Vector activity and propagule size affect dispersal potential by vertebrates

Abstract

Many small organisms in various life stages can be transported in the digestive system of larger vertebrates, a process known as endozoochory. Potential dispersal distances of these "propagules" are generally calculated after monitoring retrieval in experiments with resting vector animals. We argue that vectors in natural situations will be actively moving during effective transport rather than resting. We here test for the first time how physical activity of a vector animal might affect its dispersal efficiency. We compared digestive characteristics between swimming, wading (i.e. resting in water) and isolation (i.e. resting in a cage) mallards (Anas platyrhynchos). We fed plastic markers and aquatic gastropods, and monitored retrieval and survival of these propagules in the droppings over 24 h. Over a period of 5 h of swimming, mallards excreted 1.5 times more markers than when wading and 2.3 times more markers than isolation birds, the pattern being reversed over the subsequent period of monitoring where all birds were resting. Retention times of markers were shortened for approximately 1 h for swimming, and 0.5 h for wading birds. Shorter retention times imply higher survival of propagules at increased vector activity. However, digestive intensity measured directly by retrieval of snail shells was not a straightforward function of level of activity. Increased marker size had a negative effect on discharge rate. Our experiment indicates that previous estimates of propagule dispersal distances based on resting animals are overestimated, while propagule survival seems underestimated. These findings have implications for the dispersal of invasive species, meta-population structures and long distance colonization events.

Fig. 1

The mean number of 2-, 3-, and 4-mm markers retrieved per hour (left y-axes) and the percentage retrieved per hour (right y-axes) from mallards (Anas platyrhynchos) in isolation, wading, or swimming treatment during a the active phase (i.e. the first 5 h of the experiment where wading and swimming birds were in the flume tank) and b the inactive phase (6–12 h after the active phase), n = 12 individuals per treatment group. The significant interaction between wade and swim treatments as found by the GLM in Table 1 is visible

Fig. 2

Cumulative percent of intact snail shells retrieved from mallards swimming, wading, or in isolation after 1–12 h propagule retention