Abstract Many rigorous methods have been developed to manipulate and sample rocky benthic species, but development of field methodology is less advanced for disturbance-prone rocky habitats e.g. cobbles and boulders. These habitats can have ecological and economic importance but research is complicated by their complex physical structure that necessitates causing disturbance when sampling, possibly resulting in mobile species being lost from samples. We used simulated/controlled sampling conditions to determine the proportions of mobile species successfully captured or lost during sampling. Boulders and cobbles were lifted out of shallow water (i.e. “littoral” sampling) and sampled in simulated sublittoral conditions, targeting biota attached to the undersurfaces. Variation in sample losses associated with different methods was tested – sampling in shallow vs deep water and small vs large boulders/cobbles. Different species and locations were also compared. Tests were done to determine contexts producing minimal losses and thus deserving prioritisation. Experiments were replicated in southern Australia and NE Baltic Sea. Large proportions of littoral amphipods and isopods were not successfully captured during sampling in most contexts (overall mean loss = 35%). This means lifting boulders from shallow water cannot accurately sample them. This finding is important especially for highly researched boulder or cobble associated invasive amphipods with large ecological impacts. Seventeen percent of littoral gastropods were lost from samples in both regions; in Australia greater proportions were occasionally lost when sampling larger boulders. Chitons and one gastropod (Theodoxus fluviatilis) could be sampled from shallow water with consistently high reliability. During sublittoral sampling, mean sample loss was 13% with no significant interspecific variability. Boulder/cobble size and water depth rarely caused influence. The results reveal sampling contexts where more sophisticated methods may be used for highly accurate abundance measurements in recent research programmes of ecological restoration and conservation of rare species involving marine but also freshwater habitats.

中文翻译：

---

巨石和鹅卵石礁中常见底栖大型动物采样方法的有效性

摘要 已经开发了许多严格的方法来操纵和采样岩石底栖物种，但是对于易受干扰的岩石栖息地，例如鹅卵石和巨石，野外方法的开发还不太先进。这些栖息地可能具有生态和经济重要性，但由于其复杂的物理结构，在采样时需要引起干扰，可能导致移动物种从样本中丢失，因此研究变得复杂。我们使用模拟/控制采样条件来确定在采样过程中成功捕获或丢失的移动物种的比例。将巨石和鹅卵石从浅水区取出（即“沿海”采样）并在模拟的亚沿海条件下进行采样，目标是附着在水下的生物群。测试了与不同方法相关的样本损失变化——在浅水与深水以及小与大巨石/鹅卵石中取样。还比较了不同的物种和地点。进行了测试以确定产生最小损失并因此应优先考虑的上下文。在澳大利亚南部和波罗的海东北部重复进行了实验。在大多数情况下，在采样期间未能成功捕获大部分沿海片脚类动物和等足类动物（总体平均损失 = 35%）。这意味着从浅水中提升巨石无法准确地对它们进行采样。这一发现对于高度研究的巨石或鹅卵石相关的具有巨大生态影响的侵入性片脚类动物尤其重要。这两个地区的样本中丢失了 17% 的沿海腹足动物；在澳大利亚，当对较大的巨石取样时，偶尔会丢失更大的比例。Chitons 和一种腹足动物 (Theodoxus flaviatilis) 可以从浅水中以始终如一的高可靠性采样。在沿海采样期间，平均样本损失为 13%，没有显着的种间变异。巨石/鹅卵石大小和水深很少造成影响。结果揭示了采样环境，在最近涉及海洋和淡水栖息地的稀有物种的生态恢复和保护研究计划中，可以使用更复杂的方法进行高度准确的丰度测量。平均样本损失为 13%，没有显着的种间变异。巨石/鹅卵石大小和水深很少造成影响。结果揭示了采样环境，在最近涉及海洋和淡水栖息地的稀有物种的生态恢复和保护研究计划中，可以使用更复杂的方法进行高度准确的丰度测量。平均样本损失为 13%，没有显着的种间变异。巨石/鹅卵石大小和水深很少造成影响。结果揭示了采样环境，在最近涉及海洋和淡水栖息地的稀有物种的生态恢复和保护研究计划中，可以使用更复杂的方法进行高度准确的丰度测量。