Modelling the distribution ranges of Houbara and Macqueen’s Bustards to optimize reinforcement plans and predict potential future impacts of climate change
PhD project
Species distribution models are increasingly used in many fields of conservation biology, ecology and evolution. They have been used to assess the impact of climate land use and other environmental changes on species distributions, to support conservation planning and reserve selection. Within these contexts, and acknowledging the huge number of precise locations where bustards have been recorded (field surveys, gps of birds, nests, dusplaying males, in winter, summer, …), developing niche models using the bustards databases could bring original and highly valuable knowledge for both conservation biology of the species and theoretical ecological mecanisms.
I. Obtaining predictive species distributions with niche suitability models
Aim: to use niche suitability models to predict the potential distribution of the two species
Tools: the BIOMOD plateform, developed under R; allows to run various modelling techniques within an ensemble forecast approach.
Data to use:
-occurrence locations of undulata (nests, displaying males, GPS locations)
-occurrence locations of macqueeni (from Argos data – summer – winter – eventually migration stopover)
-environmental layers: climate (from the Worldclim database), topography (from the USGS Hydro-1K database), vegetation cover (NDVI)
-Human land use
-spatial scale: 1×1 km pixels
-Field validation of houbara occurrence and densities
Expected results: for each pixel, a suitability value between 0 and 1 across the
-predicted range (= suitable range) for undulata in north Africa
-suitable summer range & suitable winter range for macqueeni
Applications:
-defining the best places where to reinforce, reintroduce of even introduce new birds (sites of higher suitability)
-testing the past release successes with the estimated suitability values
-test if the observed dispersal is directed towards pixels of higher suitability
Extensions: should be possible to also consider demographic parameters (if linked to climate so spatially variable) within the distribution models to infer abundance or population fate across the presence area.
II. Projecting these models under climate change scenarios
Aim: to project these models using future climate and land use scenarios
Tools: still the same BIOMOD plateform
Further data to use:
-environmental layers for future climate, from the IPCC scenarios; topography similar, no use; NDVI not available but evapo-transpiration estimated from the climatic variables.
-spatial scale: 50×50 km pixels
Expected results: for each pixel, a suitability value between 0 and 1 across the
-predicted future suitability for undulata in north Africa
-same for suitable summer range & winter range for macqueeni
-predicted range of extinction, of potential colonization to be confronted to dispersal ability
-any change in distance between breeding and winter ranges in macqueeni, hence longer or shorter migration distances; towards more sedentarity?
Applications:
-defining the best places where to introduce new birds in a context of climate warming and intense land use transformation (sites of higher suitability)
-testing if the past release successes are related to the increase or loss in suitability probability from current to future conditions
-test if the observed dispersal is directed towards pixels of a larger increase in suitability
Extensions: should be possible to also consider demographic parameters within the distribution models to infer future abundance or population fate across the future presence area.
Note
The student will be co-supervised by Frédéric Jiguet (Maître de Conférences MNHN, HDR at UPMC, Ecole Doctorale Diversité du Vivant, Email: fjiguet_at_mnhn.fr) and Yves Hingrat of the ECWP (Emirates Center for Wildlife Propagation; see http://www.ecwp.org/) (Yves HINGRAT, Coordinateur Recherche Ecologie et Conservation, Email : yhingrat_at_reneco-hq.org). The project is funded by the ECWP.