Moderate Population Structure in Drosophila sturtevanti from the South American Atlantic Forest Biome

© 2021 Academia Sinica, Taiwan Open Access Moderate Population Structure in Drosophila sturtevanti from the South American Atlantic Forest Biome Bruna Memari Trava1, Rogério Pincela Mateus2, Luciana Paes de Barros Machado2, and Lilian Madi-Ravazzi1,* 1Instituto de Biociências, Letras e Ciências Exatas, Universidade Estadual Paulista “Júlio de Mesquita Filho” (UNESP), Cristóvão Colombo Street, 2265 – Jardim Nazareth, São José do Rio Preto, SP 15054-000, Brazil. *Correspondence: E-mail: [email protected] (Madi-Ravazzi). Tel: +55 17 3221-2389. Fax: +55 17 3221-2390. E-mail: [email protected] (Trava) 2Universidade Estadual do Centro-Oeste (UNICENTRO), Campus CEDETEG, Élio Antonio Dalla Vecchia Alley, 838, Vila Carli, Guarapuava, PR 85040-167, Brazil. E-mail: [email protected] (Mateus); [email protected] (Machado) Received 30 May 2020 / Accepted 25 May 2021 / Published 17 August 2021 Communicated by Chung-Pin Lin Drosophila sturtevanti is a widely distributed Neotropical species. In South America, it is abundant and adapted to different phytophysiognomies of the Atlantic Forest biome. Reproductive, chromosomal and enzymatic studies have indicated the existence of a differentiation among D. sturtevanti populations. In this work, the level of genetic diversity and the population genetic structure were analyzed using four population groupings. One hundred and twenty-six D. sturtevanti males collected from nine forest fragments were analyzed for 11 species-specific microsatellite loci. A total of 109 alleles, ranging from 2 to 16 alleles per locus, were detected. The highest mean observed heterozygosity - HO was estimated in samples from the largest collection areas, and the lowest HO was from a population where fire events are common. A low molecular variation, around 3% among populations and negative among groups, an absence of genetic and geographic correlations and a moderate genetic differentiation - FST = 0.0663 - indicated that D. sturtevanti is not strongly structured. Besides no overall genetic and geographic distance correlation, the pair of closest geographically populations Matão and Nova Granada showed the lower differentiation through FST, DC and a Neighbor Joining tree. Ribeirão da Ilha - RDI, an isolated insular population, was the most differentiated according to FST, DC and a cluster-based Bayesian analysis. The isolation of RDI that resulted in significant divergence could be ancient, because of sea level regressions/ transgressions, or more recently via founder effect/genetic drift by anthropic action carrying D. sturtevanti hosts from continent to island. This work is important for understanding the genetic variability distribution of a Neotropical forest-dwelling Drosophila species using for the first time, a wide population distribution approach. Key words: Saltans group, Forest fragmentation, Genetic diversity, Simple sequence repeats, Population differentiation. BACKGROUND The Brazilian Atlantic Forest biome is formed by a complex set of vegetation types, or phytophysiognomies, which are distributed over approximately 1.3 million km2, over 17 states, and along the coast and interior of southern Brazil (Galindo-Leal and Câmara 2003). This biome is one of the most important in terms of endemism and biodiversity in the world; however, it is also the most threatened (Mittermeier et al. 2004), and Citation: Trava BM, Mateus RP, de Barros Machado LP, Madi-Ravazzi L. 2021. Moderate population structure in Drosophila sturtevanti from the South American Atlantic Forest biome. Zool Stud 60:46. doi:10.6620/ZS.2021.60-46. Zoological Studies 60:46 (2021) doi:10.6620/ZS.2021.60-46 1 © 2021 Academia Sinica, Taiwan only between 7 and 11% of its original area currently remains (Salgueiro et al. 2004; Tabarelli et al. 2005). The different phytophysiognomies have very particular climates—for example, the dense ombrophilous forest is hot and humid; the seasonal forest (semideciduous and deciduous) has hot and humid summers, with cold and dry winters; and the mixed ombrophilous forest has a humid and cold climate (Marques 2016). It is known that South American biomes, especially the Atlantic Forest, contain much of the global biodiversity. However, the anthropogenic environmental impacts in the last 100 years are probably the most severe in human history. Deforestation is the dominant land-use trend in Latin America (Ramankutty and Foley 1999; Achard et al. 2002), and several biomes, including the Brazilian Cerrado and the Amazon and Atlantic rain forests, have been and continue to be severely affected by conversion to agriculture and pastures (Galindo-Leal and Câmara 2003; Klink and Machado 2005). This deforestation process is mainly due to traditional and export-oriented industrial agriculture (soybean production, for example) and cattle ranching (Hecht 1993; Grau et al. 2008), which indirectly have favored other forms of degradation, such as logging and fire (Nepstad et al. 1999). Thus, it becomes crucial to perform studies on the diversity of organisms that remain living in such threatened areas to gather information about the impact of these events on the conservation status of South American biodiversity. In agreement with this concept, the present study investigated the genetic diversity and structure of D. sturtevanti populations that were collected in fragments of the remaining Atlantic Forest in South America. Drosophila sturtevanti (saltans group, sturtevanti subgroup) is a Neotropical species that occurs in the Atlantic Forest biome. It is a forest-dwelling species found in fragments of this biome and has a broad geographic distribution across almost the entire range of the saltans group, from Mexico to southern Brazil, including the Caribbean islands. This species is a generalist and has a high abundance in all seasons (Magalhães 1962; Tidon-Sklorz and Sene 1992; Torres and Madi-Ravazzi 2006; Valadão et al. 2019), and some of its elements have been studied previously. For example, its populations present relatively high esterase polymorphism and polytene chromosome inversion frequencies among the species of the saltans group (Hosaki-Kobayashi and Bicudo 1997; Bernardo and Bicudo 2009). Studies on the reproductive isolation of the populations of this species are inconsistent. For example, incipient sexual isolation among the D. sturtevanti populations was detected (Dobzhansky 1944; Carareto 1994), but experiments using laboratoryraised and freshly collected strains from Mexico, Central America and Brazil did not find the same results (Hosaki-Kobayashi and Bicudo 1994). Nonetheless, morphological data analysis (aedeagus and wing morphometry) indicated that there are three different genetic sets of D. sturtevanti Brazilian populations: those from the Northeast, Southeast and South Regions (Segala 2019). This genetic differentiation, along with the wide distribution of D. sturtevanti, makes this species an important model for studies of the association between habitat fragmentation and the distribution of genetic variability. Specimens in fragmented habitats, such as the D. sturtevanti in the Atlantic Forest, are frequently isolated and/or experience reduction in population size, triggering bottlenecks and genetic drift effects (e.g., Goodman et al. 2001; Pascual et al. 2001; Broeck et al. 2017). The population isolation of Drosophila species, which have low dispersion (Wallace 1966; Spencer and Heed 1975; Markow and Castrezana 2000), could result in limited gene flow, increased interpopulation genetic differentiation, and reduced intrapopulation variability and effective population sizes. These events could, over time, make the population less resilient to environmental changes (Franknam 1997; Templeton et al. 2001). Population genetics studies are very helpful for understanding the history of a species, and most investigations of genetic variation have concentrated mainly on the (nearly) neutral genetic markers as gauges of the contemporary and historical processes affecting the maintenance and distribution of genetic variation. Therefore, in this work, for the first time, a population study of D. sturtevanti was performed using 11 species-specific microsatellite markers. The objectives were to evaluate the populations’ levels of genetic diversity, to verify the existence of population structure and to investigate whether genetic differentiation is associated with the fragmentation of the Atlantic Forest. Thus, three different groupings were analyzed: 1 - populations of the three geographic regions of Brazil (northeast, southeast and south), as suggested by previous morphological analyses; 2 - populations of two phytophysiognomies (dense ombrophilous forest and semideciduous seasonal forest); and 3 - a putative grouping from a Bayesian analysis. MATERIALS AND METHODS Drosophila sturtevanti samples A total of 126 individuals from nine populations of D. sturtevanti were collected from two phytophysiognomy types of the Atlantic Forest of three Brazilian geographic regions (Table 1, Fig. 1). page 2 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan The collections were performed with the permission of the Instituto Chico Mendes de Conservação da Biodiversidade – ICMBio, a regulatory agency responsible for environmental studies in Brazil (permission number 46752). The representative specimens were stored in 100% ethanol in the collection of the Museu de Zoologia da Universidade de São Paulo, Brazil. All flies were captured using closed traps containing fermented banana bait (Penariol et al. 2008). The identification of males was carried out by analyzing the aedeagus using specific identification keys (FreireMaia and Pavan 1949; Mourão and Bicudo 1967; Vilela Fig. 1. Map of the Brazilian Drosophila sturtevanti populations from which samples were collected and analyzed. Table 1. Geographic regions, phytophysiognomies and sizes of collection areas within the Brazilian Atlantic Forest biome of Drosophila sturtevanti Geographic region Collection site (Abbreviation) Phytophysiognomies Area size (ha) South region Aguaí (AGU) DOF 7,652 Ribeirão da Ilha (RDI) DOF 10,000 Piraí (PIR) DOF 982.6 Southeast region Serra da Cantareira (SDC) DOF 64,800 Picinguaba (PCG) DOF 47,500 Matão (MAT) SSF 2,189 Nova Granada (NGR) SSF 1,359 Northeast region Pratigi (PRA) DOF 32,000 Guaribas (GUA) SSF 4,051 DOF = Dense Ombrophilous Forest; SSF = Semidecidual Seasonal Forest. N page 3 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan and Bächli 1990; Souza et al. 2014). Samples were preserved in 96% ethanol and kept at 4°C until DNA extraction. DNA extraction and microsatellite analyses The extraction of genomic DNA was performed by individual maceration of each sampled male using the Promega kit. Eleven microsatellites (Dsturt_B, Dsturt_D, Dsturt_E, Dsturt_G, Dsturt_I, Dsturt_J, Dsturt_K, Dsturt_L, Dsturt_M, Dsturt_N and Dsturt_O) described for D. sturtevanti (Roman, B.E., Trava, B. M.; Madi-Ravazzi, L., submitted) were amplified through polymerase chain reaction (PCR) in a total volume of 25 µl, containing 0.5 µl of Taq DNA polymerase, 2.5 µl of 10x buffer, 0.2 mM dNTP, 0.2 pmol of each primer, 1.5 mM MgCl2 and 3 ng of DNA. Touchdown PCR was performed for the Dsturt_B, Dsturt_D, Dsturt_I, Dsturt_K and Dsturt_M loci as follows: denaturation cycle at 94°C for 2 minutes; 2 repetitions of 10 cycles at 94°C for 1 minute, 65°C for 1 minute (-1°C per cycle), and 72°C for 2 minutes; ending with 18 cycles at 94°C for 1 minute, 55°C for 1 minute and 72°C for 5 minutes. For all other loci, a specific primer annealing temperature (Ta) was applied: 53°C - Dsturt_ N; 55°C - Dsturt_G; 56°C - Dsturt_E and Dsturt_L; 57°C - Dsturt_O; 65°C - Dsturt_J. The PCR conditions for these loci were as follows: 94°C for 2 minutes; 30 cycles of 94°C for 1 minute, specific Ta for 1 minute and 72°C for 2 minutes; and ending with 72°C for 5 minutes. PCR amplification products were visualized in 6% polyacrylamide gel and stained with 15% silver nitrate (Sanguinetti et al. 1994). Statistical analysis Population genetic structure and diversities were assessed using 10–15 individuals from each of the nine populations. The mean number of alleles (Na), effective number of alleles (Ne), expected heterozygosity (HE), observed heterozygosity (HO), number of private alleles (Np), frequency of private alleles (Ap) and fixation index (F) were calculated for each population in GenAlEx software v.6.51b2 (Peakall and Smouse 2006 2012). The allelic polymorphic information content (PIC) was obtained using CERVUS software v.3.0.7 (Kalinowski et al. 2007). The mean allelic richness (Ar)—which is an unbiased measure of the number of alleles estimated independently of the sample size, allowing for comparison between different sample sizes (El Mousadik and Petit 1996)—was calculated in FSTAT v.2.93 (Goudet 2001). Departures from Hardy-Weinberg equilibrium (HWE) at each locus within populations were estimated in the GenAlEx software. All levels of significance were determined after a sequential Bonferroni correction for multiple tests (Holm 1979). The FreeNA software (Chapuis and Estoup 2007) was used to estimate null allele frequencies (An) for each locus and population following the Expectation Maximization (EM) algorithm of Dempster et al. (1977). A population bottleneck test, using adjusted frequencies for the presence of null alleles, was performed in BOTTLENECK software (Cornuet and Luikart 1996; Piry et al. 1999) to test the occurrence of recent demographic events. The program BOTTLENECK computed the distribution of the expected heterozygosity from the observed number of alleles when given the sample size under the assumption of mutation-drift equilibrium. The simulation of the coalescent process of n genes was performed under the two-phase model (TPM), using the Wilcoxon signed-rank test, setting the parameters as 90% single-step mutations, 10% multiplestep mutations, a variance of the geometric distribution of 12, and 1,000 iterations. These settings correspond to sensible parameter values for most microsatellites, considering that fewer than 20 loci were used (Cornuet and Luikart 1996; Piry et al. 1999). The level of genetic differentiation among populations was verified using multiple approaches: indexes of genetic differentiation, a neighbor-joining tree, the use of a nonspatial Bayesian algorithm and analyses of molecular variance (AMOVA). The FreeNA software was used to estimate two genetic differentiation indexes based on the ENA method (Chapuis and Estoup 2007): DC - Cavalli-Sforza and Edwards (1967) genetic distances, and FST (Weir 1996). For these indexes 95% confidence intervals (C.I.) were obtained using bootstrap resampling over loci. Wright (1978) qualitative classification of the genetic differentiation among populations (spatial analyses) was applied accordingly to the FST values obtained: ‘low’ (0–0.05), ‘moderate’ (0.05–0.15), ‘high’ (0.15–0.25) and ‘very high’ (> 0.25). The correlations between the genetic differentiation above (DC and FST) and geographic distance by the Mantel tests (with 9999 random permutations) were performed in GenAlEx software. A neighbor-joining tree (Saitou and Nei 1987) was obtained through maximum likelihood analysis (Felsenstein 1981) on adjusted allele frequencies (considering the presence of null alleles) of microsatellite data using PHYLIP software (version 3.7a; Felsenstein 2009). To evaluate support for the branches, a bootstrap analysis (Felsenstein 1985) was performed 1000 times. An unrooted tree was constructed using Cavalli-Sforza and Edwards (1967) distances, using a branch and bound algorithm, the majority rule option and with random addition of populations. The cluster-based Bayesian method was performed with the software STRUCTURE v.2.3.4 (Pritchard et al. 2000), page 4 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan considering the presence of null alleles and the Dsturt_I locus as X or Y linked in the genotypic matrix. For this analysis, the admixture hypothesis was used assuming the existence of correlated allele frequencies, in which each sample is partially composed of the genome of each ancestral population. Together with the allele frequency model, they allow the log likelihood L (K) for the data to be obtained. This model is considered to be the most appropriate when the a priori origin and the degree of isolation of the studied populations are unknown (Pritchard et al. 2000). The prior probability, i.e., the probability that an individual belongs to any reference K population, is defined as l/K. The K value was fixed from 1 to 11 using 10,000 burn-in, 500,000 Markov chain Monte Carlo (MCMC) replicates after burn-in and 25 iterations. After obtaining the results, a bar graph was generated with the CLUMPAK tool (Kopelman et al. 2015) using the best number of clusters (K) obtained by the STRUCTURE HARVESTER tool (Earl and Von Holdt 2012), according to the Evanno test (Evanno et al. 2005; Earl and Von Holdt 2012). The AMOVA was run in the Arlequin software (version 3.5.2.2, Excoffier and Lischer 2010), using the adjusted genetic frequencies and 1,000 permutations, to test four assumptions about the distribution of genetic variability of Brazilian D. sturtevanti populations: 1) with no grouping; 2) grouping according to geographic regions (Northeast, Southeast and South regions); 3) grouping according to Atlantic Forest phytophysiognomies (dense ombrophilous forest and semideciduous seasonal forest); and 4) groupings according to the result of the cluster-based Bayesian method performed with STRUCTURE software. The second grouping was proposed based on a previous morphological study (Segala 2019) using D. sturtevanti, (Segala 2019) using D. sturtevanti and also to test correlation between genetic variability distribution and isolation by distance. The third grouping was tested based on phylogeography and microsatellite studies with D. ornatifrons (Gustani et al. 2015; Zorzato 2015), another Neotropical forest dweeling species, which suggested that there is correlation between genetic variability and Atlantic Forest phytophysionomies. In the absense of a genetic structure, slightly negative variation can be obtained only by chance, because the true value estimated is zero (Huang et al. 2021). RESULTS Genetic diversity analysis The variability analysis of 11 microsatellite loci specific to D. sturtevanti, using 126 males from nine populations, resulted in a total of 109 alleles (Table S1). The number of alleles per locus ranged from 2 (Dsturt_ B) to 16 (Dsturt_N), with an average (± standard error – S.E.) of 9.91 ± 1.30 (Table S2). Almost all loci were polymorphic, except Dsturt_B and Dsturt_D. These loci presented the lowest total numbers of alleles (2 and 3, respectively), were monomorphic in all populations except PIR (Table S1), and showed the lowest genetic diversity values (Supplementary Table S2). The Dsturt_I locus was the only one that did not show heterozygotes (HO = 0 and F = 1.00, Table S2) despite a total of 103 males were analyzed. This indicates that this locus is most likely Xor Y-linked. The other microsatellite loci were assumed to be autosomal because all population samples showed heterozygotes. In table S2, all the data for each locus concerning the total and mean number of alleles, polymorphism information content, mean effective number of alleles, mean observed heterozygosity, mean expected heterozygosity, fixation index and null allele frequency were reported. Among the polymorphic loci, PIC ranged from 0.815 (Dsturt_J) to 0.907 (Dsturt_N), indicating that these loci are informative for population analysis. The mean HO (± S.E.) across loci ranged from 0.15 ± 0.04 (Dsturt_N) to 0.68 ± 0.07 (Dsturt_O), while HE ranged from 0.73 ± 0.03 (Dsturt_E and Dsturt_J) to 0.81 ± 0.02 (Dsturt_N). Null allele frequencies ranged from 0.0774 (Dsturt_O) to 0.3682 (Dsturt_N). The Fixation index (F) across polymorphic loci resulted in positive values, with the Dsturt_O locus showing the lowest (0.12 ± 0.08) and the Dsturt_N locus showing the highest value (0.82 ± 0.05). Across populations, the total mean HO was 0.43 ± 0.03, ranging from 0.22 ± 0.06 (GUA) to 0.39 (± 0.08 in PRA and ± 0.09 in SDC) (Table 2). Among polymorphic loci, Dsturt_O was the only locus that did not show significant deviation from Hardy-Weinberg Equilibrium (HWE) in all samples, and Dsturt_N showed deviations in all populations. The other seven polymorphic loci showed significant deviations in at least four populations, and all populations presented heterozygote deficiency. F values ranged from 0.33 ± 0.11 in SDC to 0.65 ± 0.08 in GUA, with a high overall mean (0.44 ± 0.04; Table S2). AGU and MAT were the populations with the highest number of loci with HWE deviations (7 - Dsturt_E, Dsturt_G, Dsturt_I, Dsturt_J, Dsturt_L, Dsturt_M and Dsturt_N), and PCG was the one with the lowest (4 – Dsturt_K, Dsturt_L, Dsturt_M and Dsturt_ N). Allelic richness ranged from 3.32 in RDI to 4.10 in PRA. The frequency of null alleles ranged from 0.14 ± 0.04 (PCG) to 0.22 ± 0.04 (GUA). All populations, except PCG, presented private alleles, ranging from 1 (AGU, RDI, SDC, NGR and GUA) to 3 (MAT), and frequencies from 0.03 (SDC and NGR) to 0.20 (RDI). page 5 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan The results from individual population analysis using the TPM model showed that none had recent bottleneck events, defined by significant heterozygote excess. Population Genetic Structure The Bayesian analysis of the population structure that was conducted using STRUCTURE software indicated the existence of six genetic clusters (K = 6). The interpretation of the distribution of this genetic variability allowed the identification of two groups of populations: RDI and the other seven populations (Fig. 2). These results were used to estimate the molecular variability distribution (AMOVA), summarized in table 3. When no grouping was tested, almost 97% of the variation occurred among individuals within populations, while only 3% of total variation was detected among populations. In the other three different sets of grouped populations (geographical regions, phytophysiognomies and cluster based Bayesian method obtained above), the variations among regions were negative, among population within regions were around 4%, and the variation within populations were all the highest. The DC and FST indexes showed RDI to be the most genetic differentiated population (Table 4). The lower significant differentiation was between MAT and NGR (DC = 0.3143; FST = 0.03). All pairwise DC and FST comparisons were significantly different from zero. The overall FST result showed moderate differentiation among populations (0.0674; 95% C.I. = 0.0533–0.0808). However, no correlations between the genetic (DC and FST) and geographic distances of the population pairs were obtained by the Mantel tests (r = 0.142, p = 0.284; r = -0.023, p = 0.569, respectively). The neighbor-joining tree showed only one node with higher bootstrap support (> 50%), grouping MAT and NGR. GUA and PCG were the most differentiated populations (Fig. 3). DISCUSSION The nine populations studied showed a high number of polymorphisms and similar allelic diversity for the analyzed polymorphic microsatellite loci. These results are consistent with those from other genetic markers for this species, including chromosomal inversions (Hosaki-Kobayashi and Bicudo 1997) and esterase loci (Bernardo and Bicudo 2009), which Table 2. Values of genetic diversity (± standard error) across nine populations of Drosophila sturtevanti from the Atlantic Forest biome in Brazil using 11 species-specific microsatellite loci. Na – mean number of alleles; Ne – mean effective number of alleles; Ar – mean allele richness; Np – number of private alleles; Ap – mean frequency of private alleles; HO – mean observed heterozygosity; HE – mean expected heterozygosity; An – mean frequency of null alleles; F – fixation index; HWE – list of loci with departure of Hardy-Weinberg expectations Populations NaNeArNpApHOHEAnFHWE AGU 5.09 ± 0.62 3.54 ± 0.44 3.60 1 0.07 0.27 ± 0.08 0.61 ± 0.09 0.20 ± 0.05 0.51 ± 0.12 E, G, I, J, L, M, N RDI 4.64 ± 0.88 3.40 ± 0.62 3.32 1 0.20 0.28 ± 0.09 0.56 ± 0.09 0.18 ± 0.04 0.54 ± 0.11 G, I, J, K, L, N PIR 4.64 ± 0.47 3.42 ± 0.40 3.58 2 0.10 0.34 ± 0.09 0.63 ± 0.07 0.18 ± 0.05 0.37 ± 0.13 I, J, K, M, N SDC 6.00 ± 0.80 4.15 ± 0.58 3.94 1 0.03 0.39 ± 0.09 0.65 ± 0.09 0.15 ± 0.05 0.33 ± 0.11 E, I, J, K, N PCG 5.54 ± 0.83 3.75 ± 0.56 3.64 0 - 0.37 ± 0.08 0.61 ± 0.09 0.14 ± 0.04 0.38 ± 0.09 K, L, M, N MAT 5.73 ± 0.62 4.04 ± 0.53 3.88 3 0.08 0.32 ± 0.08 0.65 ± 0.09 0.19 ± 0.05 0.41 ± 0.11 E, G, I, J, L, M, N NGR 6.09 ± 0.77 4.15 ± 0.58 3.94 1 0.03 0.31 ± 0.07 0.64 ± 0.09 0.19 ± 0.05 0.46 ± 0.10 E, I, L, M, N PRA 6.18 ± 0.71 4.37 ± 0.60 4.10 2 0.07 0.39 ± 0.08 0.66 ± 0.09 0.15 ± 0.04 0.34 ± 0.10 G, I, K, M, N GUA 4.91 ± 0.79 3.75 ± 0.62 3.68 1 0.15 0.22 ± 0.06 0.60 ± 0.09 0.22 ± 0.04 0.65 ± 0.08 E, J, K, L, N E = Dsturt_E; G = Dsturt_G; I = Dsturt_I; J = Dsturt_J; K = Dsturt_K; L = Dsturt_L; M = Dsturt_M; N = Dsturt_N. Fig. 2. Genetic structure analysis of nine Brazilian Drosophila sturtevanti populations using the Bayesian algorithm as implemented in the program STRUCTURE. Six clusters (K) were obtained (different colors). Population abbreviations are described in the Materials and Methods section. page 6 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan indicated that D. sturtevanti has more polymorphisms than any other species in the saltans group. The level of heterozygote deficiency in the D. sturtevanti populations was high. Among the events that could generate this result, such as the null alleles, the Wahlund effect, inbreeding or demographic processes (Shoemaker and Jaenike 1997; Moraes and Sene 2002; Hurtado et al. 2004; Markow and O’Grady 2008), our data allowed us to discuss the effect of the first, null allele. Their estimated frequencies were all above 0.14, higher than those detected in a Neotropical cactophilic species of Drosophila (Machado et al. 2010), to which Table 3. Analyses of Molecular Variance (AMOVA) for Drosophila sturtevanti populations, after correction for the presence of null alleles, without and assuming three different populational groupings (geographical regions, phytophysiognomies and after STRUCTURE analysis). DF = degrees of freedom; SS = sum of squares; VC = variance component; PV (%) = percentage of total variation Source DF SS VC PV (%) Without grouping Among populations 8 0.092 0.00025 3.01 Within populations 111 0.900 0.00811 96.99 Total 119 0.992 0.00836 100 Geographical regions Among regions 2 0.017 -0.00010 -1.17 Among populations within regions 6 0.074 0.00032 3.86 Within populations 111 0.900 0.00811 97.32 Total 119 0.992 0.00833 100 Phytophysiognomies Among regions 1 0.004 -0.00016 -1.95% Among populations within regions 7 0.088 0.00033 4.01% Within populations 111 0.900 0.00811 97.95% Total 119 0.992 0.00828 100% STRUCTURE Among regions 1 0.001 -0.00047 -5.88% Among populations within regions 7 0.090 0.00037 4.59% Within populations 111 0.900 0.00811 101.29% Total 119 0.992 0.00801 100% Table 4. Pairwise Cavalli-Sforza and Edwards (1967) genetic distances DC (above diagonal) and Weir (1996) FST (below diagonal) for Drosophila sturtevanti populations, after correction for the presence of null alleles. Shaded numbers represent the highest (above 0.45 for DC and 0.1 for FST), and numbers inside boxes represent the lowest values obtained AGU RDI PIR SDC PCG MAT NGR PRA GUA AGU *** 0.4506 0.4310 0.4231 0.4038 0.3687 0.3522 0.3887 0.4045 RDI 0.1178 *** 0.4842 0.4395 0.4458 0.4819 0.4321 0.4370 0.4315 PIR 0.0858 0.1104 *** 0.4094 0.4370 0.4801 0.4546 0.4442 0.5027 SDC 0.0729 0.0870 0.0419 *** 0.3665 0.3935 0.3778 0.3583 0.4295 PCG 0.0750 0.1119 0.0723 0.0523 *** 0.3651 0.3379 0.3402 0.3712 MAT 0.0416 0.1219 0.0810 0.0506 0.0494 *** 0.3143 0.3513 0.4539 NGR 0.0513 0.1129 0.0877 0.0582 0.0365 0.0300 *** 0.3393 0.3884 PRA 0.0432 0.0929 0.0648 0.0414 0.0319 0.0448 0.0421 *** 0.4276 GUA 0.0642 0.0907 0.0958 0.0608 0.0509 0.0756 0.0497 0.0427 *** page 7 of 12Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan no effect on the HWE or population differentiation was observed. However, a study of the impact of null alleles in the Maire yew, a tree from southern China (Wu et al. 2019), that depicts null allele frequencies similar to those obtained here (0.159–0.331) showed that null alleles can have significant effects on some population genetic parameters, especially HO, F and FST. Therefore, we considered that they could be one of the main generators of the high deficiency and should not be neglected. Microsatellite diversity detected in D. sturtevanti populations (HO = 0.43 ± 0.03) was higher than the mean diversity of the Brazilian populations of the endemic Drosophila species from the Atlantic Forest and open xerophytic areas (approximately 0.36 – Moraes and Sene 2007; Machado et al. 2010; Silva et al. 2015; Zorzato 2015). The highest HO values were detected in SDC and PRA, which correspond to two largest areas of collection. The first area, SDC, is located in the Paulistano Plateau and has the largest size (64,800 ha) and highest altitude (1,215 m) of the areas. The SDC region covers São Paulo, Guarulhos, Mairiporã and Caieiras municipalities, and has great ecological importance because it is the Biosphere Reserve of the Green Belt of the city of São Paulo, which comprises 7,916.52 hectares of the Atlantic Forest. The second area, PRA, is one of the Atlantic Forest conservation regions in the Pratigi Environmental Protection Area (EPA). This EPA is located on the southern coast of Bahia state, contains 85,686 ha and is an important federal conservation unit that protects the biodiversity and hydric resources associated with the Atlantic Forest in the region. The higher genetic diversity of these two populations may reflect the ecological and climatic characteristics of these regions, with high humidity favoring the occurrence of D. sturtevanti. The GUA population presented the lowest microsatellite diversity (HO = 0.22 ± 0.06) and higher distance in neighbor-joining tree. This population was collected in the Guaribas Biological Reserve, which is located on the border of the Caatinga and is one of the largest environmental preservation areas of Atlantic Forest (IBAMA 2003). However, several episodes of fire around and inside this reservation were detected, and four occurred in GUA within the fiveyear period of 2007–2012, (Alencar 2014). Certainly, these incidents could have had an impact on the local fauna, including D. sturtevanti, and they could be the cause of the observed low diversity in GUA. However, this impact was long enough ago to not be detected by the bottleneck analysis, which only identifies a recent event, when allelic diversity was still lower than the heterozygosity (Cornuet and Luikart 1996; Piry et al. 1999). The analysis of all population differentiation approaches resulted in some patterns that could be Fig. 3. Neighbor-joining tree obtained for the nine populations of Drosophila sturtevanti using Cavalli-Sforza and Edwards (1967) distances. Abbreviations are described in the Materials and Methods section. Node number refer to bootstrap values (only higher than 50% is shown). page 8 of 12 Zoological Studies 60:46 (2021) © 2021 Academia Sinica, Taiwan highlighted. The low molecular variation (3% among populations and negative among groups), the absence of genetic and geographic correlations (Mantel tests) and the moderate FST indicated that D. sturtevanti is not strongly structured, which is in disagreement with previous morphological data (Segala 2019) that distinguished South, Southeast and Northeast populations. This discrepancy is probably related to the nature of the markers while morphological characters should be under strong natural selection, microsatellite loci are considered nearly neutral. It could also be addressed that the microsatellite loci studied are not suffering hitchhiking effect of selection on genes that determine morphological characters analyzed. The genetic differentiation (FST and DC) and Bayesian analyses pointed out the higher isolation of RDI from the other populations. RDI is an insular population; thus, its differentiation could be explained by geographic isolation from the other continental populations and/or the consequences of the founder effect. In the first case, at the last glacial maximum (LGM) in the Pleistocene, the ocean level was between 80 and 130 meters lower than the current level (Suguio 2008), and RDI, which is today a separate island, could have been part of the continental forest at that time. The isolation may have arisen after the LGM, in the Holocene, as a consequence of sea level rise. Gustani et al. (2015) also showed that historical events involving sea level regressions/transgressions were important to the evolution of another Atlantic Forest Drosophila species, D. ornatifrons. Currently, gene flow could be restricted by continental/island isolation. The work of Tait et al. (2017), using microsatellites, on the recent invasion of D. suzuki in Italy is an example of higher restriction of gene flow among continental and insular populations. In the second case, considering the limited dispersion capacity of drosophilids (Spencer and Heed 1975; Markow and Castrezana 2000), some of the members of the continental D. sturtevanti population could have colonized the island by anthropic action, such as being carried by some hosts. The colonization of islands by Drosophila specimens via human activities is fairly common (Louis and David 1986; Lachaise and Silvain 2004; Jones 2005; Legrand et al. 2009 2011). For example, isolation by distance migration pattern of D. sechellia within the Seychelles archipelago is most likely explained by local genetic exchanges between neighboring populations due to the association of this species with the host plant Morinda citrifolia, which is used by humans medicinally and in fishing activities (Legrand et al. 2011). Another point to highlight is the higher genetic proximity, accordingly to the FST, DC and neighborjoining, between NGR and MAT, the populations from the interior of São Paulo state that have the smallest distance between them, 140 km. Considering the hypervariability inherent to the microsatellite markers, the detection of genetic proximity between populations indicates that different combinations of alleles arose in each generation and were sorted over geographical space, according to the isolation by distance model (Sunnucks 2000). The presence of ecological corridors could be allowing genetic interchange between these populations, maintaining their genetic similarity. Batista et al. (2018) studied six populations from the Southeast region of Brazil of D. mediopunctata, an almost exclusively forest-dwelling Neotropical species that belongs to the tripunctata group of the subgenus Drosophila (Vilela 1992; Yotoko et al. 2003; Hatadani et al. 2009; TaxoDros 2020) and detected low and moderate levels of population genetic structure using microsatellites and chromosome inversion polymorphisms, respectively. The authors proposed that the divergence among populations of this species might be a result of the association of climatic and geomorphological properties of the collection regions, rather than with forest fragmentation. In the case of the D. sturtevanti populations sampled in this work, considering the low dispersion ability of Drosophila (Spencer and Heed 1975; Markow and Castrezana 2000) and the fragmented and geographic isolated population distribution, the absence of a significant structure observed should be due to ancestral shared polymorphisms and/or recurrent mutations. CONCLUSIONS The characterization of Drosophila sturtevanti based on the analyses of species-specific microsatellites showed to be important as, for the first time, the genetic variability distribution of a Neotropical forestdwelling Drosophila species was studied using a wide population distribution approach. This work allowed the detection of moderate genetic differentiation among its populations, without general correlation with geographic and/or Atlantic Forest phytophysiognomy distributions. The weak genetic structure among fragmented and geographic isolated populations is likely to be a consequence of ancestral shared polymorphisms and/ or recurrent mutations; the similarity between MAT and NGR, geographically closest populations, might be the result of isolation by distance with gene flow provided by ecological corridors; and the higher differentiation of RDI, an insular population, could be due to its isolation after the last maximum glacial, or more recent founder effect via island colonization in association of hosts carried by anthropic action. A phylogeographical study page 9 of 12Zoological Studies 60:46 (2021)