Successfully locating a host plant is crucial for an insect herbivore to feed and/or oviposit. However, locating a host within a complex environment that may contain an array of different plant species is a difficult task. This is particularly the case for polyphagous herbivores, which must locate a host within environments that may simultaneously contain multiple suitable and unsuitable hosts. Here we review the mechanisms of host selection used by polyphagous herbivores, as well as exploring how prior experience may modify a generalist's response to host cues. We show that recent research demonstrates that polyphagous herbivores have the capacity to detect both common cues from multiple host species, as well as specific cues from individual host species. This creates a paradox in that generalists invariably rank hosts when given a choice, a finding at odds with the “neural limitations” hypothesis that says generalist insect herbivores should not have the neural capacity to identify cues specific to every possible host. To explain this paradox we propose a model, akin to parasitoid host location, that postulates that generalist herbivores use different cues sequentially in host location. We propose that initially common host cues, associated with all potential hosts, are used to place the herbivore within the host habitat and that, in the absence of any other host cues, these cues are sufficient in themselves to lead to host location. As such they are true “generalist” cues. However, once within the host habitat, we propose that the presence of a smaller group of cues may lead to further host searching and the location of preferred hosts: these are “specialist” cues. This model explains the current conflict in the literature where generalists can respond to both common and specific host‐plant cues, while also exhibiting specialist and generalist host use behavior under different conditions.

中文翻译：

---

“顺序线索假设”：解释宿主位置和多食草食动物排序的概念模型

成功定位寄主植物对于昆虫食草动物进食和/或排卵至关重要。然而，在可能包含一系列不同植物物种的复杂环境中定位宿主是一项艰巨的任务。对于多食性食草动物来说尤其如此，它们必须将宿主定位在可能同时包含多个合适和不合适宿主的环境中。在这里，我们回顾了多食食草动物使用的宿主选择机制，并探讨了先前的经验如何改变通才对宿主线索的反应。我们表明，最近的研究表明，多食性食草动物有能力检测来自多个宿主物种的共同线索，以及来自单个宿主物种的特定线索。这造成了一个悖论，即通才在有选择时总是对主机进行排名，这一发现与“神经局限性”假说不一致，即一般昆虫食草动物不应该具有识别每个可能宿主特定线索的神经能力。为了解释这个悖论，我们提出了一个类似于寄生宿主位置的模型，该模型假设通才食草动物在宿主位置顺序使用不同的线索。我们建议，最初与所有潜在宿主相关的共同宿主线索用于将食草动物放置在宿主栖息地内，并且在没有任何其他宿主线索的情况下，这些线索本身就足以导致宿主位置。因此，它们是真正的“通才”线索。然而，一旦进入宿主栖息地，我们建议存在一小部分线索可能会导致进一步的宿主搜索和首选宿主的位置：这些是“专业”线索。