Documents uploaded at Dryad: 1) Geographic_localities_Runemark_et_al.txt 2) Microsatellite_Data_Runemark_et_al.txt 3) Patch_size_He_population_level.txt 4) Individual_patch_size_He.txt 5) Clutch_size_HL.txt 6) Matings_He.txt 7) blueRegions.m 8) Test_lizard_1.jpg 9) Test_lizard_2.jpg 10) 3-way_selection_type_analysis.txt File 1) Contains a table with full population names in the first column (the names of the islets or names from the geographic locals or churches nearby), the population abbreviations used in Files 1 and 2 and in Fig. 1 in the manuscript in the seccond,habitat (i. e. if it is an islet or mainland local) in the third column, and finally the coordinates (latitude,longitude) in the fourth column. File 2) Contains the microsatellite data of the individuals included in the study. The first column entitled population includes the abbreviation for population affinity of the individuals from the paper and File 3. The following columns contain allele 1 and 2 for each loci used in the analysis; Lv319 1 is the first allele and Lv319 2 is the seccond. The loci are published in the two papers listed below: Runemark, A., Gabirot, M., Bensch, S., Svensson, E.I., Martín, J., Pafilis, P., Valakos, E.D. & Hansson, B. Cross-species testing of 27 pre-existing microsatellites in Podarcis gaigeae and P. hispanica (Squamata: Lacertidae) Mol. Ecol. Recourses 8(6):1367-1370. Wellenreuther, M., Runemark, A., Svensson, E. & Hansson, B. 2009. Ten variable polymorphic microsatellite loci for the Skyros Wall Lizard Podarcis gaigeae (Squamata: Lacertidae). Mol. Ecol. Recourses 9(3):1005-1008. File 3) Contains data on the percentage of the trunk that is covered by blue spots as well as expected heterozygosity (He) per population.The first column denotes population: population abbreviations are identical to those in the manuscript, the second column habitat of origin and the third column is expected heterozygosity (He) for the population. In the following columns mean %-age of the trunk covered by the blue patch, mean of the morphological traits and mean value of PC1 (first principal component which reflects size) are found. File 10) Contains data for the three-way analysis of whether traits subjected to different selection regimes (sexually selected vs naturally selected) are differently affected in the different habitats. Columns denote population of origin, habitat of origin,study trait, type of selection, expected heterozygosity, trait value and standardized trait value. File 4) contains individual spot sizes and expected heterozygosity for these individuals where both these are available (for some there can hence be small deviances from the population mean in He/patch size due to individuals for which not both of these pieces of information were available - no microsatellite data is available for individuals with lost tails and low quality photos were omitted from the patch size analysis but for calculating population means individual with only one of these pieces of information available were still valuable to strengthen the estimates). File 5) Contains clutch size, HL (estimate of heterozygosity, see paper), population and habitat of origin as well as snout-vent length for individual females. File 6) Contains mean number of matings per trial (including SE and 95% CI) per population and expected heterozygosity (He) for four populations. File 7) The matlab script identifying the blue patches. Note that the visual inspection stage which confirms that suitable sample areas have been selected and that the highlighted area fits with that of the patches is crucial for obtaining biologically relevant results using this software. We provide a sample image to illustrate which kind of input image that is suitable for this script. The file is a matlab file with the extension .m and doubleclicking it launches it in a computer with Matlab. Follow the command line instructions for the steps to perform. The select coarse area of interest is to be used to zoom in to the animal in pictures where the animal is only occupying a small part of the image. When you have visually evaluated the accuracy of the lizard, close the Figure windows containing images of that lizard. The script will then launch the image from the next .jpg file in the folder. Files 8-9) Sample images for testing the matlab-script described in file 7. All files except for 7 (.m) and 8 and 9 (.jpg) are .txt files. Title of paper: Has the inbreeding load for a condition-dependent sexual signalling trait been purged in insular lizard populations? Authors: Anna Runemark1*, Bengt Hansson1, Marcus Ljungqvist1, Mikkel Brydegaard2 & Erik I. Svensson1 Author affiliations: 1. Department of Biology, Ecology Building, Lund University, SE-223 62 Lund, SWEDEN. Phone: +46 46 222 3789. Fax: +46 46 222 4716 2. Department of Physics, Lund University, Professorsgatan 1, SE-22362 Lund, SWEDEN. Phone: +46 46 222 41 38. * Author for correspondence: Email: Anna.Runemark@biol.lu.se Telephone: +46 46 222 3789, Fax: +46 46 222 4717 RUNNING TITLE: Purging of the genetic load in Podarcis? KEYWORDS: Purging, inbreeding, sexual selection, island biology, Podarcis Abstract Sexually selected traits are often condition-dependent and are expected to be affected by genome-wide distributed deleterious mutations and inbreeding. However, sexual selection is a powerful selective force that can counteract inbreeding through purging of deleterious mutations. Inbreeding and purging of the inbreeding load for sexually selected traits has rarely been studied across natural populations with different degrees of inbreeding. Here we investigate inbreeding effects (measured as marker-based heterozygosity) on condition-dependent sexually selected signalling trait and other morphological traits across islet- and mainland populations (n=15) of an endemic lizard species (Podarcis gaigeae). Our data suggest inbreeding depression on a condition-dependent sexually selected signalling character among mainland subpopulations with low or intermediate levels of inbreeding, but no sign of inbreeding depression among small and isolated islet populations despite their higher overall inbreeding levels. In contrast, there was no such pattern among ten other morphological traits which are primarily naturally selected and presumably not involved in sexual signalling. These results are in line with purging of recessive deleterious alleles, or purging in combination with stochastic fixation of alleles by genetic drift, for a sexual signalling character in the islet environment, which is characterized by low population sizes and strong sexual selection. Higher clutch sizes in islet populations also raise interesting questions regarding the possibility of antagonistic pleiotropy. Purging and other non-exclusive explanations for of our results are discussed.