Amarula Elephant Research Programme, School of Life Sciences, University of Kwazulu-Natal, Durban, South Africa.
PLoS One. 2012;7(6):e38363. doi: 10.1371/journal.pone.0038363. Epub 2012 Jun 27.
The identification of temporal thresholds or shifts in animal movement informs ecologists of changes in an animal's behaviour, which contributes to an understanding of species' responses in different environments. In African savannas, rainfall, temperature and primary productivity influence the movements of large herbivores and drive changes at different scales. Here, we developed a novel approach to define seasonal shifts in movement behaviour by examining the movements of a highly mobile herbivore (elephant; Loxodonta africana), in relation to local and regional rainfall patterns.
METHODOLOGY/PRINCIPAL FINDINGS: We used speed to determine movement changes of between 8 and 14 GPS-collared elephant cows, grouped into five spatial clusters, in Kruger National Park, South Africa. To detect broad-scale patterns of movement, we ran a three-year daily time-series model for each individual (2007-2009). Piecewise regression models provided the best fit for elephant movement, which exhibited a segmented, waveform pattern over time. Major breakpoints in speed occurred at the end of the dry and wet seasons of each year. During the dry season, female elephant are constrained by limited forage and thus the distances they cover are shorter and less variable. Despite the inter-annual variability of rainfall, speed breakpoints were strongly correlated with both local and regional rainfall breakpoints across all three years. Thus, at a multi-year scale, rainfall patterns significantly affect the movements of elephant. The variability of both speed and rainfall breakpoints across different years highlights the need for an objective definition of seasonal boundaries.
CONCLUSIONS/SIGNIFICANCE: By using objective criteria to determine behavioural shifts, we identified a biologically meaningful indicator of major changes in animal behaviour in different years. We recommend the use of such criteria, from an animal's perspective, for delineating seasons or other extrinsic shifts in ecological studies, rather than arbitrarily fixed definitions based on convention or common practice.
动物运动的时间阈值或转变的识别可以使生态学家了解动物行为的变化,从而有助于理解不同环境下物种的反应。在非洲稀树草原,降雨、温度和初级生产力影响大型食草动物的运动,并在不同尺度上引发变化。在这里,我们开发了一种新方法,通过检查高度移动的食草动物(大象;Loxodonta africana)的运动,来定义运动行为的季节性转变,与当地和区域降雨模式有关。
方法/主要发现:我们使用速度来确定南非克鲁格国家公园 8 到 14 头 GPS 项圈母象的运动变化,这些象分为五个空间集群。为了检测运动的广泛模式,我们为每个个体运行了一个为期三年的每日时间序列模型(2007-2009 年)。分段回归模型为大象运动提供了最佳拟合,大象运动随时间呈现出分段的波动模式。速度的主要断点发生在每年旱季和雨季结束时。在旱季,雌性大象受到有限饲料的限制,因此它们覆盖的距离较短,变化较小。尽管降雨的年际变化很大,但速度断点与三年来的当地和区域降雨断点都有很强的相关性。因此,在多年的时间尺度上,降雨模式显著影响了大象的运动。不同年份速度和降雨断点的可变性突出表明需要客观定义季节性边界。
结论/意义:通过使用客观标准来确定行为转变,我们确定了动物行为在不同年份发生重大变化的生物意义指标。我们建议从动物的角度使用这些标准来划定季节或其他生态研究中的外部转变,而不是基于传统或惯例的任意固定定义。