School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York, United States of America.
PLoS One. 2012;7(8):e43167. doi: 10.1371/journal.pone.0043167. Epub 2012 Aug 24.
Predictive habitat models can provide critical information that is necessary in many conservation applications. Using Maximum Entropy modeling, we characterized habitat relationships and generated spatial predictions of spinner dolphin (Stenella longirostris) resting habitat in the main Hawaiian Islands. Spinner dolphins in Hawai'i exhibit predictable daily movements, using inshore bays as resting habitat during daylight hours and foraging in offshore waters at night. There are growing concerns regarding the effects of human activities on spinner dolphins resting in coastal areas. However, the environmental factors that define suitable resting habitat remain unclear and must be assessed and quantified in order to properly address interactions between humans and spinner dolphins. We used a series of dolphin sightings from recent surveys in the main Hawaiian Islands and a suite of environmental variables hypothesized as being important to resting habitat to model spinner dolphin resting habitat. The model performed well in predicting resting habitat and indicated that proximity to deep water foraging areas, depth, the proportion of bays with shallow depths, and rugosity were important predictors of spinner dolphin habitat. Predicted locations of suitable spinner dolphin resting habitat provided in this study indicate areas where future survey efforts should be focused and highlight potential areas of conflict with human activities. This study provides an example of a presence-only habitat model used to inform the management of a species for which patterns of habitat availability are poorly understood.
预测性栖息地模型可以提供许多保护应用所必需的关键信息。本研究采用最大熵模型，描述了夏威夷群岛主要岛屿上飞旋海豚（Stenella longirostris）休息栖息地的生境关系，并进行了空间预测。夏威夷的飞旋海豚表现出可预测的每日活动模式，白天在近岸海湾作为休息栖息地，晚上在近海觅食。人们越来越关注人类活动对沿海地区休息的飞旋海豚的影响。然而，定义适宜休息栖息地的环境因素仍不清楚，必须进行评估和量化，以便正确处理人类与飞旋海豚之间的相互作用。我们使用了最近在夏威夷群岛主要岛屿上进行的一系列海豚目击调查的数据，以及一系列假设对休息生境很重要的环境变量，来对飞旋海豚的休息生境进行建模。该模型在预测休息生境方面表现良好，表明与深海觅食区的接近程度、水深、浅水深湾的比例和粗糙度是飞旋海豚栖息地的重要预测因子。本研究提供了一个适宜飞旋海豚休息的栖息地预测位置，这些位置表明未来调查工作应集中在哪些地区，并突出了与人类活动潜在冲突的区域。本研究提供了一个仅基于存在数据的栖息地模型的示例，用于为一种对其栖息地可利用性模式了解甚少的物种提供管理信息。
