Dispersal of Honey Locust (Gleditsia triacanthos) Seeds by White-tailed Deer

The seminal International Primer for Ecological Restoration lists the integration of a habitat within its larger ecological landscape as a necessary attribute of a restored ecosystem (Society for Ecological Restoration 2004). Most plant species maintain this integration and connectivity through dispersal of seeds. Specifically, long-distance dispersal is a critical explanatory factor of plant metapopulation dynamics, colonization, and invasion (Cain et al. 2000). A plant species’ ability to reach favorable habitats often relies on long-distance dispersal events (Clark et al. 1998). Restored habitats are no exception to this generalization. Historically, long-distance dispersal mechanisms have been viewed as difficult or impossible to study. However, recently both genetic approaches, such as parentage analysis (Cain et al. 2000), and field-based approaches, such as sampling deer feces for viable seeds (Myers et al. 2004), have provided an increased understanding for the long-distance dispersal of plant species, and can potentially inform restoration efforts in the attempt to maintain plant habitat connectivity.

Anthropogenic changes at a landscape-wide scale, such as climate change and habitat fragmentation, may significantly shift the potential niche, and correspondingly the distribution, of many plant species. The current increase in global temperatures will make new habitats suitable, while simultaneously making some current habitats unsuitable (Gray and Hamman 2013). These changes will also likely affect animal-mediated dispersal. Landscape-wide changes can alter animal movement patterns, causing a shift (either an increase or decrease) in propagule rain received by sites (Hampe 2011). In light of these changes, a more robust understanding of basic plant dispersal mechanisms must be developed in order to predict how the ability of species to colonize suitable habitat, including target areas for restoration, will change.

A potentially important long-distance dispersal vector for plants in North America is the white-tailed deer (Odocoileus virginianus). As generalist herbivores, deer disperse many vascular species (Eycott et al. 2007), some of which are invasive (Williams 2003, Myers et al. 2004). Changes in climate, deer density, and habitat fragmentation can affect daily movement and migration patterns in deer (Nixon et al. 1991, Brinkman et al. 2005), and could in turn impact the dispersal of many plant species.

One native species that may rely on dispersal by deer is honey locust (Gleditsia triacanthos). Honey locust is a woody species native to the central United States, west of the Appalachian Mountains and east of Texas and South Dakota. It thrives in riparian habitats and is commonly cultivated for use as an ornamental, a windbreak, and a dietary supplement for livestock (Blair 1990, Wilson 1991). Fruits are produced from mid-September to mid-October, but continue to dehisce throughout the winter (Blair 1990). This phenology of honey locust may make it an attractive food for deer. Livestock, specifically cattle and sheep, are known to readily consume these fruits (Wilson 1991), but there is currently no evidence in the literature that supports the assertion that seeds are dispersed by deer (Blair 1990). Here, I provide the first direct evidence that deer disperse honey locust seeds by endozoochory.

This project grew out of a pilot study testing the relative efficacy of two methods for determining seed viability of the invasive shrub, Amur honeysuckle (Lonicera maackii). Deer fecal pellets were collected from December 2012 to January 2013 at the Miami University Ecology Research Center (ERC) in Butler County, Ohio. The ERC is a 69 hectare mosaic of secondary forests, agricultural plots, old fields, and mowed areas. An aerial survey in the vicinity of ERC estimated winter deer density to be 3.9 deer/ km² (T. Millette, Research Aviation and Mapping Science, LLC, unpub. data). Collected deer pellet groups (n = 29) were randomly divided in half. Each half pellet group was assigned one of two methods to test for germinable seeds: washing pellets with a 0.5mm sieve and removing seeds of interest or potting deer pellets intact and identifying emerging seedlings.

Figure 1.

Two honey locust seedlings germinating out of deer pellets placed on vermiculite.

I report here only seedlings germinating out of deer pellets potted intact, since seeds...