Potential of Myrica rubra seed dispersal by spitting out by sika deer during rumination and its effect on germination rate after dispersal

Zoysia japonica, a low growing grass, is tolerant to grazing and trampling. Kinkazan Island in northern Japan is inhabited by sika deer (Cervus nippon). The deer population increased in the 1970s at a shrine garden in the western part of the island, leading to expansion of Zoysia swards around the shrine garden and their colonization of remote open patches. The expansion around the garden is due to elongation of the rhizome, but expansion to remote places may be due to endozoochory by deer. This appears to be a good example of the “foliage is the fruit” hypothesis (the FF hypothesis) proposed by Janzen (1984; American Naturalist 123:338–353). To demonstrate this, we confirmed the expansion of the Zoysia swards and tested the traits of Zoysia by field surveys and indoor experiments. The Zoysia peduncles stood among the leaves, and sika deer fed on both the seeds and leaves. One deer fecal pellet contained about 20 seeds at its peak in June. In the feeding experiment, the survival rate through digestion was 38%. In the greenhouse experiment, the germination rate was 72%. An outdoor experiment showed that germination rates of the ingested seeds were 5% in a dark habitat and 58% in a bright habitat. Zoysia exhibited many of the traits presented by the FF hypothesis, and sika deer functioned as seed dispersal agents. Therefore, the Zoysia—sika deer relationship seems to support the FF hypothesis.

In a herbaceous community subjected to continual impacts of sika deer (Cervusnippon), I examined the effects of seed ingestion by deer on seeds by comparing the ripening and germination rates of seeds of two dominant species, Zoysia japonica and Hydrocotyle maritima, between seeds taken out of fecal pellets (deer‐ingested seeds) and mature seeds collected directly from living plants (control seeds). Seeds of Z. japonica were likely to have tolerance to ingestion from earlier periods of seed maturity. In contrast, only ripened seeds of H. maritima may have tolerance to ingestion. When the seeds ripened, the germination rates of two species did not differ significantly between deer‐ingested seeds and control seeds. Thus, although immature seeds may be crushed by ingestion, many mature seeds can be dispersed by sika deer with no alteration of germination rate. However, the other germination experiment showed that the germination rates were significantly higher for seeds of Juncus tenuis in artificially broken fecal pellets than for those kept confined in the pellets, and all seeds germinated from intact pellets were situated near the surface of the pellets. These results suggest that dung may physically prevent seeds inside from germinating and decomposition of dung enables herbaceous small seeds in the dung to germinate.

Deer affect their habitats in various ways. Many previous studies on the ecological functions of deer have investigated the impact of herbivory on vegetation but have rarely focused on other functions. In this study, we evaluated the seed dispersal function of the sika deer (Cervus nippon yesoensis) in Hokkaido Island, examining the process from foraging to germination. We (1) surveyed the composition of seeds dispersed by wild deer, (2) measured seed recovery rate and gut passage time by feeding trials using three fleshy‐fruited trees (Vitis coignetiae, Actinidia arguta, and Rosa rugosa), (3) conducted germination tests using defecated intact seeds, and (4) estimated seed dispersal distance and its seasonal changes using GPS movement data. In total, 73 seedlings of at least 10 species emerged from 181 fecal samples, each weighing ~1 g. The proportion of intact seeds after gut passage varied greatly with plant species and among trials (0.0%–55.0%). The germination rate of intact defecated seeds was similar to (R. rugosa and V. coignetiae) or higher (A. arguta) than that of manually washed seeds. The average seed dispersal distances during the resident period were 594 m for V. coignetiae and 610 m for A. arguta. The average seed dispersal distances during the migration period increased to 3140 m for V. coignetiae and 3617 m for A. arguta, and sometimes exceeded 18,000 m. These findings highlight that sika deer can disperse seeds over long distances, thus contributing to gene flow between distant populations and range expansion to suitable habitats.