Fu Jinbo, Zhao Linlin, Liu Changdong, Sun Bin
Department of Fisheries, Ocean University of China, Qingdao, Shandong, China.
First Institute of Oceanography, Ministry of Natural Resources, Qingdao, Shandong, China.
PeerJ. 2021 Aug 17;9:e12001. doi: 10.7717/peerj.12001. eCollection 2021.
As IUCN critically vulnerable species,the Indo-Pacific humpback dolphins () have attracted great public attention in recent years. The threats of human disturbance and environmental pollution to this population have been documented extensively. However, research on the sensitivity of this species to climate change is lacking. To understand the effect of climate change on the potential distribution of , we developed a weighted ensemble model based on 82 occurrence records and six predictor variables (e.g., ocean depth, distance to shore, mean temperature, salinity, ice thickness, and current velocity). According to the true skill statistic (TSS) and the area under the receiver operating characteristic curve (AUC), our ensemble model presented higher prediction precision than most of the single-algorithm models. It also indicated that ocean depth and distance to shore were the most important predictors in shaping the distribution patterns. The projections for the 2050s and 2100s from our ensemble model indicated a severe adverse impact of climate change on the habitat. Over 75% and 80% of the suitable habitat in the present day will be lost in all representative concentration pathway emission scenarios (RCPS) in the 2050s and 2100s, respectively. With the increased numbers of records of stranding and deaths of in recent years, strict management regulations and conservation plans are urgent to safeguard the current suitable habitats. Due to habitat contraction and poleward shift in the future, adaptive management strategies, including designing new reserves and adjusting the location and range of reserves according to the geographical distribution of , should be formulated to minimize the impacts of climate change on this species.
作为世界自然保护联盟（IUCN）的极度濒危物种，印太驼背豚近年来备受公众关注。人类干扰和环境污染对这一物种的威胁已有大量记录。然而，关于该物种对气候变化敏感性的研究却很缺乏。为了解气候变化对印太驼背豚潜在分布的影响，我们基于82个出现记录和六个预测变量（如海洋深度、离岸距离、平均温度、盐度、冰层厚度和海流速度）开发了一个加权集合模型。根据真技能统计量（TSS）和受试者工作特征曲线下面积（AUC），我们的集合模型比大多数单算法模型具有更高的预测精度。结果还表明，海洋深度和离岸距离是塑造其分布格局的最重要预测因子。我们的集合模型对2050年代和2100年代的预测表明，气候变化将对印太驼背豚的栖息地产生严重不利影响。在2050年代和2100年代的所有代表性浓度路径排放情景（RCPS）中，现今超过75%和80%的适宜栖息地将分别丧失。鉴于近年来印太驼背豚搁浅和死亡记录的增加，迫切需要制定严格的管理规定和保护计划，以保护当前的适宜栖息地。由于未来栖息地收缩和向极地转移，应制定适应性管理策略，包括设计新的保护区，并根据印太驼背豚的地理分布调整保护区的位置和范围，以尽量减少气候变化对该物种的影响。
