Understanding behavioural responses to human-induced rapid environmental change: A meta-analysis

Behavioural responses are often the first reaction of an organism to human
induced rapid environmental change (HIREC), yet current empirical evidence
provides no consensus about the main environmental features that animals
respond to behaviourally or which behaviours are responsive to HIREC. To
understand how changes in behaviour can be predicted by different forms of
HIREC, we conducted a meta-analysis of the existing empirical literature
focusing on behavioural responses to five axes of environmental change
(climate change, changes in CO2, direct human impact, changes in nutrients
and biotic exchanges) in five behavioural domains (aggression,
exploration, activity, boldness and sociability) across a range of taxa
but with a focus on fish and bird species. Our meta-analysis revealed a
general absence of directional behavioural responses to HIREC. However,
the absolute magnitude of the effect sizes was large. This means that
animals have strong behavioural responses to HIREC, but the responses are
not clearly in any particular direction. Moreover, absolute magnitude of
the effect sizes differed between different behaviours and different forms
of HIREC:  Exploration responded more strongly than activity, and
climate change induced the strongest behavioural responses. Model
heterogeneities identified that effect sizes varied primarily because of
study design, and the specific sample of individuals used in a study;
phylogeny also explains significant variation in our bird model. Based on
these results, we make four recommendations to further our
understanding:  1) a more balanced representation of laboratory
and field studies, 2) consideration of context dependency, 3)
standardisation of the methods and definitions used to quantify and study
behaviours, and 4) consideration of the role for individual differences in
behaviour.

行为响应通常是生物体对人类诱导的快速环境变化（HIREC）的第一反应，但当前实证证据尚未就动物在行为层面响应的核心环境特征，或是哪些行为会对HIREC产生响应达成共识。为探明不同类型HIREC如何预测行为变化，我们对现有实证文献开展了一项元分析，聚焦多类群动物在五大行为维度（攻击行为、探索行为、活动行为、大胆性与社会性）下对五类环境变化（气候变化、二氧化碳浓度变化、人类直接干扰、营养盐变化与生物交互变化）的行为响应，研究涵盖多类群动物且重点关注鱼类与鸟类。本元分析结果显示，动物对HIREC的行为响应总体上不存在明确的方向性，但效应量的绝对数值却十分可观——这表明动物对HIREC存在强烈的行为响应，只是这类响应并未偏向某一特定方向。此外，不同行为类型与不同HIREC类型间的效应量绝对数值存在差异：探索行为的响应强度高于活动行为，而气候变化引发的行为响应最为显著。模型异质性分析结果表明，效应量的差异主要源于研究设计与研究所采用的特定个体样本；系统发育因素同样能解释鸟类模型中存在的显著变异。基于上述结果，我们提出四项建议以深化相关研究认知：1）平衡实验室研究与野外研究的占比；2）考量情境依赖性；3）统一行为量化与研究所用的方法与定义；4）关注行为个体差异的作用。