Individual-based modelling of carapa seed uptake in Nigerian Ngel Nyaki Forest with simulations of restoration strategies and searching time

Carapa oreophila is among the plant species that can be found in only a few places in Nigeria, such as the Ngel Nyaki Forest, north eastern Nigeria. Possessing various medicinal properties, C. oreophila, along with other plants and animal species, contributes in maintaining the forest ecosystems as well as checking the global warming. The elephants now extinct in the forest were dispersing C. oreophila seeds along with the African pouched rats. However, it was recently found that the dispersal by rodents is now more important since the C. oreophila seeds lost their major dispersers. When a rodent encounters a seed, if it is hungry, it eats the seed and destroys the kernel of the seed, but sometimes the rodent stores the seed in the forest litter for future use. This seed may germinate when the rodent did not locate the cache, perhaps due to more than enough stored food in the forest. The applications of this process includes restoring degraded landscapes with the aid of mixed restoration strategy, through which seeds are placed by humans and expected to be removed by rodents. Since rodents live in different habitats, they can provide services to the disturbed landscapes. Therefore, the objectives of this study include the determination of fates of C. oreophila seeds placed in the forest plot, the determination of a method through which the loss of seeds can be minimised when a mixed restoration experiment is to be conducted, and the estimation of rodent’s searching time when encountering a seed in the forest plot. To understand the fate of C. oreophila seed removed, the stochastic individual-based model (IBM), explicitly tracking the number of seeds removed, was developed through the pure death process (PDP) and Poisson counting process (PCP) together with their deterministic counterparts. The parameters of the models were estimated from the scatter-hoarding data obtained from different habitats of the forest, core, edge and fragment. We devised two random walk models, the symmetric and intermittent walks, to determine a method that can minimise the loss of seeds in restoration projects and estimate the rodents’ searching time. We found that large proportions of removed seeds were stored in the core and edge of the forest, while only a few of such seeds were cached in the fragment. Thus, if this low dispersal is recurrent over a long time, it could pose a threat of the plant population decline in the forest fragment. Our computer simulation suggests that the loss of seeds in mixed restoration projects can be minimised by placing the seeds randomly in the forest plots. We found that the rodents searching time follows the Gumbel probability distribution. Since the population of many large seeded plant species, such as C. oreophila plant, depends on seed dispersal, studying the seed removal process can help us identify possible places within the forest where seedlings can be established, through which the forest ecosystem can be maintained.

Text FS 2020 12 ir.pdf Download (940kB)

Actions (login required)

View Item