Coral reef communities exposed to rapid temperature rises and frequent thermal anomalies were evaluated for taxonomic turnover via presence/absence information over a 27-year period experiencing large changes in the dominant taxa. Temporal turnover of the taxa within sites was consistently high (~ 40%) due to both inter-annual episodic and directional changes. Turnover with time displayed a rapid increase and slow decline after sequential cool and warm thermal anomalies between 1996 and 1998. Subsequent warm temperature anomalies caused fewer broad-scale changes. Directional change for all sites combined indicated three overall gains and losses in taxa—Montipora being the only dominant taxon that declined in both abundance and presence/absence. The studied marine reserves had higher local but lower between-site taxonomic richness than fished reefs. Despite similar mean turnover, there were fewer gains than losses in marine reserves (7 gains and 20 losses in 5 sites) than fished sites (16 gains and 15 losses in 7 sites). Changes in taxonomic cover and presence/absence turnover data were not correlated, indicating that turnover detects finer scale taxonomic change likely to be missed when the cover of the dominant taxa is evaluated—especially in the higher richness marine reserves. High spatial richness, community change, and thermal acclimation in these shallow reef lagoons may have prevented higher net losses of taxa. Consequently, the probabilities of reducing local extirpations of taxa may be best achieved by planning and management that promotes spreading more evenly-spread access restrictions to reef areas with high between-site diversity rather than focusing restrictions to sites with high within-site diversity.

中文翻译：

---

热应激珊瑚分类群的十年更替支持海洋保护区设计的风险分散方法

在 27 年的主要分类群经历了巨大变化的情况下，通过存在/不存在信息评估了暴露于快速升温和频繁热异常的珊瑚礁群落的分类周转。由于年际间的情节和方向变化，站点内分类群的时间周转率一直很高（~40%）。1996-1998年间连续出现冷暖热异常后，随着时间的推移，周转量呈现快速增加和缓慢下降的趋势。随后的暖温异常引起的大范围变化较少。所有地点的方向变化加起来表明分类群的三个整体收益和损失 - 蒙蒂波拉是唯一在丰度和存在/不存在方面都下降的主要分类群。所研究的海洋保护区比鱼礁具有更高的本地但更低的站点间分类丰富度。尽管平均周转率相似，但海洋保护区（5 个地点的 7 个增益和 20 个损失）的收益少于捕捞地点（7 个地点的 16 个收益和 15 个损失）。分类覆盖的变化与存在/不存在周转数据不相关，这表明周转检测到在评估主要分类群的覆盖时可能会遗漏的更精细的分类变化——尤其是在更丰富的海洋保护区。这些浅礁泻湖的高空间丰富度、群落变化和热适应可能阻止了更高的分类群净损失。最后，