Data from: Habitat monitoring and projections for threatened Canada lynx: linking the Landsat archive with carnivore occurrence and prey density

Habitat degradation is the leading threat to US federally listed species, but monitoring changes in habitat extent, fragmentation or quality is not generally included in species recovery plans. Consecutive satellite images of the same area through time are commonly used to assess landscape changes, and image data are increasingly used to generate predictor variables for species–habitat models. A time series of satellite images has, however, rarely been used in conjunction with a species–habitat model to map changes in habitat extent or to predict wildlife–habitat dynamics through time. We used a time series of forest disturbance encompassing 16 530 km2 of forestland in Maine to map and quantify change in forest types that influence probability of occurrence (POC) of the threatened Canada lynx Lynx canadensis. Integration with a predictive occurrence model developed from landscape conditions c. 2004 allowed us to retrospectively (1973–2004) assess changes at 2- to 4-year intervals, update predictions (2004–2010) and project the spatial distribution and extent of lynx foraging habitat over the next decade. Between 1975 and 1991, 40% of mature conifer forest in our study area was removed by stand-replacing harvests, coincident with the last regional outbreak of eastern spruce budworm Choristoneura fumiferana (Clem.). As those areas regenerated between 1991 and 2010, POC increased dramatically as broad-scale habitat conditions for lynx's primary prey, snowshoe hare Lepus americanus, improved. Projections suggest that continued forest regrowth will provide new habitat over the next decade, but fragmentation will increase, and by 2022 the habitat distribution will shift southwards into areas where climate change or interspecific competition may limit lynx population expansion. Synthesis and applications. Conservation planning for vagile species that depend on ephemeral forest types, such as the Canada lynx in the north-eastern USA, requires habitat monitoring strategies that can adapt to changing landscape conditions. Although Maine's lynx population has increased since the 1990s, we recommend that continued monitoring of lynx habitat and population status is critical given habitat and climate uncertainty. We also recommend that habitat management efforts be targeted not only at areas currently occupied by lynx but also where habitat loss is expected to occur near-term from succession or in anticipation of the next budworm outbreak.生境退化是威胁美国联邦保护物种的首要因素，但针对生境范围、破碎化程度或质量变化的监测，通常并未纳入物种恢复计划之中。 研究人员通常利用同一区域的连续时序卫星影像评估景观变化，且愈发多地借助影像数据为物种-生境模型生成预测变量。然而，时序卫星影像鲜少与物种-生境模型结合，用于绘制生境范围变化图谱，或是动态预测野生动物与生境的时空变化关系。 本研究针对美国缅因州16530平方千米的林地，利用森林干扰时序数据，绘制并量化了影响受威胁物种加拿大猞猁（Lynx canadensis）出现概率（probability of occurrence, POC）的森林类型变化。结合基于约2004年景观条件构建的预测性出现模型，我们得以回溯评估1973年至2004年间每2至4年的生境变化、更新2004年至2010年的生境预测结果，并对未来十年内猞猁觅食生境的空间分布与范围进行推演。 1975年至1991年间，研究区域内40%的成熟针叶林因皆伐作业被移除，这一时间恰逢东部云杉芽卷蛾（Choristoneura fumiferana (Clem.)）最后一次区域性大暴发。1991年至2010年间，受干扰区域逐渐恢复，猞猁的主要猎物雪兔（Lepus americanus）的大范围生境条件得到改善，猞猁的出现概率随之大幅提升。模型推演结果显示，未来十年内森林持续恢复将催生新的生境，但生境破碎化程度将加剧；至2022年，生境分布将向南迁移至气候变化或种间竞争可能限制猞猁种群扩张的区域。 研究总结与应用启示：对于依赖短命森林类型的漫游性物种（如美国东北部的加拿大猞猁），其保护规划需要能够适配景观动态变化的生境监测策略。尽管缅因州的猞猁种群自1990年代以来有所恢复，但鉴于生境与气候的不确定性，持续监测猞猁生境与种群现状仍至关重要。此外，生境管理工作不仅应聚焦于猞猁当前的栖息区域，还应针对因群落演替即将发生生境丧失的区域，或是为应对下一次云杉芽卷蛾暴发提前规划的区域开展。