C.T. de Wit Graduate School for Production Ecology and Resource Conservation, Wageningen University, 6708 PB Wageningen, THE NETHERLANDS.
Prosopis chilensis, Prosopis juliflora and Prosopis pallida were widely introduced in tropical drylands, where they have become naturalized and invasive. Invasion management is partially hampered by species misidentification because of morphological similarities between species and between species and their hybrids. In Kenya, species identification is even more challenging as several species were introduced within sites. Since biological invasions are invoked by either the susceptibility of a habitat to invasion (invasibility) or the invasive traits of a species (invasiveness), proper knowledge of these two factors is required for proper invasion management. In this dissertation, we used field, greenhouse and laboratory studies to: 1) evaluate invasibility of Turkwel riverine forest, and 2) invasiveness of Prosopis species by determining: 2) ecological impacts of Prosopis invasion, 3) underlying mechanisms for observed ecological impacts and 4) species composition and genetic diversity of Prosopis populations in Kenya.
Invasibility was evaluated by comparing Prosopis occurrence between undisturbed invasion resistant forests with invasion susceptible forest gaps; ecological impacts of Prosopis invasion determined by comparing herbaceous species and tree regeneration among Acacia tortilis, Prosopis and a mixture of A. tortilis and Prosopis species canopies; underlying mechanisms of observed ecological impacts determined by assessing the impact of A. tortilis and Prosopis litter on A. tortilis and Prosopis species seed germination and seedling growth; and species composition and genetic diversity of Kenyan Prosopis populations determined by comparing Kenyan populations at Bamburi, Bura, Isiolo, Marigat, Taveta and Turkwel with P. chilensis, P. juliflora and P. pallida references for relatedness and genetic diversity.
We found that both the forest and the forest gaps were equally susceptible to Prosopis invasion, suggesting that invasion was spontaneous and independent of the assumed habitat invasibility variation. The ecological impact study revealed reduction of ground vegetation cover, herbaceous species diversity and termination of A. tortilis regeneration by Prosopis. Termination of A. tortilis regeneration may be attributed to strong reduction of A. tortilis and Prosopis species seed germination by increasing Prosopis litter concentration in the soil in greenhouse studies. Relatedness studies between the Kenya Prosopis populations and the reference species revealed the clustering of P. chilensis with Taveta population, P. juliflora with Bura, Marigat and Isiolo and P. pallida with Bamburi. Results suggested that Turkwel population was a likely hybrid between P. chilensis and P. juliflora. Genetic diversity of populations at Bamburi, Isiolo and Marigat was higher than that of the reference species that they clustered with. Bamburi, Isiolo, Taveta and Turkwel populations revealed genetic uniqueness, as demonstrated by generation of private markers by a specific primer in each of the population. Prosopis juliflora and its hybrid occurred in areas currently classified as most invaded.
Based on the results, we concluded that P. juliflora and its hybrid are the most aggressive invaders and that riverine forest invasion was invoked more by species invasiveness and not habitat invasibility. The invasiveness traits found in our study were: 1) Allelopathy of Prosopis litter which reduces herbaceous species ground vegetation cover and herbaceous species diversity; and inhibited regeneration of A. tortilis. 2) Unique site adaptations of introduced germplasm evident from the genetic differentiation of introduced germplasm depending on site, and 3) hybridization that was evident from a P. chilensis–P. juliflora hybrid in Turkwel.