The first step was selecting a set of forest bird species occurring in the Eastern Arc Mountains that could be used as model species within the designed modelling framework. A first requirement was availability of population demographic data. A second requirement for species selection was their dependency on montane evergreen forest. Seven bird species fulfilled these conditions: Arizelocichla masukuensis, A. milanjensis, Chamaetylas fuelleborni, Illadopsis rufipennis, Modulatrix stictigula, Phyllastrepus placidus, Sheppardia sharpei. These species differ in their dependence on intact forest, climate associations and dispersal capacity.
We modeled bird population dynamics using a sex specific stage-structured model as this would allow to restrict dispersal to the juveniles and to account for sex-biased dispersal. Survival rates were available to ConservaTEAM from a long-term population monitoring study in Tanzania. Fecundity had to estimated as no reliable data were available.
To model dispersal mechanistically in function of the landscape, RS requires a cost surface with each cell reflecting the relative preference of forest birds to select it. I created a set of cost surfaces that either reflected a relatively permeable or impermeable matrix.
Current climate and future predictions were generated based on remotely sensed climate data (MODIS for temperature and CHIRPS for rainfall). Data from 2006-2015 were used to calculate a set of 14 relevant climatic variables centered around 2010. Subsequently, projections for the future were calculated at decadal intervals for the period 2020-2090 according to two alternative emission pathways (RCP4.5 and RCP8.5) using a regional downscaled model (AFRICLIM).
Environmental suitability for each species was modeled using Maxent. This model is recognized to represent the best option when using presence-only-data. Species occurrences were compiled in close collaboration with ornithological experts in the EAM region and environmental suitability subsequently estimated using a set of relevant and uncorrelated climate variables across a range of model settings.
RangeShifter was used to evaluate the ecological effectiveness of a set of management scenarios on the persistence of P. cabanisi. In this study, recommendations from a stakeholder workshop in the Taita Hills (SE Kenya) were used to develop scenarios aimed at increasing species persistence in this mountain bloc. The study made use of a version of RS that did not yet incorporate effects of climate change. We found that the largest population increase was predicted to occur under scenarios increasing habitat area. However, the effectiveness was sensitive to spatial planning. Compared to adding one large patch to the habitat network, adding several small patches yielded mixed benefits: although overall population sizes increased, specific newly created patches acted as dispersal sinks, which compromised population persistence in some existing patches. The study nicely demonstrated that the effectiveness of spatial management is strongly driven by patterns of individual dispersal across landscapes.
One of the main objectives of the project was to develop conservation management scenarios together with stakeholders. However, due to limitations of RangeShifter in representing climate change encountered in the process, we had to change our plans regarding stakeholder engagement in formulating and designing management scenarios that could mitigate the impacts of climate change. However, I found it important to communicate with stakeholders involved in conservation planning in the EAM about the possible effects of climate change on biodiversity. For this workshop, I analysed further the climate change projections and calculated measures of climate loss for the different EAM mountain blocs. This workshop sparked a number of very interesting ideas and we are now in the process to incorporate these in ongoing work.
Publications:
Aben, J., Bocedi, G., Palmer, S.C.F. Pellikka, P., Strubbe, D., Hallmann, C., Travis, J.M.J. Lens, L., & Matthysen, E. 2016. The importance of realistic dispersal models in conservation planning: application of a novel modelling platform to evaluate management scenarios in an Afrotropical biodiversity hotspot. Journal of Applied Ecology 53: 1055-1065.
Travis, J.M.T. Watts, K., Palmer, S., Aben, J., Henry, R., Synes, N. & Atkinson, N. 2017. Computer-aided landscape design. Woodland Conservation News (Spring issue). Woodland Trust, UK.
Aben, J., Pellikka, P., & Travis, J.M.J. Accepted. A call for viewshed ecology: advancing our understanding of the ecology of information. Methods in Ecology and Evolution.
