Rapid adaptation to global change can counter vulnerability of species to population declines and extinction. Theoretically, under such circumstances both genetic variation and phenotypic plasticity can maintain population fitness, but empirical support for this is currently limited. Here, we aim to characterise the role of environmental and genetic diversity, and their prior evolutionary history (via haplogroup profiles) in shaping patterns of life history traits during biological invasion. Data were derived from both genetic and life history traits including a morphological analysis of 29 native and invasive populations of topmouth gudgeon Pseudorasbora parva coupled with climatic variables from each location. General additive models were constructed to explain distribution of somatic growth rate (SGR) data across native and invasive ranges, with model selection performed using Akaike’s Information Criteria. Genetic and environmental drivers that structured the life history of populations in their native range were less influential in their invasive populations. For some vertebrates at least, fitness related trait shifts does not seems to be dependent on the level of genetic diversity or haplogroup make-up of the initial introduced propagule, nor of the availability of local environmental conditions being similar to those experienced in their native range. As long as local conditions are not beyond the species physiological threshold, its local establishment and invasive potential are likely to be determined by local drivers, such as density dependent effects linked to resource availability or to local biotic resistance.

The dataset comprised of populations of P. parva across the native (China, Japan and Taiwan) and invasive range in Europe, all collected between June and August 2010. Following the sampling of each population (by fish trapping, seine netting or electric fishing, method dependent on the habitat sampled), 50 fish were randomly selected and euthanized (overdose of anaesthetic, 120 mg/L benzocaïne), with fin tissues collected and preserved in 98% ethanol prior to the whole fish being preserved in 10% formalin. Each sampling site was geo-located using a GPS and typology of the site recorded.

For each fish, fork length (FL, mm), weight (W, g) and sex were recorded. For the production of all data for somatic growth rate analyses, ages of individual fish were obtained from ageing of scales, which were collected above the lateral line and below the insertion of the dorsal fin. Ageing was completed on a projecting microscope by counting the number of annual growth checks present. Each fish length was then plotted against age (years), and the von Bertalanffy growth model parameters length infinity (Linf) and k (see Sainsbury, 1980) calculated. In addition, the growth metrics somatic growth rate (SGR (cm/year); FL _{at capture}-FL_{age 1})/ age at capture) and FL at year one (FL1) were determined, which also corresponded to the age at maturity in all populations (Gozlan et al., 2010a).

Reproductive traits were assessed using fecundity and gonado-somatic index (GSI) of female fish. Ovaries were extracted from each female and weighed (to 0.01 g; Wo). A sub-sample of each ovary, cut from the middle of the gonad, was weighed (Ws) and the number of oocytes counted under a binocular microscope. Each oocyte was classified by size categories using a fitted micrometer (i.e. <0.1; [0.4-0.5]; [0.6-0.8]; [0.9-1]; >1.1 mm ± 0.05mm) with the sum of oocytes (n) representing potential fecundity (oocytes <0.1 mm were not included in counts). The potential fecundity was then calculated as [FEC = n*Wo/WS]. GSI was calculated as [Gonad weight / Total fish weight] x 100 (Strum, 1978) and only females with GSI above 12 % were considered in reproductive state and used in analyses.

To examine patterns of relative growth, raw data from 30 mensural characters, including fork length (FL; see also Záhorská et al., 2009), were measured from digital photographs taken by a Pentax optio S10 camera, with analysis using IMPOR 2.31E software.

The DRYAD data are self explanatory. All other information are provided directly in the published paper.