Plant characters related to slug feeding: a re‐analysis of feeding studies using trait data from botanical databases and literature

Slugs feed on some plant species but not on others for reasons that are not yet fully understood. We re‐examined the two largest studies on the preference of Arion lusitanicus for different plant species. Briner & Frank (1998) measured the consumption index of slugs (CIB) based on the dry leaf mass eaten per day per gram of slug mass. Kozlowski & Kozlowska (2009) also measured a consumption index of slugs (CIK) but on the basis of the area eaten per day per gram of slug mass. The two indices showed a moderate positive rank correlation. Briner & Frank (1998) found that slugs preferred annual to perennial plants and crops rather than wild species. On the other hand, Kozlowski & Kozlowska (2009) found no statistically significant difference in slug preference for annuals, biennials or perennials. Using botanical databases we selected plant traits related to slug herbivory and analysed 141 species. We first analysed the two data sets separately and found that CIB was higher for seedlings with a higher relative growth rate although this result was only marginally statistically significant. In addition, we found that annuals were consumed more than perennials and that cultivated plants were consumed more than wild plants. This is in agreement with the conclusions of Briner & Frank (1998). For CIK slugs consumed less from plants with high dry matter content (DMC). The use of classification and regression tree models solves the problem of analysing datasets with missing values. Regarding CIB, the tree models indicated that life cycle was the major branch. Within perennials, cultivated/wild origin of plants was the most important predictor of slug feeding. Within wild perennials slugs ate more from plants with low specific leaf area (SLA), i.e. relatively thick leaves. Within annuals we found the unexpected result that slugs ate more from plant species with either alkaloids or glucosinolates than from species without one of these compounds. For CIK, the results of tree models showed that slugs preferred shade‐intolerant plants to shade‐tolerant plant species. Within the shade‐tolerant group slugs fed more on leaves with low DMC. We only partially understand which factors make plants acceptable to slugs. We discuss this issue along with the pros and cons of the two methods used to calculate CI.