Secondary dispersal of forest herb seeds from raccoon dung: contrasting service by multiple vectors

Abstract

Secondary movement of seeds deposited in dung potentially improves their chances of seedling establishment but has rarely been documented in temperate-zone deciduous forest. We followed mayapple (Podophyllum peltatum) seeds after deposition in raccoon (Procyon lotor) dung to test for secondary dispersal in a long-established deciduous forest in southeastern Ohio, USA. Marked seeds were incorporated into raccoon dung, placed in natural forest sites, and observed at regular intervals until the subsequent Spring. Abiotic seed movement was monitored during natural and simulated rain events. To identify animal vectors and describe their interaction with seeds, motion-triggered wildlife cameras were trained on dung with embedded mayapple seeds. Seed movement was initially limited by adhesion in the viscid dung matrix, but 15% of seeds eventually moved >20 cm from the point of placement, a minimal estimate of a shoot’s competitive neighborhood (maximum movement distance = 130 cm). Significantly greater movement occurred on slopes and in the absence of leaf litter. Rainfall moved seeds as far as 51 cm. White-footed mice (Peromyscus leucopus) and chipmunks (Tamias striatus) began removing seeds within 24 h of dung placement, presumably caching a portion of them outside the 1.5-m radius of observation. At least four other animal species interacted with dung in ways unrelated to seed presence, which nevertheless resulted in seed movement on a scale of centimeters. Thus, multiple secondary processes caused seed movement, each resulting in a distinctive seed distribution. The forest herb potentially benefits through reduction in seed density and distribution to a variety of microsites.