Observing Seed Removal: Remote Video Monitoring of Seed Selection, Predation and Dispersal

What happens to seeds once they have reached a surface – here referred to as post-dispersal seed fate (cf. Chambers and MacMahon, 1994) – is becoming an increasingly important question. A large body of literature is available evaluating postdispersal seed fate (reviewed in Hulme, 1998, 2002; Crawley, 2000; Moles and Westoby, 2003; see Hulme and Kollmann, Chapter 2, this volume; Vander Wall and Longland, Chapter 18, this volume). Many studies investigating post-dispersal seed fate, however, measure only seed removal from experimental plots, and implicitly or explicitly assume that removed seeds were killed by seedeaters (see Hulme and Kollmann, Chapter 2, this volume; Forget and Wenny, Chapter 23, this volume; Wenny, Chapter 21, this volume). This approach may be inadequate, as several recent studies in which removed seeds were tracked have shown that many of these seeds are secondarily dispersed rather than consumed (e.g. Vander Wall, 1992, 1997; Forget, 1993; Levey and Byrne, 1993; Brewer and Rejmanek, 1999; Hoshizaki et al., 1999; Jansen et al., 2002). Scatterhoarding animals eat some of the seeds they gather but bury most seeds as food reserves in scattered caches in the soil (e.g. Jansen et al., 2002; Jansen et al., 2004). Scatterhoarding may enhance seed survival and seedling establishment because it moves seeds away from the parent and siblings, secures seeds from seed predators such as insects and wild pigs, and puts seeds in situations that often enhance germination and establishment (Vander Wall, 1990). Cached seeds can only establish seedlings, however, if animals leave some proportion of their food reserves untouched. Several studies of secondary dispersal (e.g. Vander Wall, 1994; Jansen, 2003) have shown that scatterhoarded seeds may indeed escape consumption and develop into seedlings. Thus, seed survival and seedling establishment can occur not only right after primary dispersal, but also after one or more subsequent bouts of secondary dispersal (Price and Jenkins, 1986). Moreover, some studies of large-seeded species have observed seeds dying from desiccation or seed predation if they were not removed and buried by scatterhoarding animals (Shaw, 1968; Jansen et al., 2004). Hence, traditional studies of post-dispersal seed survival that simply measure seed removal are insufficient for estimating seed mortality (see