Predicting the genetic impact of stocking in Brook Charr (Salvelinus fontinalis) by combining RAD sequencing and modeling of explanatory variables

In fisheries management, intensive stocking programs are commonly used to enhance population abundance and maintain stock productivity. However, such practices are increasingly raising concerns since multiple studies documented adverse genetic and evolutionary impacts of stocking on wild populations. Improvement of stocking management relies on a better understanding of the dynamic of introgressive hybridization between wild and domestic population and on assessment of the genetic state of wild populations after stocking cessation. In Québec, Canada, over five million captive reared Brook Charr (Salvelinus fontinalis) are stocked every year to support recreational fishing activities. Here we investigated how variation in stocking history and environmental variables, including water temperature, pH and dissolved oxygen, may influence the impact of stocking practices on the genetic integrity of wild Brook Charr populations. We collected DNA samples (n = 862, average of 30 individuals per lake) from 29 lakes that underwent different stocking intensity through time and also collected environmental parameters for each sampled lake. An average of 4580 high quality filtered SNPs was obtained for each population using Genotyping-By-Sequencing (GBS) which were then used to quantify the mean domestic membership of each sampled population. An exhaustive process of model selection was conducted to obtain a best-fitted model that explained 56% of the variance observed in mean domestic genetic membership. The number of years since the mean year of stocking was the best explanatory variable to predict variation in mean domestic genetic membership whereas environmental characteristics had little influence on observed patterns of admixture. Our model predictions also revealed that each sampled wild population could potentially return to a wild genetic state (absence of domestic genetic background) after stocking cessation. Overall, our study provides new insights on factors determining level of introgressive hybridization and suggests that stocking impacts could be reversible with time.

在渔业管理领域，集约化放养计划常被用于提升种群丰度、维持种群生产能力。然而，此类做法正日益引发担忧，多项研究已证实放养会对野生种群产生不利的遗传与进化影响。优化放养管理，有赖于更深入理解野生与家养种群间的渐渗杂交动态，以及评估放养停止后野生种群的遗传状态。加拿大魁北克地区每年会放养超过500万尾人工繁育的溪红点鲑（Brook Charr, *Salvelinus fontinalis*），以支撑休闲垂钓活动。本研究旨在探究放养历史差异，以及水温、pH值、溶解氧等环境变量如何影响放养行为对野生溪红点鲑种群遗传完整性的作用效果。我们从29个历史放养强度各异的湖泊中采集了DNA样本（总样本量n=862，每个湖泊平均采集30尾个体），同时为每个采样湖泊收集了环境参数。本研究通过简化基因组测序（Genotyping-By-Sequencing, GBS）技术，为每个种群获取了平均4580个经过质量筛选的高质量单核苷酸多态性位点，随后利用这些位点量化每个采样种群的平均家养遗传占比。研究通过穷尽式模型选择流程构建出最优拟合模型，该模型可解释平均家养遗传占比观测变异的56%。结果显示，预测平均家养遗传占比变异的最优解释变量为距平均放养年份的间隔年数，而环境特征对观测到的遗传混合模式影响甚微。模型预测还表明，所有采样的野生种群在放养停止后均有可能恢复至野生遗传状态（即无家养遗传背景）。综上，本研究为解析决定渐渗杂交水平的影响因素提供了新视角，并表明放养带来的遗传影响可随时间推移得到逆转。