Population dynamics of exploited species in a stochastic environment

我的研究的长期目标是提高我们对野生动物种群动态的理解，并开发新的工具来模拟人类开发和自然扰动对基于标记动物的种群动态的影响。我将专注于具有不可观测状态的多状态捕获-标记-再捕获（CMR）模型，这是最近开发的一类强大的模型。状态可以是个体在重新捕获时的状况（例如繁殖者/非繁殖者），不可观察的状态是当非繁殖者不在繁殖场所时，因此不能被捕获。我们用一个独特的数据集来解决这些问题，该数据集来自北极地区正在进行的大雪雁标记计划（自1989年以来，我们对60，000只鸟进行了标记）。考虑到我们对物种的理解水平不断提高，这个种群是一个很好的模型系统，可以用来测试生态理论和野生动物管理中的一般问题。鹅在北极繁殖，这是一个受气候变化影响的高度随机的环境，这一种群目前受到春季狩猎，这是北美独特的保护措施，因为它们被认为是过剩的。我的目标是1-应用混合信息的多状态模型（活的遭遇和死亡的恢复），以测试春季狩猎对招募率，第一次繁殖年龄和雪雁生存的影响; 2-确定信息的混合是否允许多状态模型中限制性较低的假设，并可用于研究生存和狩猎死亡率之间的关系，季节性的基础上;（3）确定鸟类集群中的长期聚集在多大程度上违反了CMR分析中关于命运独立性的假设;（4）继续发展卡尔曼滤波建模方法，这是一种新的生态学数学方法，旨在减少种群模型预测的不确定性;研究北极气候因素的变化如何影响雪雁的生命率，并对这些影响进行建模。该项目开发的模型将适用于人类利用的各种动物物种。我将在北极野生动物种群管理领域培养几名高素质的人才（博士、硕士和博士后）。

The long-term goals of my research are to improve our understanding of the population dynamics of wildlife species, and to develop new tools to model the effect of human exploitation and natural perturbations on the dynamics of populations based on marked animals. I will focus on multistate capture-mark-recapture (CMR) models with unobservable states, a powerful class of models recently developed. A state can be the condition of an individual at recapture (e.g. breeder/non-breeder) and an unobservable state is when non-breeders are absent from the breeding site, and thus cannot be captured. We address these questions with a unique dataset from an on-going marking program of Greater Snow Geese in the Arctic (since 1989, we banded 60,000 birds). Given our increasing level of understanding of the species, this population is an excellent model system to test general questions in ecological theory and wildlife management. Geese breed in the Arctic, a highly stochastic environment impacted by climate change, and this population is currently subject to a spring hunt, a unique conservation measure in North America because they are considered overabundant. My objectives are to 1- apply multistate models with mixture of information (live encounters and dead recoveries) to test the effects of spring hunt on recruitment rate, age at first breeding, and survival in snow geese; 2- determine if mixture of information allows less restrictive assumptions in multistate models and can be used to examine the relationship between survival and hunting mortality on a seasonal basis; 3- determine to what extent long-lasting aggregations in flocking birds violate the assumption of independence of fate in CMR analyses; 4- continue to develop the Kalman Filter modelling approach, a novel mathematical approach in ecology designed to reduce uncertainties in population model predictions; 5- examine how variability in climatic factors in the Arctic affects vital rates of snow geese and model these effects on population dynamics. The models developed in this project will be applicable to a wide range of animal species exploited by humans. I will train several highly qualified people (PhD, MSc and post-doctorate) in the area of management of arctic wildlife population.