Abstract Phenotypic variation in behavior exists among species and populations, as well as among and within individuals. The pace-of-life syndrome hypothesis predicts covariation between life-history strategies, ranging from slow to fast, and behavior, ranging from shy, inactive, and flexible to bold, active, and less flexible. This covariation is expected to exist at multiple hierarchical levels, from the species down to the individual. We predict that fast-lived species will differ in average levels of behavior, and additionally show lower within-individual and among-individual variation than slow-lived ones. Shrews represent a highly suitable model to test these predictions, as they comprise a range of genera which differ tremendously in life-history strategy and metabolism. We performed repeated tests of boldness and aggression on 155 wild-caught individuals of five species of shrews, two species of the slow-lived genus Crocidura , two of the fast-lived genus Sorex , and one of the intermediate-paced genus Neomys . To compare not only average levels of behavior but also its variance components between those groups, we calculated coefficients of variation at within- and among-individual levels. Our results support our first prediction that, following the framework of pace-of-life-syndromes, fast-lived species should exhibit bolder behavior than slow-lived ones. However, our prediction of lower within- and among-individual variation in fast-lived species was not supported. Instead, our data suggest that other ecological factors might influence the expression of behavioral variation in shrew species, such as the variability in habitat choice and differences in anti-predator strategies. Significance statement The behavior and life history of animals are often structured into so-called pace-of-life syndromes (POLS), with slow-lived individuals being rather shy, inactive, and flexible and fast-lived individuals rather bold, active, and less flexible. Comparing the behavior in five species of shrews, we tested if such a gradient can also be found on the species level. While the average levels of species’ behavior indeed matched their pace of life, their individual behavior and behavioral stability did not. It was rather explained by an interplay of ecological and physiological factors, among them the variability in habitat choices and differences in anti-predator strategies. Our study shows that behavioral variation cannot be explained by just one factor like POLS at different hierarchical levels, but rather by a combination of factors including the animals’ life-history and ecological and physiological background.

中文翻译：

---

五种鼩鼱物种内部和之间的个体行为、行为稳定性和生活节奏

摘要 行为的表型变异存在于物种和种群之间，以及个体之间和个体内部。生活节奏综合症假说预测了生活史策略（从缓慢到快速）和行为（从害羞、不活跃和灵活到大胆、积极和不那么灵活）之间的协变。预计这种协变存在于多个层次级别，从物种到个体。我们预测，快活物种的平均行为水平会有所不同，此外，与慢活物种相比，个体内部和个体间的差异会更低。鼩鼱代表了一个非常适合测试这些预测的模型，因为它们包含一系列在生活史策略和新陈代谢方面差异巨大的属。我们对 155 名野生鼩鼱、两种慢活属 Crocidura、两种快活属 Sorex 和一种中等速度的 Neomys 进行了大胆和侵略性的重复测试。为了不仅比较这些组之间的平均行为水平，还比较其方差分量，我们计算了个体内部和个体间的变异系数。我们的结果支持我们的第一个预测，即遵循生命节奏综合症的框架，快活物种应该比慢活物种表现出更大胆的行为。然而，我们对快活物种个体内部和个体间差异较低的预测没有得到支持。相反，我们的数据表明，其他生态因素可能会影响鼩鼱物种行为变异的表达，例如栖息地选择的可变性和反捕食者策略的差异。意义声明 动物的行为和生活史通常被划分为所谓的生活节奏综合症 (POLS)，慢活的个体相当害羞、不活跃，而灵活和快活的个体则相当大胆、活跃和不太灵活。比较五种鼩鼱的行为，我们测试了这种梯度是否也可以在物种水平上找到。虽然物种行为的平均水平确实与它们的生活节奏相匹配，但它们的个体行为和行为稳定性却没有。而是通过生态和生理因素的相互作用来解释的，其中包括栖息地选择的可变性和反捕食者策略的差异。