Large-area imaging reveals biologically driven non-random spatial patterns of corals at a remote reef
Large-area imaging reveals biologically driven non-random spatial patterns of corals at a remote reef
复制标题
DOI：
10.1007/s00338-017-1624-3
复制
发表时间：
2017-12-01
期刊：
影响因子：
3.5
通讯作者：
Sandin, Stuart A.
中科院分区：
文献类型：
--
作者：
Edwards, Clinton B.;Eynaud, Yoan;Sandin, Stuart A.
对于诸如造礁珊瑚之类的无根生物，个体在景观上分散程度的差异可能是生活史策略的重要差异造成的，也可能反映了栖息地可用性的模式。因此，对空间格局的描述不仅有助于确定构成生态系统的关键生物和物理机制，而且还有助于提供产生和检验生态理论所需的数据。在这里，我们使用原位成像技术在太平洋中部的Palmyra Atoll创建了16个样地的大面积显微图像，每个样地覆盖了100米(2)的底栖动物栖息地。我们绘制了44,008个珊瑚群落的位置，并将每个群落识别到尽可能低的分类水平。使用空间离散度量，我们测试了偏离空间随机性的情况。我们还使用目标模型拟合来探索导致分类群空间格局差异的候选过程。多数分类群呈聚类，聚类程度因分类群而异。少数类群没有明显偏离随机性，也没有显示出空间均匀性的证据。重要的是，容易分裂或通过部分死亡耐受压力的分类群更聚集。几乎没有例外，聚类模式与破碎化和分散限制模式一致。在一些分类群中，散度与丰度呈线性相关，表明空间格局的密度依赖性。石珊瑚的空间格局是非随机的，反映了该分类群的基本生活史特征，这表明在许多情况下，珊瑚礁景观可能具有重要的空间可预测性。
For sessile organisms such as reef-building corals, differences in the degree of dispersion of individuals across a landscape may result from important differences in life-history strategies or may reflect patterns of habitat availability. Descriptions of spatial patterns can thus be useful not only for the identification of key biological and physical mechanisms structuring an ecosystem, but also by providing the data necessary to generate and test ecological theory. Here, we used an in situ imaging technique to create large-area photomosaics of 16 plots at Palmyra Atoll, central Pacific, each covering 100 m(2) of benthic habitat. We mapped the location of 44,008 coral colonies and identified each to the lowest taxonomic level possible. Using metrics of spatial dispersion, we tested for departures from spatial randomness. We also used targeted model fitting to explore candidate processes leading to differences in spatial patterns among taxa. Most taxa were clustered and the degree of clustering varied by taxon. A small number of taxa did not significantly depart from randomness and none revealed evidence of spatial uniformity. Importantly, taxa that readily fragment or tolerate stress through partial mortality were more clustered. With little exception, clustering patterns were consistent with models of fragmentation and dispersal limitation. In some taxa, dispersion was linearly related to abundance, suggesting density dependence of spatial patterning. The spatial patterns of stony corals are non-random and reflect fundamental life-history characteristics of the taxa, suggesting that the reef landscape may, in many cases, have important elements of spatial predictability.