too late in genetically differentiated from those captured at the shore. their ontogeny as earlier developmental stages might be Note that we are here referring to genetic differentiation more susceptible to exhibit a plastic response than later at neutral markers and not to relatively restricted highly stages or that the treatment was not strong enough. Yet, diverged regions that might differentiate the ecotypes the strength of our treatment was comparable to other (e.g., Malinsky et al. 2015). This lack of genetic structur- studies using a constant water ﬂow in an attempt to ing is in agreement with previous investigations based on induce a plastic response in ﬁsh (Peres-Neto and Magnan microsatellite markers that did not ﬁnd any evidence for 2004; Franssen et al. 2013). more than one genetic cluster in Crater Lake As. Mana- Alternatively, the treatment may have been biologically gua (Barluenga and Meyer 2010; Kusche et al. 2014). It is unrealistic in mimicking the natural conditions. Gener- important to note, though, that these studies differed ally, limnetic ﬁsh are thought to exhibit a more fusiform from our approach in that they exclusively used individu- body shape as an adaptation for increased swimming als captured at the shore. demand (Webb 1984, 1988), yet this is not driven by a Because samples for RAD-seq were collected exclusively constant water current, but rather related to the mode of from the shore, we could not use this data to test for foraging in the open water. Enclosure experiments in the genetic differentiation between ﬁsh from the limnetic and wild restricting individuals to the shore or the open-water benthic zones; this part of the project was started before habitat for foraging would be a more accurate way of we sampled ﬁsh for the ﬁrst time from the middle of the testing a role of phenotypic plasticity (Robinson and Par- lake. Nonetheless, as the two groups do not seem to be sons 2002), but such experiments have so far not been differentiated at neutral markers (yet), our demographic feasible in Nicaragua for logistical reasons. Nonetheless, inferences should not be affected by using only ﬁsh from whether driven by a different mode of foraging, predator the shore and our estimates of the colonization time and avoidance, or any other potentially unknown reason, the population size therefore unbiased. According to our relevant biological stimulus leading to a more elongated demographic inferences, Crater Lake As. Managua has body shape is most likely the resulting increased swim- been colonized only around 800 (516–1284, 95% CI) gen- ming demand. We think our treatment did effectively erations ago and by a very small founder population ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5351 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. consisting of only around 32 (0–71, 95% CI) individuals. allowed for enough time to build up genetic differentia- The estimate of the current population size of around tion throughout the genome. Alternatively, the population 19,500 individuals seems biologically reasonable; Midas may be stalled in its divergence, and the speciation pro- cichlids are by far the most abundant ﬁsh in L. As. Mana- cess may never be completed (Matessi et al. 2001; Nosil gua (pers. observation) and are today likely at carrying et al. 2009b). capacity in the relatively small crater lake (900 m in diameter; surface area of 0.74 km ). Yet, right after colo- Midas cichlids in Crater Lake As. Managua: a nization of the new underexploited environment of the case for incipient sympatric speciation? crater lake, the small founder population will almost cer- tainly not have been limited by resources. Whether population divergence will proceed and ulti- Generally, selection pressures in the beginning will mately lead to speciation depends foremost on the most likely have been directed at better adapting the strengths of disruptive selection and assortative mating founder population as a whole to the crater lake environ- (Gavrilets 2005; Bolnick 2011). Determining the strength ment in general: Crater lakes are very deep and their of selection acting on the benthic–limnetic divergence has water is usually very clear in contrast to the great lakes so far proofed not feasible in Midas cichlids due to the (the source) that are relatively shallow (mean depth of difﬁculty of performing experiments in the ﬁeld (e.g. 8–12 m) and turbid (Barlow 1976; Elmer et al. 2010b). In placing enclosures in the middle of a crater lake) or real- this regard, it is interesting that A. tolteca has evolved a istically resembling the open-water niche of a crater lake distinct morphology from the source population (Reck- (up to 200 m deep) in the laboratory. Yet, the fact that nagel et al. 2013b) in only 800 generations. It is possible sympatric speciation along the benthic–limnetic axis has that the founder effect that we have identiﬁed here has happened in two other crater lakes (Elmer et al. 2014) facilitated the rapid divergence of the crater lake popula- suggests that selection pressures along the benthic–lim- tion as a whole in allopatry (e.g., Kolbe et al. 2012), yet netic axis in Nicaraguan crater lakes have been at least in population divergence after the colonization of a new some cases sufﬁcient in driving sympatric speciation in environment due to selection can commence extremely Midas cichlids. However, the speciﬁcs matter and every fast (Lescak et al. 2015). lake environment and every population’s demographic In any case, only with time will the population have history will result in different conditions that may or may become large enough for intraspeciﬁc competition for not be conducive to sympatric speciation (Bolnick 2011; resources to elicit frequency-dependent disruptive selec- Martin 2013). tion. Resource limitation due to a high population den- While our results suggest that ﬁsh in Crater Lake As. sity is a necessary condition for stable disruptive selection Managua differentially use the benthic and limnetic habi- (Bolnick 2011). Hence, it seems possible that the pro- tats, it is currently unclear if this would readily translate cesses of disruptive selection and assortative mating have to reproductive isolation by habitat isolation like in phy- been at work in A. tolteca, but that genetic differentiation tophagous insects, for example (Rice 1987; Feder 1998; has not built up at neutral markers across the genome Via 1999). Midas cichlids form seasonally monogamous yet. Neutral genetic differentiation may not be expected pairs that breed at the shore (Barlow 1992), and hence, in the earliest stages of divergence (Elmer et al. 2010c; the spatial segregation breaks down during the time of Colborne et al. 2016). It seems likely that only few breeding (Baylis 1976). Yet, if pair formation happened in regions in the genome, that is, small genomic islands, that the respective habitat before pairs move to the shore to harbor the genetic basis for the observed differences in breed, differential habitat use would effectively result in morphology and trophic ecology are differentiated assortative mating by habitat (Gavrilets et al. 2007). between the benthic and limnetic ecomorphs in Crater Behavioral experiments have shown that sympatric ben- Lake As. Managua, as has been found recently between thic and limnetic species from Crater Lakes Apoyo and littoral and benthic ecomorphs of crater lake cichlids in Xiloa mate assortatively even under laboratory conditions Tanzania (Malinsky et al. 2015) or carrion and hooded (i.e., in the absence of different habitats), but that pair crows in Europe (Poelstra et al. 2014). Note that our formation happens before the pairs establish territories RAD-seq data set did not explicitly include individuals for breeding (Baylis 1976; Kautt et al., unpublished data). from the two habitats and we could thus not perform Hence, active mate preference seems to be a strong mech- outlier tests for signatures of selection. anism leading to assortative mating in these species, but Altogether, it seems that the population in Lake As. this does not negate the possibility that habitat isolation Managua is at the earliest stages of population divergence, is still contributing to reproductive isolation or has played and we propose that the very young age of the population a role in the initial divergence of limnetic and benthic and the even later onset of disruptive selection have not species in these two crater lakes. Different reproductive 5352 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? barriers often act in concert, and their contribution to and limnetic habitats. Whether these conditions are sufﬁ- overall reproductive isolation can change over time (Nosil cient for sympatric speciation to occur remains to be and Schluter 2011; Nosil 2012). Whether, and to what tested (e.g., Bolnick 2011; Martin 2013). Here, we did extent, assortative mating due to habitat isolation or mate not ﬁnd evidence for neutral genetic divergence in the preference exists in A. tolteca remains to be tested, but population. This could either mean that the strengths of our study shows that the conditions for assortative mat- disruptive selection and assortative mating are not strong ing by habitat isolation can exist even in a very small cra- enough and that the population is therefore stalled in its ter lake such as As. Managua. divergence (Matessi et al. 2001) or that we are dealing with a very recent population divergence (Elmer et al. 2010c; Colborne et al. 2016). While ecological experi- Conclusions ments are needed to test the former explanation, our Divergence along the benthic–limnetic axis is a common demographic inferences suggest that it is plausible that theme in the diversiﬁcation of freshwater ﬁshes (Schluter the apparent lack of genetic divergence is due to a lack of and McPhail 1992; Robinson and Wilson 1994; Hulsey time. In conclusion, our study shows how knowledge et al. 2013; Colborne et al. 2016), and in at least two cra- about the demographic history can inform on studies of ter lakes in Nicaragua, Midas cichlid ﬁsh have speciated speciation and that some of the necessary conditions for along this axis in sympatry (Elmer et al. 2014; Kautt et al. sympatric speciation do occur in nature and can occur 2016). Theory predicts that disruptive selection due to even in such a small environment as Crater Lake intraspeciﬁc competition for resources and assortative Asososca Managua. mating due to habitat isolation are necessary conditions for this process (Gavrilets et al. 2007). However, this pre- Acknowledgments diction has never been tested before in Midas cichlids. In complement to previous studies that have focused on the Sample collection permits were granted by MARENA, benthic and limnetic species in Crater Lakes Apoyo and Nicaragua. We are grateful to Empresa Nicaraguense de Xiloa (Barluenga et al. 2006; Kautt et al. 2012; Elmer Acueductos y Alcantarillados (ENACAL) and particularly et al. 2014), in this study we investigated a much earlier Ingeniero Martin Brenes for support of this study. The stage of the divergence continuum: population divergence coalescence analyses were partly performed on the compu- along the benthic–limnetic axis in a species of Midas tational resource bwUniCluster funded by the Ministry of cichlids, A. tolteca (Recknagel et al. 2013b), endemic to Science, Research and Arts and the Universities of the the extremely young and small Crater Lake As. Managua. State of Baden-Wurttemberg, € within the framework pro- More speciﬁcally, we studied whether some of the neces- gram bwHPC and partly on the High Performance Com- sary conditions for sympatric speciation due to intraspeci- puting (HPC) cluster of the University of Konstanz. We ﬁc competition for resources and habitat isolation are thank the Limnological Institute of the University of Kon- given in Lake As. Managua. stanz for help with the stable isotope analyses, in particu- In agreement with the prediction of a match between lar Elizabeth Yohannes, and Hilmar Hofmann for help phenotype and habitat, we found that individuals caught with measuring the water current in the phenotypic plas- in the limnetic habitat are more elongated than ﬁsh col- ticity experiment. Moreover, we want to thank two anony- lected in the benthic habitat. Stable isotope analyses – by mous reviewers whose comments helped to improve this integrating diet over longer time spans – further con- manuscript. A. F. Kautt was supported by the Landes- ﬁrmed that ﬁsh from the limnetic habitat also exhibit a graduiertenforderung € (LGFG) of the State of Baden- more limnetic lifestyle. Thus, we conclude that individu- Wurttemberg € and the International Max Planck Research als differentially use the two habitats. Together with pre- School (IMPRS) for Organismal Biology. G. Machado- vious evidence (Franchini et al. 2014), our experiments Schiafﬁno was supported by the Alexander von Humboldt conducted in the laboratory suggest that the differences Foundation and a grant from the Deutsche Forschungsge- we found in the most relevant ecological trait, body meinschaft (MA 6144/1-1). J. Torres-Dowdall was sup- shape elongation, are unlikely to be due to phenotypic ported by an EU FP7 Marie Curie Zukunftskolleg plasticity and are probably strongly genetically deter- Incoming Fellowship Program of the University of Kon- mined. Altogether, our data therefore support the notion stanz (grant number 291784) and a grant from the that some of the necessary conditions for sympatric spe- Deutsche Forschungsgemeinschaft (TO 914/2-1). The ciation are present in Midas cichlids in Crater Lake As. study was funded by support of the University of Kon- Managua: Individuals vary in genetically determined eco- stanz and a European Research Council advanced grant morphological traits and differentially use the benthic (ERC “GenAdap” 293700) to A. Meyer. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5353 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. divergence in a species characterized by a trophic Data Accessibility polymorphism. J. Evol. Biol. 29:633–644. Morphometric measurements, stable isotope data, and Comeault, A. A., S. M. Flaxman, R. Riesch, E. Curran, V. microsatellite genotypes are provided in Table S1, Sup- Soria-Carrasco, Z. Gompert, et al. 2015. Selection on a porting information. The variant call format (VCF) of the genetic polymorphism counteracts ecological speciation in a RAD-seq data is available as Data set S1, Supporting stick insect. Curr. Biol. 25:1975–1981. information. DeNiro, M. J., and S. Epstein. 1981. Inﬂuence of diet on the distribution of nitrogen isotopes in animals. Geochim. Cosmochim. Acta 45:341–351. Conﬂict of Interest Edelaar, P., A. M. Siepielski, and J. Clobert. 2008. Matching habitat choice causes directed gene ﬂow: a neglected None declared. dimension in evolution and ecology. Evolution 62:2462–2472. Elmer, K. R., S. Fan, H. M. GunterJ. C. Jones, S. Boekhoff, S. References Kuraku, et al. 2010a. Rapid evolution and selection inferred Barlow, G. W. 1976. The Midas cichlid in Nicaragua. Pp. 333– from the transcriptomes of sympatric Crater lake cichlid 358 in T. B. Thorson, ed. Investigations of the ichthyology ﬁshes. Mol. Ecol. 19:197–211. of Nicaraguan lakes. Univ. of Nebraska Press, Lincoln. Elmer, K. R., H. Kusche, T. K. Lehtonen, and A. Meyer. Barlow, G. W. 1992. Is mating different in monogamous species? 2010b. Local variation and parallel evolution: morphological The Midas cichlid ﬁsh as a case study. Am. Zool. 32:91–99. and genetic diversity across a species complex of neotropical Barluenga, M., and A. Meyer. 2010. Phylogeography, crater lake cichlid ﬁshes. Philos. Trans. R. Soc. Lond. B Biol. colonization and population history of the Midas cichlid Sci. 365:1763–1782. species complex (Amphilophus spp.) in the Nicaraguan Elmer, K. R., T. K. Lehtonen, A. F. Kautt, C. Harrod, and A. crater lakes. BMC Evol. Biol. 10:326. Meyer. 2010c. Rapid sympatric ecological differentiation of Barluenga, M., K. N. Stolting, W. Salzburger, M. Muschick, crater lake cichlid ﬁshes within historic times. BMC Biol. and A. Meyer. 2006. Sympatric speciation in Nicaraguan 8:60. Crater lake cichlid ﬁsh. Nature 439:719–723. Elmer, K. R., S. H. Fan, H. Kusche, M. L. Spreitzer, A. F. Baylis, J. R. 1976. Quantitative study of long-term courtship: Kautt, P. Franchini, et al. 2014. Parallel evolution of 1. Ethological Isolation between sympatric populations of Nicaraguan crater lake cichlid ﬁshes via non-parallel routes. the Midas cichlid, Cichlasoma citrinellum, and the Arrow Nat. Commun. 5:5168. Cichlid, C. zaliosum. Behaviour 59:59–69. Excofﬁer, L., and H. E. L. Lischer. 2010. Arlequin suite ver Beaudoin, C. P., E. E. Prepas, W. M. Tonn, L. I. Wassenaar, 3.5: a new series of programs to perform population and B. G. Kotak. 2001. A stable carbon and nitrogen isotope genetics analyses under Linux and Windows. Mol. Ecol. study of lake food webs in Canada’s Boreal Plain. Freshw. Resour. 10:564–567. Biol. 46:465–477. Excofﬁer, L., G. Laval, and S. Schneider. 2005. Bolnick, D. I. 2011. Sympatric speciation in threepine Arlequin (version 3.0): an integrated software package for stickleback: why not? Int. J. Ecol. 2011:15. population genetics data analysis. Evol. Bioinform. Online Bolnick, D. I., L. K. Snowberg, C. Patenia, W. E. Stutz, T. 1:47–50. Ingram, and O. L. Lau. 2009. Phenotype-dependent native Excofﬁer, L., I. Dupanloup, E. Huerta-Sanchez, V. C. Sousa, habitat preference facilitates divergence between Parapatric and M. Foll. 2013. Robust demographic inference from Lake and stream stickleback. Evolution 63:2004–2016. genomic and SNP data. PLoS Genet. 9:e1003905. Bruford, M. W., O. Hanotte, J. F. Y. Brookﬁeld, and T. Burke. Feder, J. L. 1998. The apple maggot ﬂy, Rhagoletis pomonella. 1998. Multilocus and single locus DNA ﬁngerprinting. Pp. Pp. 130–144 in D. J. Howard and S. H. Berlocher, eds. 287–336 in A. R. Hoelzel, ed. Molecular genetics analysis of Endless forms: species and speciation. Oxford Univ. Press, populations: a practical approach. Oxford Univ. Press, Oxford, U.K. Oxford, U.K. Feder, J. L., S. P. Egan, and P. Nosil. 2012. The genomics of Catchen, J. M., A. Amores, P. A. Hohenlohe, W. Cresko, and speciation-with-gene-ﬂow. Trends Genet. 28:342–350. J. H. Postlethwait. 2011. Stacks: building and genotyping Felsenstein, J. 1981. Skepticism towards santa rosalia, or why loci de novo from short-read sequences. G3: Genes - are there so few kinds of animals. Evolution 35:124–138. Genomes - Genetics 1:171–182. France, R. L. 1995a. Differentiation between littoral and Catchen, J., P. A. Hohenlohe, S. Bassham, A. Amores, and W. pelagic food webs in lakes using stable carbon isotopes. A. Cresko. 2013. Stacks: an analysis tool set for population Limnol. Oceanogr. 40:1310–1313. genomics. Mol. Ecol. 22:3124–3140. France, R. L. 1995b. C-13 enrichment in benthic compared to Colborne, S. F., S. R. Garner, F. K. Longstaffe, and B. D. Neff. planktonic algae – foodweb implications. Mar. Ecol. Prog. 2016. Assortative mating but no evidence of genetic Ser. 124:307–312. 5354 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Franchini, P., C. Fruciano, M. L. Spreitzer, J. C. Jones, K. R. Kerschbaumer, M., L. Postl, M. Koch, T. Wiedl, and C. Elmer, F. Henning, et al. 2014. Genomic architecture of Sturmbauer. 2011. Morphological distinctness despite large- ecologically divergent body shape in a pair of sympatric scale phenotypic plasticity-analysis of wild and pond-bred crater lake cichlid ﬁshes. Mol. Ecol. 23:1828–1845. juveniles of allopatric populations of Tropheus moorii. Franssen, N. R., L. K. Stewart, and J. F. Schaefer. 2013. Naturwissenschaften 98:125–134. Morphological divergence and ﬂow-induced phenotypic Kiljunen, M., J. Grey, T. Sinisalo, C. Harrod, H. Immonen, plasticity in a native ﬁsh from anthropogenically altered and R. I. Jones. 2006. A revised model for lipid-normalizing stream habitats. Ecol. Evol. 3:4648–4657. delta C-13 values from aquatic organisms, with implications Gavrilets, S. 2003. Perspective: models of speciation: what have for isotope mixing models. J. Appl. Ecol. 43:1213–1222. we learned in 40 years? Evolution 57:2197–2215. Kolbe, J. J., M. Leal, T. W. Schoener, D. A. Spiller, and J. B. Gavrilets, S. 2004. Fitness landscapes and the origin of species. Losos. 2012. Founder effects persist despite adaptive Princeton Univ. Press, Princeton, NJ. differentiation: a ﬁeld experiment with lizards. Science Gavrilets, S. 2005. “Adaptive speciation” – it is not that easy: a 335:1086–1089. reply to Doebeli et al.. Evolution 59:696–699. Kusche, H., H. Recknagel, K. R. Elmer, and A. Meyer. 2014. Gavrilets, S., A. Vose, M. Barluenga, W. Salzburger, and A. Crater lake cichlids individually specialize along the benthic- Meyer. 2007. Case studies and mathematical models of limnetic axis. Ecol. Evol. 4:1127–1139. ecological speciation. 1. Cichlids in a crater lake. Mol. Ecol. Lescak, E. A., S. L. Bassham, J. Catchen, O. Gelmond, M. L. 16:2893–2909. Sherbick, F. A. von Hippel, et al. 2015. Evolution of Guo, B. C., F. J. J. Chain, E. Bornberg-Bauer, E. H. Leder, and stickleback in 50 years on earthquake-uplifted islands. Proc. J. Merila.2013. Genomic divergence between nine- and Natl Acad. Sci. USA 112:E7204–E7212. three-spined sticklebacks. BMC Genom. 14:756. Li, H., and R. Durbin. 2009. Fast and accurate short read Gutenkunst, R. N., R. D. Hernandez, S. H. Williamson, and C. alignment with Burrows-Wheeler transform. Bioinformatics D. Bustamante. 2009. Inferring the joint demographic 25:1754–1760. history of multiple populations from multidimensional SNP Machado-Schiafﬁno, G., F. Henning, and A. Meyer. 2014. frequency data. PLoS Genet. 5:e1000695. Species-speciﬁc differences in adaptive phenotypic plasticity Hecky, R. E., and R. H. Hesslein. 1995. Contributions of in an ecologically relevant trophic trait: hypertrophic lips in benthic algae to lake food webs as revealed by stable isotope Midas cichlid ﬁshes. Evolution 68:2086–2091. analysis. J. N. Am. Benthol. Soc. 14:631–653. Machado-Schiafﬁno, G., A. F. Kautt, H. Kusche, and A. Meyer. Henning, F., J. C. Jones, P. Franchini, and A. Meyer. 2013. 2015. Parallel evolution in Ugandan crater lakes: repeated Transcriptomics of morphological color change in evolution of limnetic body shapes in haplochromine cichlid polychromatic Midas cichlids. BMC Genom. 14:171. ﬁsh. BMC Evol. Biol. 15:9. Hulsey, C. D., R. J. Roberts, Y. H. E. Loh, M. F. Rupp, and J. Malinsky, M., R. J. Challis, A. M. Tyers, S. Schiffels, Y. Terai, T. Streelman. 2013. Lake Malawi cichlid evolution along a B. P. Ngatunga, et al. 2015. Genomic islands of speciation benthic/limnetic axis. Ecol. Evol. 3:2262–2272. separate cichlid ecomorphs in an East African crater lake. Johansson, F., and J. Andersson. 2009. Scared ﬁsh get lazy, and Science 350:1493–1498. lazy ﬁsh get fat. J. Anim. Ecol. 78:772–777. Manousaki, T., P. M. Hull, H. Kusche, G. Machado-Schiafﬁno, Kautt, A. F., K. R. Elmer, and A. Meyer. 2012. Genomic signatures P. Franchini, C. Harrod, et al. 2013. Parsing parallel of divergent selection and speciation patterns in a natural evolution: ecological divergence and differential experiment’, the young parallel radiations of Nicaraguan crater gene expression in the adaptive radiations of thick- lake cichlid ﬁshes. Mol. Ecol. 21:4770–4786. lipped Midas cichlid ﬁshes from Nicaragua. Mol. Ecol. Kautt, A. F., G. Machado-Schiafﬁno, and A. Meyer. 2016. 22:650–669. Multispecies outcomes of sympatric speciation after Martin, C. H. 2013. Strong assortative mating by diet, color, admixture with the source population in two radiations of size, and morphology but limited progress toward sympatric nicaraguan crater lake cichlids. PLoS Genet. 12:e1006157. speciation in a classic example: Cameroon crater lake doi:10.1371/journal.pgen.1006157 cichlids. Evolution 67:2114–2123. Kavembe, G. D., A. F. Kautt, G. Machado-Schiafﬁno, and A. Martin, R. A., and D. W. Pfennig. 2009. Disruptive Meyer. 2016. Eco-morphological differentiation in Lake selection in natural populations: the roles of ecological Magadi tilapia, an extremophile cichlid ﬁsh living in hot, specialization and resource competition. Am. Nat. alkaline and hypersaline lakes in East Africa. Mol. Ecol. 174:268–281. 25:1610–1625. Martin, C. H., J. S. Cutler, J. P. Friel, C. D. Touokong, G. Kellogg, K. A., J. A. Markert, J. R. Stauffer, and T. D. Kocher. Coop, and P. C. Wainwright. 2015. Complex histories of 1995. Microsatellite variation demonstrates multiple repeated gene ﬂow in Cameroon crater lake cichlids cast paternity in Lekking Cichlid ﬁshes from Lake Malawi, doubt on one of the clearest examples of sympatric Africa. Proc. R. Soc. B 260:79–84. speciation. Evolution 69:1406–1422. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5355 Incipient Sympatric Speciation in Lake As. Managua? A. F. Kautt et al. Matessi, C., A. Gimelfarb, and S. Gavrilets. 2001. Long-term whole-genome association and population-based linkage buildup of reproductive isolation promoted by disruptive analyses. Am. J. Hum. Genet. 81:559–575. selection: how far does it go? Selection 2:41–64. R Development Core Team. 2014. R: a Language and Meyer, A. 1987. Phenotypic plasticity and heterochrony in Environment for Statistical Computing. Available at http:// Cichlasoma managuense (Pisces, Cichlidae) and their www.R-project.org. (accessed 20 March 2016). implications for speciation in cichlid ﬁshes. Evolution Ravigne, V., U. Dieckmann, and I. Olivieri. 2009. Live where 41:1357–1369. you thrive: joint evolution of habitat choice and local Meyer, A., and A. F. Kautt. 2014. Sympatric speciation. Oxford adaptation facilitates specialization and promotes diversity. bibliographies in evolutionary biology. DOI:10.1093/obo/ Am. Nat. 174:E141–E169. 9780199941728-0053. Recknagel, H., K. R. Elmer, and A. Meyer. 2013a. A hybrid Minagawa, M., and E. Wada. 1984. Stepwise enrichment of genetic linkage map of two ecologically and morphologically N-15 along food-chains – further evidence and the relation divergent Midas cichlid ﬁshes (Amphilophus spp.) obtained between delta-N-15 and animal age. Geochim. Cosmochim. by massively parallel DNA sequencing (ddRADSeq). G3: Genes - Genomes - Genetics 3:65–74. Acta 48:1135–1140. Noack, K., A. B. Wilson, and A. Meyer. 2000. Broad Recknagel, H., H. Kusche, K. R. Elmer, and A. Meyer. 2013b. taxonomic applicability of microsatellites developed for the Two new endemic species in the Midas cichlid species complex from Nicaraguan crater lakes: Amphilophus tolteca highly polymorphic neotropical cichlid, Amphilophus citrinellum. Anim. Genet. 31:151–152. and Amphilophus viridis (Perciformes, Cichlidae). Aqua Nosil, P. 2012. Ecological speciation. Oxford Univ. Press, New 19:207–224. York. Recknagel, H., K. R. Elmer, and A. Meyer. 2014. Crater lake habitat predicts morphological diversity in adaptive Nosil, P., and D. Schluter. 2011. The genes underlying the process of speciation. Trends Ecol. Evol. 26:160–167. radiations of cichlid Fishes. Evolution 68:2145–2155. Nosil, P., D. J. Funk, and D. Ortiz-Barrientos. 2009a. Rice, W. R. 1987. Speciation via habitat specialization: the evolution of reproductive isolation as a correlated character. Divergent selection and heterogeneous genomic divergence. Mol. Ecol. 18:375–402. Evol. Ecol. 1:301–314. Nosil, P., L. J. Harmon, and O. Seehausen. 2009b. Ecological Rieseberg, L. H. 2001. Chromosomal rearrangements and explanations for (incomplete) speciation. Trends Ecol. Evol. speciation. Trends Ecol. Evol. 16:351–358. 24:145–156. Robinson, B. W., and K. J. Parsons. 2002. Changing times, Pardo, N., D. R. Avellan, J. L. Macias, T. Scolamacchia, and spaces, and faces: tests and implications of adaptive D. Rodriguez. 2008. The ~1245 yr BP Asososca maar: new morphological plasticity in the ﬁshes of northern postglacial lakes. Can. J. Fish Aquat. Sci. 59:1819–1833. advances on recent volcanic stratigraphy of Managua (Nicaragua) and hazard implications. J. Volcanol. Geoth. Robinson, B. W., and D. S. Wilson. 1994. Character release Res. 176: and displacement in ﬁshes – a neglected literature. Am. Nat. 493–512. 144:596–627. Peres-Neto, P. R., and P. Magnan. 2004. The inﬂuence of Salzburger, W., I. Braasch, and A. Meyer. 2007. Adaptive swimming demand on phenotypic plasticity and sequence evolution in a color gene involved in the morphological integration: a comparison of two formation of the characteristic egg-dummies of male haplochromine cichlid ﬁshes. BMC Biol. 5:51. polymorphic charr species. Oecologia 140:36–45. Peterson, B. K., J. N. Weber, E. H. Kay, H. S. Fisher, and H. Sanetra, M., F. Henning, S. Fukamachi, and A. Meyer. 2009. E. Hoekstra. 2012. Double digest RADseq: an inexpensive A microsatellite-based genetic linkage map of the cichlid method for de novo SNP discovery and genotyping in ﬁsh, Astatotilapia burtoni (Teleostei): a comparison of model and non-model species. PLoS One 7:e37135. genomic architectures among rapidly speciating cichlids. Poelstra, J. W., N. Vijay, C. M. Bossu, H. Lantz, B. Ryll, I. Genetics 182:387–397. Mueller, et al. 2014. The genomic landscape underlying Schluter, D. 1996. Ecological speciation in postglacial ﬁshes. phenotypic integrity in the face of gene ﬂow in crows. Philos. Trans. R. Soc. Lond. B Biol. Sci. 351:807–814. Science 344:1410–1414. Schluter, D., and J. D. McPhail. 1992. Ecological character Post, D. M. 2002. Using stable isotopes to estimate trophic displacement and speciation in sticklebacks. Am. Nat. position: models, methods, and assumptions. Ecology 140:85–108. 83:703–718. Servedio, M. R., G. S. Van Doorn, M. Kopp, A. M. Frame, Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference and P. Nosil. 2011. Magic traits in speciation: ‘magic’ but of population structure using multilocus genotype data. not rare? Trends Ecol. Evol. 26:389–397. Genetics 155:945–959. Smadja, C. M., and R. K. Butlin. 2011. A framework for Purcell, S., B. Neale, K. Todd-Brown, L. Thomas, M. A. R. comparing processes of speciation in the presence of gene Ferreira, D. Bender, et al. 2007. PLINK: a tool set for ﬂow. Mol. Ecol. 20:5123–5140. 5356 ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. A. F. Kautt et al. Incipient Sympatric Speciation in Lake As. Managua? Thibert-Plante, X., and A. P. Hendry. 2011. The consequences Wu, C. I. 2001. The genic view of the process of speciation. J. of phenotypic plasticity for ecological speciation. J. Evol. Evol. Biol. 14:851–865. Biol. 24:326–342. Zanden, M. J. V., and J. B. Rasmussen. 1999. Primary Trickett, A. J., and R. K. Butlin. 1994. Recombination consumer delta C-13 and delta N-15 and the trophic suppressors and the evolution of new species. Heredity position of aquatic consumers. Ecology 80:1395–1404. 73:339–345. Zardoya, R., D. M. Vollmer, Craddock, J. T. Streelman, S. Udovic, D. 1980. Frequency-dependent selection, disruptive Karl, and A. Meyer. 1996. Evolutionary conservation of selection, and the evolution of reproductive isolation. Am. microsatellite ﬂanking regions and their use in resolving the Nat. 116:621–641. phylogeny of cichlid ﬁshes (Pisces: Perciformes). Proc. R. Van Oosterhout, C., W. F. Hutchinson, D. P. M. Willis, and Soc. B 263:1589–1598. P. Shipley. 2004. Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol. Supporting Information Ecol. Notes 4:535–538. Via, S. 1999. Reproductive isolation between sympatric races Additional Supporting Information may be found online of pea aphids. I. Gene ﬂow restriction and habitat choice. in the supporting information tab for this article: Evolution 53:1446–1457. Figure S1. Experimental setup of the phenotypic plasticity Via, S. 2009. Natural selection in action during speciation. experiment. Proc. Natl Acad. Sci. USA 106:9939–9946. Figure S2. Schematic illustration of all 13 tested demo- Webb, P. W. 1984. Body form, locomotion and foraging in graphic models. aquatic vertebrates. Am. Zool. 24:107–120. Table S1. Morphometric measurements, stable isotope Webb, P. W. 1988. Simple physical principles and vertebrate data, and microsatellite genotypes. aquatic locomotion. Am. Zool. 28:709–725. Table S2. Support for tested demographic models. Wigginton, J. E., D. J. Cutler, and G. R. Abecasis. 2005. A note Data set S1. RADseq genotype data in variant call for- on exact tests of Hardy-Weinberg equilibrium. Am. J. Hum. mat. Genet. 76:887–893. ª 2016 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 5357

Journal

Ecology and Evolution – Wiley

Published: Aug 1, 2016

Keywords: ; ; ; ; ;

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

Incipient sympatric speciation in Midas cichlid fish from the youngest and one of the smallest crater lakes in Nicaragua due to differential use of the benthic and limnetic habitats?

References (112)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies