Solar Radiation Determines Site Occupancy of Coexisting Tropical and Temperate Deer Species Introduced to New Zealand Forests

Assemblages of introduced taxa provide an opportunity to understand how abiotic and biotic factors shape habitat use by coexisting species. We tested hypotheses about habitat selection by two deer species recently introduced to New Zealand’s temperate rainforests. We hypothesised that, due to different thermoregulatory abilities, rusa deer (Cervus timorensis; a tropical species) would prefer warmer locations in winter than red deer (Cervus elaphus scoticus; a temperate species). Since adult male rusa deer are aggressive in winter (the rut), we also hypothesised that rusa deer and red deer would not use the same winter locations. Finally, we hypothesised that in summer both species would prefer locations with fertile soils that supported more plant species preferred as food. We used a 250 × 250 m grid of 25 remote cameras to collect images in a 100-ha montane study area over two winters and summers. Plant composition, solar radiation, and soil fertility were also determined for each camera location. Multiseason occupancy models revealed that direct solar radiation was the best predictor of occupancy and detection probabilities for rusa deer in winter. Multistate, multiseason occupancy models provided strong evidence that the detection probability of adult male rusa deer was greater in winter and when other rusa deer were present at a location. Red deer mostly vacated the study area in winter. For the one season that had sufficient camera images of both species (summer 2011) to allow two-species occupancy models to be fitted, the detection probability of rusa deer also increased with solar radiation. Detection probability also varied with plant composition for both deer species. We conclude that habitat use by coexisting tropical and temperate deer species in New Zealand likely depends on the interplay between the thermoregulatory and behavioural traits of the deer and the abiotic and biotic features of the habitat.

引入类群（introduced taxa）的集合为探究非生物与生物因子如何塑造共存物种的生境利用（habitat use）提供了研究契机。我们针对新近被引入新西兰温带雨林的两种鹿类，开展了生境选择（habitat selection）相关假说的验证工作。基于体温调节能力的差异，我们提出假说：泽鹿（rusa deer，*Cervus timorensis*，热带物种）在冬季会相较于马鹿（red deer，*Cervus elaphus scoticus*，温带物种）更偏好温暖生境。鉴于成年雄性泽鹿在冬季（发情期，rut）具有攻击性，我们进一步提出假说：泽鹿与马鹿不会共享同一冬季生境。最后，我们推测夏季时两种鹿类均会偏好土壤肥沃、可提供更多偏好性食用植物的生境。我们在一处100公顷的山地研究区域（montane study area）内，以250×250米的网格布设25台远程相机（remote cameras），在两个冬季与两个夏季期间收集影像数据。同时，我们测定了每个相机布设点位的植物组成（plant composition）、太阳辐射（solar radiation）与土壤肥力（soil fertility）。多季节占用模型（multiseason occupancy models）分析显示，直接太阳辐射是影响泽鹿冬季占用率与检测率的最优预测因子。多状态多季节占用模型（multistate, multiseason occupancy models）提供了强有力的证据：冬季以及当同一点位存在其他泽鹿时，成年雄性泽鹿的检测率会升高。马鹿在冬季大多会撤离研究区域。仅在2011年夏季，两种鹿类均有充足的相机影像可供开展双物种占用模型（two-species occupancy models）拟合，此时泽鹿的检测率同样随太阳辐射强度升高而增加。此外，两种鹿类的检测率均会随植物组成发生变化。我们认为，新西兰共存的热带与温带鹿类的生境利用，大概率取决于鹿类的体温调节与行为特征，以及生境的非生物与生物特征之间的相互作用。