Impacts of Land Cover Data Selection and Trait Parameterisation on Dynamic Modelling of Species’ Range Expansion

Dynamic models for range expansion provide a promising tool for assessing species’ capacity to respond to climate change by shifting their ranges to new areas. However, these models include a number of uncertainties which may affect how successfully they can be applied to climate change oriented conservation planning. We used RangeShifter, a novel dynamic and individual-based modelling platform, to study two potential sources of such uncertainties: the selection of land cover data and the parameterization of key life-history traits. As an example, we modelled the range expansion dynamics of two butterfly species, one habitat specialist (Maniola jurtina) and one generalist (Issoria lathonia). Our results show that projections of total population size, number of occupied grid cells and the mean maximal latitudinal range shift were all clearly dependent on the choice made between using CORINE land cover data vs. using more detailed grassland data from three alternative national databases. Range expansion was also sensitive to the parameterization of the four considered life-history traits (magnitude and probability of long-distance dispersal events, population growth rate and carrying capacity), with carrying capacity and magnitude of long-distance dispersal showing the strongest effect. Our results highlight the sensitivity of dynamic species population models to the selection of existing land cover data and to uncertainty in the model parameters and indicate that these need to be carefully evaluated before the models are applied to conservation planning.用于物种分布范围扩张的动态模型，为评估物种通过向新区域迁移以应对气候变化的能力提供了极具前景的工具。然而此类模型存在诸多不确定性，可能会影响其在面向气候变化的保护规划中的应用效果。我们采用RangeShifter——一款新型的基于个体的动态建模平台——来探究此类不确定性的两大潜在来源：土地覆被数据的选择，以及关键生活史特征的参数化设置。作为示例，我们对两种蝴蝶的分布范围扩张动态进行了建模：一种为栖息地专性物种（Maniola jurtina），另一种为广适性物种（Issoria lathonia）。研究结果显示，种群总规模、占据网格单元数量以及平均最大纬度迁移范围的预测结果，均明显取决于两种数据选择的差异：一是使用CORINE土地覆被数据（CORINE land cover data），二是使用来自三个备选国家数据库的更精细草地覆被数据。分布范围扩张同样对所考量的四项生活史特征的参数化设置较为敏感，这四项特征分别为长距离扩散事件的强度与发生概率、种群增长率以及环境容纳量，其中环境容纳量与长距离扩散强度的影响最为显著。本研究结果凸显了物种种群动态模型对现有土地覆被数据选择以及模型参数不确定性的敏感性，并指出在将此类模型应用于保护规划之前，需对上述因素进行审慎评估。