Small islands of short, fire-avoiding forest exist in a wider sea of fire-prone shrublands in the southwestern Cape of South Africa. These forests are restricted to growing in extremely rocky habitats, such as on cliffs and amongst rock outcrops and screes. The dynamics of these forests is different to that of the shrublands, where post-fire recruitment from long-lived seed banks of non-fleshy fruits dominates. For example, forest tree fruits are generally soft and non-dormant. Heeria argentea and Hartogiella schinoidesare important constituents of these forests. They comprise narrowly endemic taxa from mono-specific genera, considered to be palaeo-remnants of a tropical flora. The question arises as to what disperses these fruits; birds have not been observed to do this and are unlikely dispersers because fruits are dull-colored and leathery.
We investigated the dispersal of these two species, using camera traps to identify animals interacting with fruits and reverse-wound cotton threads to track fruits and determine fates. Germination tests were done on eaten H. argentea fruits. Lastly, we measured the impact of recent fires on H. argentea and a co-occurring, shrubland tree species in relation to degree of rockiness and tree size.
We determined that fruit removal of H. argentea and H. schinoides is only by the nocturnal Namaqua rock rat (Micaelamys namaquensis). This rock-dwelling, generalist rodent moves the fruits to rock crevices before eating only the thin, fleshy pericarp of the fruit. Experiments with eaten and control H. argentea fruits show lack of germination in control fruits. We show the benefits of this directed dispersal is to ensure establishment and growth in rocky, fire refugia.
This is a clear example of directed dispersal, with disproportionate arrival of fruits to rare sites that provide an increased survival advantage in these fire-swept landscapes. These plants are entirely dependent on this single rodent frugivore for both dispersal and germination. Since these trees are locally endemic and this mutualism appears highly adapted to survival in rock habitats of the Cape, it suggests these are not palaeo-remnants with primitive fruit characteristics from a tropical flora. We suggest that a community of Cape trees has converged in distribution and dispersal mechanism due to the selective pressure of fire. There are a number of other rock-restricted plants with similar life-history and fruit traits both regionally and globally, which may utilise similar dispersal strategies.