Historical, Human, and Environmental Drivers of Genetic Diversity in the Red Swamp Crayfish (Procambarus Clarkii) Invading the Iberian Peninsula

i. Title Page 1 Article Title 2 Historical, human and environmental drivers of genetic diversity in the red swamp crayfish 3 (Procambarus clarkii) invading the Iberian Peninsula. 4 The full names of the authors 5 Lucía Acevedo-Limón1*#, Francisco J. Oficialdegui2#, Marta I. Sánchez3, Miguel Clavero1 6 Correspondence: 7 Lucía Acevedo-Limón. Estación Biológica de Doñana– CSIC. Avenida Américo Vespucio 26, 8 41092, Sevilla, Spain. Telephone: +34954232340. Fax: +34954621125 9 Email: *[email protected] 10 # These authors contributed equally to this work 11 The author's institutional affiliations where the work was carried out, with a footnote for the 12 author’s present address if different from where the work was carried out 13 1. Department of Conservation Biology, Doñana Biological Station (EBD-CSIC), Seville, Spain. 14 2. Department of Wetland Ecology, Doñana Biological Station (EBD-CSIC), Seville, Spain. 15 3. Department of Plant Biology and Ecology, Faculty of Biology, University of Seville, Seville, 16 Spain. 17 E-mails: 18 Francisco J. Oficialdegui; [email protected]; ORCID: 0000-0001-6223-736X 19 Marta I. Sánchez; [email protected]; ORCID: 0000-0002-8349-5410 20 Miguel Clavero; migue[email protected]; ORCID: 0000-0002-5186-0153 21 Keywords 22 environmental suitability; genetic structure; human-mediated dispersal; invasive species; 23 microsatellite; multiple introduction; 24 25 ii. Summary 26 1. Patterns of genetic diversity in invasive populations can be modulated by a range of factors 27 acting at different stages of the invasion process, including the genetic composition of the source 28 population(s), the introduction history (e.g. propagule pressure), the environmental suitability of 29 recipient areas and the features of secondary introductions. 30 2. The North-American red swamp crayfish, Procambarus clarkii, is one of the most widely 31 introduced freshwater species worldwide. It was legally introduced into Spain twice, near the 32 city of Badajoz in 1973 and in the Guadalquivir marshes in 1974. Thereafter the species rapidly 33 colonized almost the entire Iberian Peninsula. 34 3. We used seven nuclear microsatellites to describe the genetic diversity and structure of 28 35 locations distributed across the Iberian Peninsula and to explain the expansion process of the red 36 swamp crayfish. Additionally, we analysed the relationship between environmental suitability 37 and genetic diversity of the studied locations. 38 4. The red swamp crayfish had a clear spatial genetic structure in the Iberian Peninsula, probably 39 determined by the two independent introduction events in the 1970s, which produced two main 40 clusters separated spatially, one of which was dominant in Portugal and the other in Spain. 41 5. The human-mediated dispersal process seemed to have involved invasion hubs, hosting highly 42 genetically diverse areas and acting as sources for subsequent introductions. Genetic diversity 43 also tended to be higher in more suitable environments across the Iberian Peninsula. 44 6. Our results showed that the complex and human-mediated expansion of the red swamp crayfish 45 in the Iberian Peninsula has involved several longand short-distance movements and that both 46 ecological and anthropogenic factors have shaped the genetic diversity patterns resulting from 47 this invasion process. Early detection of potential invasion hubs may help to halt multiple short48 distance translocations and thus the rapid expansion of highly prolific invasive species over non49 native areas. 50 iii. Main text 51 Introduction 52 Biological invasions are one of the main threats to biodiversity globally (Bellard, Cassey & 53 Blackburn, 2016). The intensification of global trade and human movements, as well as the increase 54 of activities such as aquaculture, pet trade or gardening, have led to an acceleration of the global55 scale exchange of biota (Hulme et al., 2008; Ricciardi, 2007), which is blurring the traditionally 56 described biogeographical barriers (Capinha et al., 2015). The number of species introduced outside 57 their native ranges has been increasing in last decades and is expected to keep growing (Seebens et 58 al., 2017). Only a fraction of the introduced species is able to establish self-sustained populations, 59 thrive and spread, and only a fraction among them causes biodiversity losses, disruptions of 60 ecosystem functioning and economic impacts (Walsh et al., 2016). Understanding why some 61 introduced species succeed and become invasive, while other fail, is a central topic in invasion 62 science (Blackburn & Duncan, 2001; Blackburn, Prowse, Lockwood & Cassey, 2013; Facon et al., 63 2006). 64 The genetic diversity of introduced populations can influence their ability to adapt to novel 65 environments and, thus, determine their invasiveness (Lavergne & Molofsky, 2007; but see 66 Bossdorf, Richards & Pigliucci, 2008; Hawes et al., 2018). Biological invasions are a multistep 67 process often described as a series of stages (transport, introduction, establishment and spread) 68 separated by different barriers that can impede the progress of an invasion (Blackburn et al., 2011). 69 Overcoming each of these barriers can generate population bottlenecks and alter the genetic 70 diversity patterns in invasive populations (Hardesty et al., 2012; Okada, Lyle & Jasieniuk, 2009). 71 Different factors can modulate the intensity of population bottlenecks in each barrier of the invasion 72 process, including genetic diversity of the source population, propagule pressure, environmental 73 suitability of the reciepient area and/or the characteristics of secondary introductions. Genetic 74 admixture (hereafter admixture) occurs when multiple divergent genetic lineages come into contact 75 and interbreed, increasing the genetic diversity of a population, as can occur in the source 76 population of the native range before the transport stage (Dlugosch & Parker, 2008; Oficialdegui et 77 al., 2019; Rius & Darling, 2014; van Boheemen et al., 2017). During the introduction stage, 78 propagule pressure (i.e., number of introduction events, inoculum size or both) modulates resulting 79 genetic diversity patterns since more introduction events and/or a large number of introduced 80 individuals promote higher genetic diversity in the introduced population (Blackburn, Prowse, 81 Loockwood & Cassey, 2013; Drolet & Locke, 2016). During the establishment stage, biotic (i.e. 82 niche competition) and abiotic (i.e. environmental suitability) factors can affect the genetic diversity 83 of an introduced population through modulation of survival and its associated population bottleneck 84 (Banks et al., 2013; Ellegren & Galtier, 2016). As such, the environmental suitability refers to the 85 climatic and physiographic variables of the introduced range. During the spread stage, founding 86 events, involving all the previous cited modulators of genetic diversity, take place whenever a 87 secondary introduction occurs (i.e., the source population being itself introduced). Therefore, range 88 expansions are generally associated with decreasing genetic diversity (i.e., allelic richness and 89 expected heterozygosity) along the expansion front (Austerlitz, Jung-muller, Godelle & Gouyon, 90 1997; Excoffier, Foll, & Petit, 2009). 91 The red swamp crayfish (Procambarus clarkii), native to North-Eastern Mexico and South92 Central United States, has been broadly introduced around the world, to the point that it is present in 93 up to 40 countries of four continents (Oficialdegui, Sánchez & Clavero, 2020). It was intentionally 94 introduced to southern Spain in the early 1970s, through two independent shipments from Louisiana 95 (U.S.A.). Both introductions had legal authorisations and were motivated by the high 96 socioeconomic value that crayfish was attaining in Spain (Clavero, 2016). The first introduction 97 took place in Badajoz (Spain) in 1973, and involved the release of around 300 individuals, the 98 survivors of an original batch of 500 crayfish (Habsburgo-Lorena, 1978). One year later, a larger 99 batch (around 500 kg) was imported to the marsh area of the Lower Guadalquivir River (Puebla del 100 Río, Seville), although only 100 kg (around 6,500 individuals) survived. The red swamp crayfish 101 immediately established self-sustained and abundant populations in the initial introduction areas 102 and rapidly spread over the Iberian Peninsula (Gutiérrez-Yurrita et al., 1999; Oficialdegui, Sánchez 103 & Clavero, 2020), aided by both intrinsic traits (e.g., short life cycle, high fecundity, high 104 environmental tolerance; Geiger, Alcorlo, Baltanas & Montes, 2005) and by multiple (arguably, 105 thousands) and uncontrolled secondary introductions (Clavero, 2016; Oficialdegui et al., 2019). 106 Shortly after the introduction, by 1982, there were already reports of the red swamp crayfish in the 107 Tablas de Daimiel National Park and the Ebro Delta (some 320 and 730 km straight-line, 108 respectively, from the Lower Guadalquivir introduction site) (Clavero 2016). Once introduced and 109 established, this species often becomes dominant in the occupied freshwater habitats, producing 110 severe ecological impacts and losses of ecosystem services (Gherardi, 2006; Souty-Grosset et al. 111 2016). 112 In this study, we examine the spatial patterns of genetic diversity of the red swamp crayfish in 113 the Iberian Peninsula to test various hypotheses about the invasion history. Based on the analysis of 114 nuclear microsatellites, we aim to analyse the present-day genetic structure of the red swamp 115 crayfish, to then explore the drivers that may have modulated the dynamics of genetic diversity 116 during the invasion process. The main questions we address are: (i) is there a relationship between 117 the number and size of initial introduction events and the present-day genetic diversity?; (ii) how 118 was the pattern of spread of the swamp crayfish among the Iberian Peninsula?; (iii) is there a 119 relationship between environmental suitability and genetic diversity? We hypothesize that 120 populations originated from Lower Guadalquivir would have a higher genetic diversity than the 121 ones originated from Badajoz, as the inoculum size was around 20 times larger in the Lower 122 Guadalquivir. Besides, genetic patterns within the Iberian Peninsula would be mainly explained by 123 human-mediated dispersal, with a negligible influence of natural dispersal. We also consider two 124 dispersion processes: the jump-dispersal and the invasion hub scenario. The jump-dispersal scenario 125 assumes that the spread has occurred through successive small-scale secondary introductions, 126 supposing that genetic diversity would tend to diminish with increasing distances to the initial foci, 127 due to the accumulation of genetic bottlenecks at each secondary introduction. The invasion hub 128 scenario involves also large-scale translocations and relevant sources other than the initial foci (i.e. 129 the invasion hubs). It supposes that the high genetic diversity in the invasion hubs could enhance 130 genetic diversity in neighbouring introduced populations. Finally, we hypothesize that suitable 131 environmental conditions would reduce the intensity of population bottlenecks, so that crayfish 132 introduced in suitable areas would present higher genetic diversity than the ones in unsuitable areas. 133 134 Methods 135 Sample collection, DNA extraction and microsatellite genotyping 136 A total of 903 adult red swamp crayfish were collected from 28 locations distributed across the 137 Iberian Peninsula (Table 1; Fig. 1). A piece of abdominal muscle tissue was extracted from each 138 crayfish and stored in 96% ethanol at room temperature until subsequent analyses. 139 Total genomic DNA was extracted from approximately 10 mg of dried muscle tissue using a 140 modified DNA salt-extraction protocol (Aljanabi, 1997) containing NaCl 25 mM, Tris 12.5 mM 141 (pH 8.0), EDTA 12.5 mM (pH 8.0), 31.5 µL SDS 10%, 230 µL deionized water and Proteinase K. 142 After overnight incubation at 34 °C, DNA samples were extracted with a Tecan robot, Freedom Evo 143 model. Resulting DNA was diluted 1:10 and preserved at -20ºC for genotyping analyses. We 144 designed two multiplex PCRs for fragment analysis, with Mix 1 (PCSH0002, PCSH0006, PclG-17, 145 PclG-29) and Mix 2 (PCSH0038, PCSH0065, PclG-15, PclG-48) containing microsatellite loci 146 previously developed by Belfiore and May (2000) and Jiang et al. (2015). A multiplex polymerase 147 chain reaction (PCR) was performed on both Mix 1 and Mix 2 (Table S1). All PCR amplifications 148 were performed in 15 µL reactions containing 4 µL of template DNA, 3 µL buffer 5x PROMEGA, 149 2.5 mM dNTP, 25 mM MgCl2, 2 µL of Primer Mix (forward primer endlabelled with [32P] ATP), 150 0.75 U Taq polymerase PROMEGA, and deionized water up to the final volume of 15 μL. The 151 thermocycling regime of the Mix 1 consisted of an initial denaturation step at 95 ºC for 3 min, 152 followed by 8 cycles of denaturing at 95 ºC for 30 s, annealing at 60 ºC (decreasing 1 ºC for each 153 cycle) for 30 s, and extension at 72 ºC for 30 s, followed by 23 cycles of denaturing at 95 ºC for 30 154 s, annealing at 52 ºC for 30 s and 72 ºC for 30 s with a final extension at 72 ºC for 10 min. 155 Thermocycling conditions of the Mix 2 were 95 ºC for 3 min followed by 10 cycles of 95 ºC for 30 156 s, 60 ºC (decreasing 1 ºC for each cycle) for 30 s, 72 ºC for 30 s, followed by 23 cycles of 95 ºC for 157 30 s, 50 ºC for 30 s, 72 ºC for 30 s with a final extension at 72 ºC for 10 min. Genotyping of 158 amplified products were performed by using an ABI3130xl Genetic Analyser (Applied Biosystem, 159 UK) and allele size was determined using the Genescan 500-LIZ size standard and 160 electrophoretograms were scored in Genemapper version 4.0 (Applied Biosystems). All peaks were 161 manually verified by the lead author to ensure genotyping accuracy. 162 163 Genetic structure and diversity 164 MICROCHECKER v.2.2.3 was used to assess the presence of null alleles, large allele drop-outs and 165 scoring errors due to stuttering (van Oosterhout, Hutchinson, Wills & Shipley, 2004). GENEPOP 166 v.4.7.0 software (Rousset, 2008) was used to detect deviation from Hardy-Weinberg equilibrium 167 (HWE) and linkage disequilibrium (LD) between pairs of loci and each locus across locations. 168 While HWE test provided possible departures from equilibrium in our locations, which may 169 indicate systematic genotyping errors and other biases (Salanti, Amountza, Ntzani & Joannidis, 170 2005); LD test was used to assess the independence between analysed loci. Exact tests were used 171 with specified Markov chain parameters of 10,000 dememorization steps, followed by 5,000 172 batches of 5,000 iterations per batch. Statistical significance levels were adjusted according to 173 Bonferroni’s procedure to counteract the problem of multiple testing in HWE and linkage 174 disequilibrium (Rice, 1989). 175 In order to characterise the genetic diversity of the red swamp crayfish in the Iberian Peninsula, 176 we estimated the total number of observed alleles (NA), the effective number of alleles (NE), the 177 expected and the observed heterozygosity (HE and HO, respectively) and the inbreeding coefficients 178 (FIS) for each locus in each location by using GENALEX v.6.503 software (Peakall & Smouse, 179 2012). The allelic richness (AR) and the number of private alleles (PA) were calculated with ADZE 180 software (Szpiech, Jakobsson & Rosenberg, 2008), a rarefaction method to be able to compare 181 locations with different sampling sizes. In order to infer the genetic differentiation among locations, 182 pairwise FST values were calculated by using ARLEQUIN v.3.1 (Excoffier, Laval & Schneider, 183 2005). Bonferroni’s correction was performed to adjust the significance for multiple pairwise 184 comparisons in FST values (Rice, 1989). 185 BOTTLENECK v.1.2.02 was used to identify locations that have recently experienced a 186 significant reduction in effective population size (Piry, Luikart & Cornuert, 1999). This software 187 performs a test of heterozygosity based on the assumption that the number of alleles decreases 188 faster than the heterozygosity when a population experience a bottleneck. The stepwise-mutation 189 (SMM) and two-phased (TPM) models with 10,000 replicates were used to test population 190 bottlenecks. Variance for TPM was set to 30 and the proportion of SMM in TPM was set to 80%. 191 The Wilcoxon’s test was used to establish whether the number of loci showing heterozygosity 192 excess was significantly greater than expected in locations at equilibrium. 193 Isolation by distance (IBD) analysis was used to evaluate the relationship between genetic (FST) 194 and geographic (based on X-Y coordinates) distances among pairs of locations (Wright, 1943). A 195 Mantel test with 100,000 replicates was performed using ade4 package in R software (Dray & 196 Dufour, 2007). To calculate the geographic distances among Iberian locations we used the 197 geosphere (Hijmans, Williams, & Vennes, 2017) and Imap (Wallace, 2015) packages in R v3.2.3 (R 198 Development Core Team, 2014). 199 STRUCTURE v.2.3.4 was used to characterize the genetic structure of red swamp crayfish in 200 the Iberian Peninsula, and particularly to test whether the two introduction foci can explain the 201 present-day observed genetic structure (Pritchard, Stephens, & Donnelly, 2000). This Bayesian 202 clustering method assigns individuals to a given number of genetic clusters (K) based on their 203 genotypes. In order to identify the number of clusters, we first analysed the likelihood of models 204 with a number of clusters ranging from K = 1 to 27 (n-1). Due to the large number of clusters, we 205 performed 20 independent runs for each K, each run involving a Markov Chain Monte Carlo using 206 2000 burn-in followed by 10,000 iteration steps. Once preliminary results were obtained and to get 207 more accuracy, another analysis was performed from K = 1 to 8 with 20 independent runs for each 208 K, each run involving a Markov Chain Monte Carlo using 200,000 burn-in followed by 1,000,000 209 iteration steps. Admixture ancestry models and correlated allele frequencies (with default 210 parameters) were considered in all cases. The most likely value of real number of clusters in the 211 genetic dataset was estimated by examining the log probability of data [Ln Pr(X|K)] and the ΔK 212 method (Evanno et al., 2005) using STRUCTURE HARVESTER (Earl & vonHoldt, 2012). We 213 summarized the clustering results of multiple runs for each K value and these were visually 214 evaluated in CLUMPAK (http://clumpak.tau.ac.il) (Kopelman, Mayzel, Jakobsson, Rosenberg & 215 Mayrose, 2015). Additionally, a discriminant analysis of principal components (DAPC) was 216 performed to identify the number of different clusters without assuming marker linkage neither 217 HWE (Jombart, Devillard & Balloux, 2010). This multivariate method consists of a two-step 218 procedure to characterize population subdivision, being a Principal Component Analysis (PCA) as a 219 prior step to Discriminant Analysis (DA) (Jombart, Devillard & Balloux, 2010). DAPC was 220 performed using adegenet version 2.1.1 (Jombart, 2008) in the R environment. 221 222 Historical, human and environmental drivers of genetic diversity 223 We identified two robust genetic groups among red swamp crayfish locations in the Iberian 371 Peninsula, which arguably derive from the quasi-independent expansion of the two crayfish batches 372 introduced to Spain (1973 in Badajoz and 1974 in Lower Guadalquivir). Despite the sources of 373 crayfish for both introductions are arguably close areas in Louisiana (i.e., the native range), the lack 374 of a strong genetic structure and the large degree of genetic admixture in Louisiana (Oficialdegui et 375 al., 2019) could have favoured a random genetic distinction between the two transported batches 376 due to founder effect. Once in the Iberian Peninsula, the Badajoz group would have expanded 377 mainly westward into Portugal, but also, though less intensely, eastward (JAE location). The Lower 378 Guadalquivir group comprised most of the red swamp crayfish Spanish range, including also a 379 location in North-eastern Portugal (VLR location). The probability of belonging to a given group 380 was very high around the introduction foci of both groups (i.e., near Badajoz or around the Lower 381 Guadalquivir, Table S4), a pattern that strengthen the assumption that the observed genetic groups 382 clusters correspond to those initial introduction events. Similarly, in a previous work based on the 383 mtDNA, Oficialdegui et al. (2019) found one haplotype (Hap_06) that was present in most 384 Portuguese locations, but was not detected in the Lower Guadalquivir basin. 385 Since most alien species in inland waters are dispersed by human vectors (Cerri, Ciappelli, 386 Lenuzza, Zaccaroni & Nocita, 2018; Strayer, 2010), clear spatial structuring of genetic variability is 387 often lacking among populations of invasive freshwater species (Audzijonyte et al., 2017; Blakeslee 388 et al., 2017). However, we observed a strong genetic structure among red swamp crayfish 389 populations in the Iberian Peninsula, likely generated by the two expansion ways from both 390 introduction foci. A similar pattern was observed in the invasion process of the European green crab 391 in North America, in which two separately introduction events led to two genetically distinct groups 392 (FitzGerald et al., 2017; Jeffery et al., 2017). Although natural and artificial barriers in rivers can 393 define the spatial distribution or expansion of freshwater invasive species (see Teixeira et al., 2020), 394 the limited admixture in both genetic groups of red swamp crayfish (Badajoz and Lower 395 Guadalquivir) suggests an effect related to the political border between Spain and Portugal. The 396 border has apparently favoured the existence of two quasi-independent expansion processes, despite 397 both countries share several river basins through which natural dispersion of invasive species may 398 occur (Gago, Anastácio, Gkenas, Banha & Ribeiro, 2016). In fact, the red swamp crayfish may had 399 entered in Portugal through natural dispersion since the first record in the country is very near to the 400 introduction area in Badajoz (Ramos & Pereira, 1981), although short-distance human transport 401 cannot be discarded. However, present-day genetic patterns suggest that most subsequent human402 driven translocations have remained within the political border of Portugal, with few additional 403 introductions from Spain (even though at least one other did occur, see VLR). Contrastingly, the 404 expansion of the red swamp crayfish across Spain relied on the transport of individuals belonging 405 mainly to the Guadalquivir group (except JAE location). The genetic structure within the Lower 406 Guadalquivir group did not follow any clear spatial pattern (absence of isolation by distance and 407 spatial distribution of clusters 2 and 3), fitting well with the patterns reported for other widely 408 spread freshwater species (see above). The political border between Spain and Portugal may thus 409 act as an actual ecological barrier, as has been already described for other taxa (Arrondo et al., 410 2018; García et al., 2018). In the red swamp crayfish case this barrier does not seem to be related to 411 policy differences between countries (as reported Arrondo et al., 2018), but to the behaviour of the 412 people stocking crayfish, who apparently tended to remain within national limits. We thus 413 emphasize the importance of political boundaries as invisible barriers that can determine the 414 structure of wild populations, especially so for those species translocated by humans, calling for an 415 international coordination management measures accordingly (Dresser, Pierson, & Fitzpatrick, 416 2018; Rollins, Woolnough, Wilton, Sinclair & Sherwin, 2009). 417 418 Drivers of genetic diversity 419 Propagule pressure is a key factor modulating the probability of establishment in introduced 420 populations and their dynamics of genetic diversity (Lockwood, Cassey & Blackburn, 2005). 421 Overall, genetic diversity indices were lower in locations of the Badajoz group than in those of the 422 Lower Guadalquivir group, a pattern probably related to the higher propagule size of the 423 introduction event into the Lower Guadalquivir. In fact, the number of crayfish involved in the 424 Guadalquivir introduction was around 20 times larger than in the one taking place in Badajoz 425 (Habsburgo-Lorena, 1978). 426 The nature of transport events (short-distance, long-distance or both) and the existence of one 427 or several invasion hubs acting as genetically diverse sources of individuals (e.g., Fig. 2A and 2B) 428 may influence the genetic diversity spatial patterns of an already established invasive species. In the 429 Iberian Peninsula, it seems that large stocks of red swamp crayfish specimens were long-distance 430 translocated without intermediary bottlenecks (Gutierrez-Yurita et al., 1999), generating high 431 genetically diverse invasion hubs that subsequently acted as source for multiple secondary 432 introduction events. Accordingly, we found that genetic diversity tended to decrease in locations 433 that were farther to either its respective introduction focus or invasion hub. This steady decline in 434 genetic diversity along the invasion process corroborate the genetic consequences described for 435 population expansion range (Austerlitz et al., 1997; Excoffier et al., 2009). For example, White et 436 al. (2013) found a significant decline in genetic diversity across the expansion range of the bank 437 vole (Myodes glareolus) in Ireland. Based on historical information and environmental 438 characteristics, we had selected a priori the Ebro Delta and the Valencia Albufera as plausible 439 invasion hubs, but additional invasion hubs could have existed. Candidate areas for this role could 440 be the Tablas de Daimiel National Park, where the red swamp crayfish was introduced in 1982, or 441 the northern Spanish plateau, were a long-lasting tradition around crayfish consumption could have 442 favoured the occurrence of several introduction events (Clavero, 2016). Detecting possible invasion 443 hubs is a determining factor to predict and prevent potential range expansion of invasive species 444 over non-native territories (Muirhead & MacIsaac, 2005). 445 How the genetic diversity is affected by the distance to the introduction focus and by the 446 accumulation of bottleneck effects have been well studied in invasion biology (van Boheemen et al., 447 2017). However, our results also showed a positive relationship between genetic diversity and 448 environmental suitability for the red swamp crayfish in the Iberian Peninsula, highlighting the 449 importance of ecological factors in shaping the genetic patterns of invasive species. Similar patterns 450 have been reported for other invertebrate species (Ortego, Aguirre, Noguerales & Cordero, 2015). 451 They support the influence of environmental suitability on evolutionary processes through 452 demographic mechanisms that affect the effective population size (Wang, 2012) and ultimately 453 module the genetic patterns of populations. Furthermore, the relevance of the suitability-genetic 454 diversity relationships for the management of biological invasions can be modulated by climate 455 change (Tilman, Balzer, Hill & Befort, 2011). Capinha et al. (2012) predicted that suitable areas for 456 the red swamp crayfish in the Iberian Peninsula would show moderate changes (including both 457 suitability increases and decreases) from the present situation. But warmer future environments will 458 enhance climate suitability for the red swamp crayfish across several European areas (Zhang et al., 459 2019), which according to our results, could end up hosting viable and more genetically diverse red 460 swamp crayfish populations. 461 462 Conclusions 463 We have described clear patterns in genetic structure of red swamp crayfish in the Iberian 464 Peninsula, determined by the two introduction events that took place in the 1970s, a complex 465 human-mediated dispersal process involving the presence of invasion hubs, and a tendency of 466 higher levels of genetic diversity occurring in more suitable environments. These results help to 467 comprehend the invasion history of the red swamp crayfish in the Iberian Peninsula and how the 468 natural and anthropogenic factors modulate it. Our study thus highlights the importance of 469 analysing patterns of genetic variability to understand the invasion processes, a knowledge that can 470 be applied to manage current invasions and prevent plausible future ones. 471 472 iv. Acknowledgements 473 We thank R. López, M. A. Bravo, and Á. Vidal for field and lab assistance, as well as the many 474 colleagues who kindly provided samples from Spain [A. Pradilla (CCEDCV - EL PALMAR), L. M. 475 Álvarez (Asturias Government), N. Franch (Parc Natural del Delta de l'Ebre), L. Valls, F. J. 476 Galindo (Andalusian Government), F. Unzu and L. Lopo (La Rioja Government), J. Bosch, F. J. 477 Oliva, I. Vedia] and Portugal (F. Ribeiro and A. F. Filipe). We also thank Laboratorio de Ecología 478 Molecular (LEM-EBD) at Estación Biológica de Doñana (CSIC) as well as C. Lejeusne and C. 479 Daguin-Thiébaut for providing logistical support. We are especially grateful to César Capinha for 480 sharing with us his dataset. We thank two anonymous reviewers who gave helpful comments to 481 improve the final version of this manuscript. F.J.O. was supported by an Andalusian Government 482 grant. The study was funded by the Andalusian Government (RNM-936) and PICS (PIC2015FR4). 483 484 Conflicts of Interest 485 The authors declare no conflict of interest. 486 487 v. References 488 Aljanabi, S. (1997). Universal and rapid salt-extraction of high quality genomic DNA for PCR489 based techniques. Nucleic Acids Research, 25(22), 4692–4693. 490 https://doi.org/10.1093/nar/25.22.4692 491 Arrondo, E., Moleón, M., Cortés-Avizanda, A., Jiménez, J., Beja, P., Sánchez-Zapata, J. A., & 492 Donázar, J. A. (2018). Invisible barriers: Differential sanitary regulations constrain vulture 493 movements across country borders. Biological Conservation, 219, 46–52. 494 https://doi.org/10.1016/j.biocon.2017.12.039 495 Audzijonyte, A., Baltrūnaitė, L., Väinölä, R., & Arbačiauskas, K. (2017). Human-mediated lineage 496 admixture in an expanding Ponto-Caspian crustacean species Paramysis lacustris created a 497 novel genetic stock that now occupies European waters. Biological Invasions, 19(8), 2443– 498 2457. https://doi.org/10.1007/s10530-017-1454-9 499 Austerlitz, F., Jung-Muller, B., Godelle, B., & Gouyon, P. H. (1997). Evolution of coalescence 500 times, genetic diversity and structure during colonization. Theoretical Population 501 Biology, 51(2), 148-164. 502 Banks, S. C., Cary, G. J., Smith, A. L., Davies, I. D., Driscoll, D. A., Gill, A. M., … Peakall, R. 503 (2013). How does ecological disturbance influence genetic diversity? Trends in Ecology and 504 Evolution, 28(11), 670–679. https://doi.org/10.1016/j.tree.2013.08.005 505 Bates, D., Maechler, M., Bolker, B., Walker, S., Christensen, R. H. B., Singmann, H., ... & Rcpp, L. 506 (2015). Package ‘lme4’. Convergence, 12(1). 507 Belfiore, N. M., & May, B. (2000). Variable microsatellite loci in red swamp crayfish, 508 Procambarus clarkii, and their characterization in other crayfish taxa. Molecular Ecology, 9, 509 2231–2234. https://doi.org/10.1046/j.1365-294X.2000.105339.x 510 Bellard, C., Cassey, P., & Blackburn, T. M. (2016). Alien species as a driver of recent 511 extinctions. Biology letters, 12(2), 20150623. https://doi.org/10.1098/rsbl.2015.0623 512 Blackburn, T. M., & Duncan, R. P. (2001). Determinants of establishment success in introduced 513 birds. Nature, 414(6860), 195–197. https://doi.org/10.1038/35102557 514 Blackburn, T. M., Prowse, T. A. A., Lockwood, J. L., & Cassey, P. (2013). Propagule pressure as a 515 driver of establishment success in deliberately introduced exotic species: Fact or artefact? 516 Biological Invasions, 15(7), 1459–1469. https://doi.org/10.1007/s10530-013-0451-x 517 Blackburn, T. M., Pyšek, P., Bacher, S., Carlton, J. T., Duncan, R. P., Jarošík, V., … Richardson, 518 D. M. (2011). A proposed unified framework for biological invasions. Trends in Ecology and 519 Evolution, 26(7), 333–339. https://doi.org/10.1016/j.tree.2011.03.023 520 Blakeslee, A. M. H., Kamakura, Y., Onufrey, J., Makino, W., Urabe, J., Park, S., … Miura, O. 521 (2017). Reconstructing the invasion history of the Asian shorecrab, Hemigrapsus sanguineus 522 (De Haan 1835) in the Western Atlantic. Marine Biology, 164:47. doi: 10.1007/s00227-017523 3069-1. 524 Bossdorf, O., Richards, C. L., & Pigliucci, M. (2008). Epigenetics for ecologists. Ecology letters, 525 11(2), 106-115. https://doi.org/10.1111/j.1461-0248.2007.01130.x 526 Capinha, C., & Anastácio, P. (2011). Assessing the environmental requirements of invaders using 527 ensembles of distribution models. Diversity and Distributions, 17(1), 13–24. 528 https://doi.org/10.1111/j.1472-4642.2010.00727.x 529 Capinha, C., Anastácio, P., & Tenedório, J. A. (2012). Predicting the impact of climate change on 530 the invasive decapods of the Iberian inland waters: An assessment of reliability. Biological 531 Invasions, 14(8), 1737–1751. https://doi.org/10.1007/s10530-012-0187-z 532 Capinha, C., Essl, F., Seebens, H., Moser, D., & Pereira, H. M. (2015). The dispersal of alien 533 species redefines biogeography in the Anthropocene. Science, 348(6240), 1248–1251. 534 https://doi.org/10.1126/science.aaa8913 535 Cerri, J., Ciappelli, A., Lenuzza, A., Zaccaroni, M., & Nocita, A. (2018). Recreational angling as a 536 vector of freshwater invasions in Central Italy: perceptions and prevalence of illegal fish 537 restocking. Knowledge & Management of Aquatic Ecosystems, (419), 38. 538 https://doi.org/10.1051/kmae/2018028 539 Clavero, M. (2016). Species substitutions driven by anthropogenic positive feedbacks: Spanish 540 crayfish species as a case study. Biological Conservation, 193, 80–85. 541 https://doi.org/10.1016/j.biocon.2015.11.017 542 Clavero, M., López, V., Franch, N., Pou-Rovira, Q., & Queral, J. M. (2015). Use a seasonally 543 flooded rice fields by fish and crayfish in a Mediterranean wetland. Agriculture, Ecosystems & 544 Environment, 213, 39-46. https://doi.org/10.1016/j.agee.2015.07.022 545 Dias, P. J., Gilg, M. R., Lukehurst, S. S., Kennington, W. J., Huhn, M., Madduppa, H. H., ... 546 McDonald, J. I. (2018). Genetic diversity of a hitchhiker and prized food source in the 547 Anthropocene: the Asian green mussel Perna viridis (Mollusca, Mytilidae). Biological 548 invasions, 20(7), 1749-1770. 549 Díez-del-Molino, D., Carmona-Catot, G., Araguas, R. M., Vidal, O., Sanz, N., García-Berthou, E., 550 & García-Marín, J. L. (2013). Gene flow and maintenance of genetic diversity in invasive 551 mosquitofish (Gambusia holbrooki). PLoS ONE, 8(12), e82501. 552 https://doi.org/10.1371/journal.pone.0082501 553 Dlugosch, K. M., & Parker, I. M. (2008). Founding events in species invasions: Genetic variation, 554 adaptive evolution, and the role of multiple introductions. Molecular Ecology, 17(1), 431–449. 555 https://doi.org/10.1111/j.1365-294X.2007.03538.x 556 Dray, S., & Dufour, A.-B. (2007). The ade4 Package: Implementing the Duality Diagram for 557 Ecologists. Journal of Statistical Software, 22(4), 1-20. https://doi.org/10.18637/jss.v022.i04 558 Dresser, C. M., Pierson, T. W., & Fitzpatrick, B. M. (2018). Isolation by distance, local adaptation, 559 and fortuitous coincidence of geo-political boundaries with spatial-genetic clusters in southern 560 Bog Turtles. Global Ecology and Conservation, 16, e00474. 561 https://doi.org/10.1016/j.gecco.2018.e00474 562 Drolet, D., & Locke, A. (2016). Relative importance of propagule size and propagule number for 563 establishment of non-indigenous species: A stochastic simulation study. Aquatic Invasions, 564 11(1), 101–110. https://doi.org/10.3391/ai.2016.11.1.11 565 Earl, D. A., & vonHoldt, B. M. (2012). STRUCTURE HARVESTER: A website and program for 566 visualizing STRUCTURE output and implementing the Evanno method. Conservation 567 Genetics Resources, 4(2), 359–361. https://doi.org/10.1007/s12686-011-9548-7 568 Ellegren, H., & Galtier, N. (2016). Determinants of genetic diversity. Nature Reviews Genetics, 569 17(7), 422–433. https://doi.org/10.1038/nrg.2016.58 570 Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the number of clusters of individuals using 571 the software STRUCTURE: A simulation study. Molecular Ecology, 14(8), 2611–2620. 572 https://doi.org/10.1111/j.1365-294X.2005.02553.x 573 Excoffier, L., Laval, G., & Schneider, S. (2005). Arlequin (version 3.0): an integrated software 574 package for population genetics data analysis. Evolutionary bioinformatics, 1, 47-50. 575 https://doi.org/10.1177/117693430500100003 576 Excoffier, L., Foll, M., & Petit, J. (2009). Genetic Consequences of Range Expansions. Annual 577 Review of Ecology, Evolution, and Systematics, 40, 481-501. 578 https://doi.org/10.1146/annurev.ecolsys.39.110707.173414 579 Facon, B., Genton, B. J., Shykoff, J., Jarne, P., Estoup, A., & David, P. (2006). A general eco580 evolutionary framework for understanding bioinvasions. Trends in Ecology and Evolution, 581 21(3), 130–135. https://doi.org/10.1016/j.tree.2005.10.012 582 FitzGerald, J., Bradbury, I. R., Nadukkalam Ravindran, P., DiBacco, C., Stanley, R. R. E., 583 Matheson, K., … Van Wyngaarden, M. (2017). RAD sequencing reveals genomewide 584 divergence between independent invasions of the European green crab (Carcinus maenas) in 585 the Northwest Atlantic. Ecology and Evolution, 7(8), 2513–2524. 586 https://doi.org/10.1002/ece3.2872 587 Gago, J., Anastácio, P., Gkenas, C., Banha, F., & Ribeiro, F. (2016). Spatial distribution patterns of 588 the non‐native European catfish, Silurus glanis, from multiple online sources–a case study for 589 the River Tagus (Iberian Peninsula). Fisheries Management and Ecology, 23(6), 503-509. 590 https://doi.org/10.1111/fme.12189 591 García, E., López-Bao, J. V., Ferrand, N., Blanco, J. C., Palacios, V., Cortés, Y., … Rio-Maior, H. 592 (2018). Cryptic population structure reveals low dispersal in Iberian wolves. Scientific Reports, 593 8(1), 1–14. https://doi.org/10.1038/s41598-018-32369-3 594 Geiger, W., Alcorlo, P., Baltanas, A., & Montes, C. (2005). Impact of an introduced Crustacean on 595 the trophic webs of Mediterranean wetlands. Biological Invasions, 7(1), 49-73. 596 https://doi.org/10.1007/1-4020-3870-4_6 597 Gherardi, F. (2006). Crayfish invading Europe: The case study of Procambarus clarkii. Marine and 598 Freshwater Behaviour and Physiology, 39(3), 175–191. 599 https://doi.org/10.1080/10236240600869702 600 Gutiérrez-Yurrita, P. J., Martinez, J. M., Ilhéu, M., Bravo-Utrera, M. A., Bernardo, J. M., & 601 Montes, C. (1999). The status of crayfish populations in Spain and Portugal. Crustacean 602 Issues, 11, 161-192. 603 Habsburgo-Lorena, A. S. (1978). Present situation of exotic species of crayfish introduced into 604 Spanish continental water. Freshwater Crayfish, 4, 175–184. 605 Hardesty, B. D., Le Roux, J. J., Rocha, O. J., Meyer, J. Y., Westcott, D., & Wieczorek, A. M. 606 (2012). Getting here from there: Testing the genetic paradigm underpinning introduction 607 histories and invasion success. Diversity and Distributions, 18(2), 147–157. 608 https://doi.org/10.1111/j.1472-4642.2011.00832.x 609 Hawes, N. A., Fidler, A. E., Tremblay, L. A., Pochon, X., Dunphy, B. J., & Smith, K. F. (2018). 610 Understanding the role of DNA methylation in successful biological invasions: a review. 611 Biological Invasions, 20(9), 2285-2300. 612 Hijmans, R. J., Williams, E., & Vennes, C. (2017). Package ‘geosphere’ version 1.5-7. 613 https://CRAN.R-project.org/package=geosphere 614 Hulme, P. E., Bacher, S., Kenis, M., Klotz, S., Kühn, I., Minchin, D., … Vilà, M. (2008). Grasping 615 at the routes of biological invasions: A framework for integrating pathways into policy. 616 Journal of Applied Ecology, 45(2), 403–414. https://doi.org/10.1111/j.1365617 2664.2007.01442.x 618 Table 2. Summary of genetic diversity values for seven microsatellite loci in the 28 sampled 747 locations of red swamp crayfish distributed across the Iberian Peninsula. N = average number of 748 crayfish in each location/loci, NA = mean number of alleles observed, NE = mean number of 749 effective alleles, PA = mean number of private alleles corrected by the sampling size, AR = mean 750 allelic richness, HO = mean observed heterozygosity, HE = mean expected heterozygosity, and FIS = 751 mean fixation index. Values are shown by mean of the seven microsatellite markers and standard 752 error (SE). 753 Sampled locations N NA NE PA AR HO HE FIS ABF Mean 25.429 8.571 4.947 0.261 7.138 0.736 0.728 -0.022 ± SE 0.149 0.644 0.592 0.137 1.112 0.022 0.026 0.024 ANC Mean 30.000 6.143 3.606 0.103 5.364 0.605 0.687 0.114 ± SE 0.000 0.400 0.294 0.091 0.684 0.032 0.020 0.044 ARE Mean 19.857 4.857 2.632 0.000 4.472 0.568 0.590 0.015 ± SE 0.071 0.442 0.146 0.000 0.710 0.016 0.023 0.033 BDJ Mean 31.000 5.286 2.497 0.000 4.444 0.465 0.560 0.219 ± SE 0.000 0.340 0.152 0.000 0.466 0.051 0.027 0.060 BRU Mean 30.000 4.857 3.280 0.071 4.554 0.619 0.624 0.001 ± SE 0.000 0.335 0.355 0.063 0.647 0.028 0.030 0.015 CID Mean 29.000 7.571 4.682 0.013 6.347 0.768 0.749 -0.028 ± SE 0.000 0.697 0.388 0.012 0.980 0.021 0.019 0.017 DEB Mean 28.000 8.714 5.046 0.037 7.184 0.760 0.783 0.034 ± SE 0.000 0.531 0.340 0.035 0.863 0.025 0.012 0.021 GDP Mean 49.857 7.857 4.263 0.127 6.062 0.714 0.714 -0.002 ± SE 0.071 0.694 0.455 0.123 0.911 0.021 0.022 0.012 GIJ Mean 12.571 6.000 3.726 0.000 5.921 0.758 0.709 -0.066 ± SE 0.101 0.267 0.202 0.000 0.525 0.029 0.018 0.025 GUA Mean 49.857 9.429 4.753 0.045 7.045 0.667 0.764 0.120 ± SE 0.071 0.625 0.357 0.034 0.813 0.012 0.015 0.020 HUE Mean 39.571 7.143 4.139 0.046 5.978 0.741 0.750 0.016 ± SE 0.149 0.369 0.151 0.046 0.500 0.019 0.010 0.014 JAE Mean 30.000 5.143 2.712 0.059 4.381 0.505 0.603 0.161 ± SE 0.000 0.335 0.140 0.059 0.511 0.027 0.023 0.036 JIL Mean 15.000 6.571 4.781 0.000 6.416 0.724 0.759 0.045 ± SE 0.000 0.434 0.364 0.000 0.842 0.029 0.019 0.030 LEZ Mean 30.000 8.286 4.589 0.027 6.846 0.714 0.755 0.052 ± SE 0.000 0.508 0.384 0.026 0.762 0.016 0.014 0.016 LGQ Mean 47.286 10.571 6.079 0.126 7.932 0.787 0.795 0.009 ± SE 0.237 0.770 0.600 0.089 1.043 0.020 0.019 0.014 LOU Mean 30.000 6.714 3.541 0.000 5.685 0.686 0.693 0.009 ± SE 0.000 0.459 0.203 0.000 0.682 0.020 0.018 0.019 MAD Mean 30.000 6.286 3.740 0.001 5.467 0.743 0.701 -0.068 ± SE 0.000 0.389 0.223 0.001 0.637 0.019 0.023 0.015 MUN Mean 21.000 5.286 3.227 0.002 4.950 0.633 0.629 -0.019 ± SE 0.000 0.322 0.283 0.002 0.605 0.036 0.033 0.034 OLI Mean 48.714 5.857 3.100 0.034 4.643 0.633 0.645 0.026 ± SE 0.143 0.442 0.203 0.034 0.576 0.026 0.021 0.018 REG Mean 30.000 6.857 4.099 0.000 5.995 0.695 0.730 0.051 ± SE 0.000 0.550 0.282 0.000 0.792 0.021 0.016 0.009 REQ Mean 29.857 5.286 3.438 0.000 4.765 0.646 0.679 0.043 ± SE 0.071 0.340 0.234 0.000 0.581 0.021 0.019 0.025 ROC Mean 30.000 8.571 4.581 0.001 6.817 0.714 0.737 0.035 ± SE 0.000 0.635 0.415 0.001 0.953 0.026 0.021 0.010 SOP Mean 20.000 5.857 3.318 0.000 5.400 0.621 0.649 0.041 ± SE 0.000 0.442 0.240 0.000 0.753 0.040 0.031 0.042 STG Mean 50.000 9.714 5.086 0.054 6.737 0.777 0.745 -0.054 ± SE 0.000 0.696 0.564 0.046 1.093 0.019 0.023 0.027 VAL Mean 49.857 4.714 3.008 0.004 4.141 0.604 0.626 0.016 ± SE 0.071 0.389 0.240 0.004 0.647 0.028 0.023 0.042 VIL Mean 31.429 6.429 3.854 0.000 5.434 0.629 0.666 0.062 ± SE 0.101 0.606 0.415 0.000 0.891 0.035 0.032 0.012 VLB Mean 29.000 7.286 4.223 0.001 6.162 0.739 0.726 -0.027 ± SE 0.000 0.574 0.324 0.001 0.830 0.020 0.022 0.020 VLR Mean 29.714 6.143 3.404 0.000 5.034 0.654 0.659 0.005 ± SE 0.092 0.631 0.271 0.000 0.792 0.024 0.026 0.009 TOTAL Mean 32.030 6.857 3.941 0.036 5.761 0.675 0.695 0.028 ± SE 0.748 0.213 0.134 0.011 0.192 0.011 0.009 0.010 754 755 Table 3. Univariate and multivariate general linear models assessing the influence of different 756 predictors on the estimators of genetic diversity (allelic richness and expected heterozygosity). 757 Results are provided in terms of the direction of the relationship (positive “POS” or negative 758 “NEG”) for continuous predictors and comparing Lower Guadalquivir (LGQ) and Badajoz (BDJ) 759 groups, for the genetic group factor, with an indication of the statistical significance. The coefficient 760 of determination of the final multivariate models (selected following a backward procedure) is also 761 showed (for full models see Table S5). 762 Genetic group N Step distance Hub distance Environmental suitability Allelic richness Univariate LGQ>BDJ POS ** NEG * NEG** POS Multivariate (R2= 0.54) POS NEG* Expected heterozygosity Univariate LGQ>BDJ POS NEG NEG * POS Multivariate (R2= 0.13) NEG * (*) p < 0.05; (**) p < 0.01 763 764 vi. Figure captions 765 Figure 1. Genetic structure of the red swamp crayfish in the Iberian Peninsula, as resulting from 766 STRUCTURE outputs. The upper map shows the spatial distribution of the 28 locations and the 767 proportion of association to each of the genetic clusters defined for the most plausible K value (K = 768 3). Lower panels show the probability of assignment of red swamp crayfish individuals to the 769 genetic clusters defined for plausible K values (K = 2, K = 3 and K = 6, after the ΔK method, Fig. 770 S2). In these panels, each vertical line represents an individual, with individuals being grouped by 771 locations (codes as in Table 1), and genetic clusters are represented by different colours. 772 773 774 Figure 2. Schematic representation of plausible dispersal patterns of the red swamp crayfish 775 (Procambarus clarkii) across the Iberian Peninsula. In the jump-dispersal scenario (A), the 776 accumulation of bottlenecks due to successive introduction events would involve a reduction of 777 genetic diversity with increasing distance from the introduction focus (BDJ, LGQ). Contrastingly, 778 in the invasion hub scenario (B), long-distance transport of genetically diverse crayfish batches 779 (e.g., due to high propagule pressure, which is represented by arrow thickness) could have 780 generated invasion hubs (ABF, DEB), acting as additional sources for secondary introductions. 781 Thus, genetic diversity would decrease with increasing distances to either original introduction foci 782 or to invasion hubs. 783 784 Figure 3. Genetic diversity indices (allelic richness and expected heterozygosity) of the 28 red 785 swamp crayfish locations in relation to: i) the minimum distance of each location to its respective 786 introduction focus (Badajoz or Lower Guadalquivir, see Figure 1) or the closest invasion hub 787 (Valencia Albufera or Ebro Delta) (left panels); and ii) the environmental suitability of the area 788 occupied by those locations (right panels). Each location is assigned to one of the two genetic 789 groups identified (Badajoz group, blue dots; Lower Guadalquivir group, orange dots). Linear 790 regression lines and associated coefficients of determination for the two genetic groups pooled and 791 with the Badajoz population excluded (marked in all panels with an arrow, see results) are also 792 shown. 793 794 Supplementary Material 795 Table S1. Characteristics of the 8 polymorphic microsatellite loci for the red swamp crayfish. Ta = 796 annealing Temperature (ºC); NA = number of observed alleles, HO = observed heterozygosity and 797 HE = expected heterozygosity. 798 Locus Ta (ºC) Primer concentration NA HO/HE GeneBank accession no. Multiplex 1 PCSH38 60 0.4 8 0.606/0.613 KJ607979 PCSH65 55 0.4 14 0.629/0.673 KJ607985 PclG-15 60 0.8 21 0.807/0.823 AF290227 PclG-48 51 0.8 14 0.659/0.685 AF290241 Multiplex 2 PCSH02 60 0.8 18 0.739/0.812 KP675952 PCSH06 60 0.4 9 0.666/0.682 KP675956 PclG-17 50 0.4 14 0.674 /0.687 AF290229 PclG-29 50 0.4 10 0.311/0.641 AF290934 799 800 Table S2. Genetic diversity values of the 28 red swamp crayfish locations for seven microsatellite 801 loci. NA = number of alleles; NE = number of effective alleles; HO = observed heterozygosity; HE = 802 expected heterozygosity; HW = P-values for deviation of Hardy-Weinberg equilibrium; FIS = 803 fixation index (positive value indicates homozygosity excess); AR = mean allelic richness and N = 804 number individuals. Significant values of deviation of HW after Bonferroni correction are indicated 805 in bold. 806 Microsatellite loci Locations PCSH02 PCSH06 PCSH38 PCSH65 PclG-15 PclG-17 PclG-48 ABF NA 13 5 6 6 13 10 7 NE 9.324 2.295 2.370 3.173 9.398 4.881 3.189 HO 0.769 0.615 0.708 0.600 0.920 0.846 0.692 HE 0.893 0.564 0.578 0.685 0.894 0.795 0.686 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.138 -0.090 -0.225 0.124 -0.030 -0.064 -0.009 ANC NA 6 4 4 9 9 5 6 NE 3.061 2.946 3.704 6.818 4.045 2.145 2.525 HO 0.300 0.533 0.700 0.700 0.833 0.600 0.567 HE 0.673 0.661 0.730 0.853 0.753 0.534 0.604 HW 0.000 1.000 1.000 0.000 1.000 1.000 1.000 FIS 0.554 0.193 0.041 0.180 -0.107 -0.124 0.062 ARE NA 7 4 3 3 9 3 5 NE 3.374 2.036 1.802 2.532 3.587 1.831 3.265 HO 0.474 0.500 0.550 0.600 0.650 0.500 0.700 HE 0.704 0.509 0.445 0.605 0.721 0.454 0.694 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.327 0.017 -0.236 0.008 0.099 -0.102 -0.009 BDJ NA 8 3 4 4 6 5 7 NE 3.331 1.495 2.364 1.783 1.981 3.599 2.925 HO 0.484 0.194 0.613 0.161 0.258 0.839 0.710 HE 0.700 0.331 0.577 0.439 0.495 0.722 0.658 HW 0.000 0.000 1.000 0.02 0.000 0.000 1.000 FIS 0.309 0.415 -0.062 0.633 0.479 -0.161 -0.078 BRU NA 5 5 2 4 8 5 5 NE 3.435 3.579 1.965 3.152 7.171 1.861 1.796 HO 0.667 0.800 0.467 0.633 0.800 0.500 0.467 HE 0.709 0.721 0.491 0.683 0.861 0.463 0.443 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.060 -0.110 0.050 0.072 0.070 -0.080 -0.053 CID NA 11 7 6 4 14 7 4 NE 7.313 6.395 2.930 3.617 6.810 2.717 2.993 HO 0.828 0.897 0.793 0.759 0.862 0.586 0.655 HE 0.863 0.844 0.659 0.724 0.853 0.632 0.666 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.041 -0.063 -0.204 -0.048 -0.010 0.072 0.016 DEB NA 12 7 6 6 13 9 8 NE 6.588 3.613 3.409 3.970 8.340 4.854 4.545 HO 0.929 0.536 0.679 0.714 0.893 0.821 0.750 HE 0.848 0.723 0.707 0.748 0.880 0.794 0.780 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS -0.095 0.259 0.040 0.045 -0.014 -0.035 0.038 GDP NA 11 5 5 4 14 9 7 NE 5.405 2.539 2.303 3.347 9.182 3.010 4.052 HO 0.860 0.660 0.560 0.700 0.857 0.680 0.680 HE 0.815 0.606 0.566 0.701 0.891 0.668 0.753 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS -0.055 -0.089 0.010 0.002 0.038 -0.018 0.097 GIJ NA 8 5 5 4 7 7 6 NE 3.388 3.414 4.174 2.043 5.541 4.173 3.347 HO 0.667 0.923 0.833 0.500 0.923 0.692 0.769 HE 0.705 0.707 0.760 0.510 0.820 0.760 0.701 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.054 -0.305 -0.096 0.020 -0.126 0.089 -0.097 GUA NA 13 8 6 7 15 9 8 NE 6.964 4.227 2.804 3.597 7.859 3.870 3.953 HO 0.700 0.740 0.580 0.640 0.592 0.700 0.720 HE 0.856 0.763 0.643 0.722 0.873 0.742 0.747 HW 1.000 1.000 1.000 1.000 0.000 1.000 1.000 FIS 0.183 0.031 0.099 0.114 0.322 0.056 0.036 HUE NA 10 5 5 6 9 8 7 NE 5.452 3.674 2.875 3.879 4.520 4.161 4.414 HO 0.897 0.675 0.575 0.750 0.816 0.725 0.750 HE 0.817 0.728 0.652 0.742 0.779 0.760 0.773 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS -0.099 0.073 0.118 -0.011 -0.048 0.046 0.030 JAE NA 7 4 2 5 7 5 6 NE 3.000 2.233 1.980 3.303 3.442 3.377 1.651 HO 0.567 0.467 0.233 0.667 0.567 0.600 0.433 HE 0.667 0.552 0.495 0.697 0.709 0.704 0.394 HW 1.000 1.000 0.764 1.000 1.000 1.000 1.000 FIS 0.150 0.155 0.529 0.044 0.201 0.148 -0.099 JIL NA 8 5 4 4 10 8 7 NE 7.143 3.600 2.432 3.261 7.500 4.639 4.891 HO 0.733 0.800 0.600 0.533 0.933 0.867 0.600 HE 0.860 0.722 0.589 0.693 0.867 0.784 0.796 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.147 -0.108 -0.019 0.231 -0.077 -0.105 0.246 LEZ NA 12 7 7 5 12 7 8 NE 5.422 3.550 3.141 3.352 8.824 4.423 3.409 HO 0.667 0.667 0.700 0.733 0.900 0.667 0.667 HE 0.816 0.718 0.682 0.702 0.887 0.774 0.707 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.183 0.072 -0.027 -0.045 -0.015 0.139 0.057 LGQ NA 16 7 8 6 16 10 11 NE 11.267 5.284 2.853 3.137 9.596 5.251 5.166 HO 0.917 0.771 0.653 0.739 0.913 0.826 0.688 HE 0.911 0.811 0.650 0.681 0.896 0.810 0.806 HW 1.000 1.000 1.000 1.000 1.000 1.000 0.412 FIS -0.006 0.049 -0.005 -0.085 -0.019 -0.020 0.147 LOU NA 10 4 4 5 9 8 7 NE 5.028 2.317 2.503 3.719 4.027 4.557 2.635 HO 0.667 0.533 0.633 0.700 0.867 0.767 0.633 HE 0.801 0.568 0.601 0.731 0.752 0.781 0.621 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS 0.168 0.062 -0.055 0.043 -0.153 0.018 -0.021 MAD NA 8 6 4 5 10 5 6 NE 4.800 3.854 2.568 3.758 5.263 1.883 4.054 HO 0.867 0.733 0.667 0.800 0.833 0.567 0.733 HE 0.792 0.741 0.611 0.734 0.810 0.469 0.753 HW 1.000 1.000 1.000 1.000 1.000 1.000 1.000 FIS -0.095 0.010 -0.092 -0.090 -0.029 -0.209 0.027 MUN NA 8 3 5 5 5 4 7 Figure S2. The most likely value of real population clusters in the genetic dataset estimated by the 836 ΔK method (Evanno et al., 2005) after STRUCTURE results. Values of ΔK, showing peaks at K = 837 2, 3 and 6. 838 839 840 Figure S3. Discriminant analysis of principal components (DAPC) for the red swamp crayfish 841 locations in the Iberian Peninsula. The graph represents the individuals as dots and the populations 842 as inertia ellipses. PCA and DA eigenvalues are displayed in the inset where the retained 843 eigenvalues are depicted in black and the number of bars represents the number of discriminant 844 functions that retained in the analysis, respectively. See Table 1 for location. 845 846