The Environment Institute and School of Biological Sciences, University of Adelaide, Adelaide, South Australia, 5005, Australia.
Department of Biosciences, Swansea University, Swansea, SA2 8PP, United Kingdom.
Ecol Appl. 2020 Jun;30(4):e02083. doi: 10.1002/eap.2083. Epub 2020 Mar 9.
The European rabbit (Oryctolagus cuniculus) is a notorious economic and environmental pest species in its invasive range. To better understand the population and range dynamics of this species, 41 yr of abundance data have been collected from 116 unique sites across a broad range of climatic and environmental conditions in Australia. We analyzed this time series of abundance data to determine whether interannual variation in climatic conditions can be used to map historic, contemporary, and potential future fluctuations in rabbit abundance from regional to continental scales. We constructed a hierarchical Bayesian regression model of relative abundance that corrected for observation error and seasonal biases. The corrected abundances were regressed against environmental and disease variables in order to project high spatiotemporal resolution, continent-wide rabbit abundances. We show that rabbit abundance in Australia is highly variable in space and time, being driven primarily by internnual variation in temperature and precipitation in concert with the prevalence of a non-pathogenic virus. Moreover, we show that internnual variation in local spatial abundances can be mapped effectively at a continental scale using highly resolved spatiotemporal predictors, allowing "hot spots" of persistently high rabbit abundance to be identified. Importantly, cross-validated model performance was fair to excellent within and across distinct climate zones. Long-term monitoring data for invasive species can be used to map fine-scale spatiotemporal fluctuations in abundance patterns when accurately accounting for inherent sampling biases. Our analysis provides ecologists and pest managers with a clearer understanding of the determinants of rabbit abundance in Australia, offering an important new approach for predicting spatial abundance patterns of invasive species at the near-term temporal scales that are directly relevant to resource management.
欧洲兔（Oryctolagus cuniculus）在其入侵范围内是一种臭名昭著的经济和环境害虫。为了更好地了解该物种的种群和分布动态，我们从澳大利亚广泛的气候和环境条件下的 116 个独特地点收集了 41 年的丰度数据。我们分析了这一时间序列的丰度数据，以确定气候条件的年际变化是否可用于绘制从区域到大陆尺度的历史、当代和潜在未来兔子丰度波动的图。我们构建了一个相对丰度的分层贝叶斯回归模型，该模型纠正了观测误差和季节性偏差。将校正后的丰度与环境和疾病变量进行回归，以预测具有高时空分辨率的全大陆范围的兔子丰度。我们表明，澳大利亚的兔子丰度在空间和时间上具有高度的可变性，主要由温度和降水的年际变化以及一种非致病性病毒的流行驱动。此外，我们表明，使用高分辨率的时空预测因子，可以有效地在大陆尺度上绘制局部空间丰度的年际变化，从而识别出持续高兔子丰度的“热点”。重要的是，在不同的气候区内部和之间，交叉验证的模型性能都是公平到优秀的。当准确考虑到固有的采样偏差时，入侵物种的长期监测数据可用于绘制精细的时空丰度波动模式。我们的分析为生态学家和害虫管理者提供了对澳大利亚兔子丰度决定因素的更清晰的理解，为预测入侵物种在直接与资源管理相关的近期时间尺度上的空间丰度模式提供了一种重要的新方法。
