Name: Campos JC

Biodeserts supervisor: Brito JC

Title: Landscape connectivity and Remote Sensing applications for assessing biodiversity patterns in desert environments

Institution: University of Porto

Status: Completed



Landscape connectivity reduces population isolation and allow dispersal of individuals to areas with favourable climatic conditions. Assessing landscape connectivity patterns in desert environments may constitute a major priority for future conservation planning in those areas. Remote Sensing (RS) has proven to be an effective tool for studies focused in ecology, biodiversity distribution and conservation, given its power in detecting key drivers of biodiversity status/change across the globe. Also, RS provides effective representations of the landscape structure, which constitutes a significant source of information for improving landscape connectivity studies. These techniques are particularly useful for studying biodiversity patterns in extremely remote regions, such as such as deserts and arid regions, where field investigations are difficult to perform.

Global biodiversity is currently facing severe losses and many species are on the brink of extinction. Climate change and habitat fragmentation represent two major factors enhancing the catastrophic degradation of biodiversity worldwide. These impacts are normally exacerbated in extreme regions, such as deserts and arid regions. These regions are frequently regarded as naturally poor and homogenous regions, although comprising unique and fragile biodiversity in need of global attention. For instance, the Sahara-Sahel holds many endemics and relict faunal populations with different biogeographical origins, such as the West African crocodile (Crocodylus suchus). Saharan relict populations are subjected to effects of extreme isolation and are deeply exposed to extreme climatic oscillations.

Accordingly, the main objective of this thesis is to verify the importance of landscape connectivity for desert organisms and how current assessments of biodiversity patterns in desert environments can profit from the application of RS tools. Concretely, four crucial goals were delineated in order to achieve this principal objective: 1) Evaluate the current state of structural connectivity methods for application in ecology, evolution and conservation; 2) Attest potential contributions of RS to the assessment of biodiversity distribution patterns in global drylands; 3) Identify and map in detail important landscape features for the assessment of local biodiversity patterns in the West Sahara-Sahel; 4) Analyse the distribution and the effects of climatic oscillations on population structure and connectivity of West African crocodiles in the West Sahara-Sahel.

For achieving the first goal, it was developed a review encompassing basal theoretical concepts and listing major advantages and disadvantages of structural connectivity methods. This review is focused on insufficiently reviewed methods and on the most recent methodological developments for measuring structural connectivity. Additionally, insights concerning the applicability of these methods in ecology, evolution and conservation are provided, and future directions for improving landscape connectivity studies are also discussed. Future landscape connectivity studies should focus on the development of integrative frameworks. This may be accomplished by incorporating complementary methods/outputs that will improve our understanding of species-landscape relationships at structural and functional levels. Increased efforts on the development of computational resources and on data collection (e.g., genetic and movement data) will allow the employment of sophisticated methods to evaluate population connectivity and the preservation of landscape connectivity patterns.

For answering to the second goal, it was performed an assessment of the most useful RS bio-indicators for describing terrestrial vertebrate distribution patterns in drylands. Twenty seven bio-indicators (averages and inter-annual variability of seasonally aggregated proxies for net primary production, rainfall, soil moisture, rain use efficiency, and soil moisture use efficiency) derived from MERIS data under the European Space Agency - Diversity II Project are used to predict functional species richness of 739 terrestrial vertebrates distributed across five global drylands. RS-derived variables related to water availability, particularly precipitation and soil moisture, reveal closer relations with vertebrate species richness (both total and functional richness) in most of the analysed drylands, in comparison to estimates of primary productivity, rain-use efficiency and soil moisture-use efficiency. The strongest relations are observed between water availability indicators and distribution patterns of vertebrate species with small body size (particularly amphibians).

To accomplish the third goal, it was derived a land cover map of the West Sahara-Sahel (30x30 m resolution) based on a stepwise methodology for land cover assessment. A collection of GPS control points with detailed descriptive traits were used for Landsat image classification. A robust regional classification (83.9% points correctly classified) of land cover is obtained for the West Sahara-Sahel, particularly for the most arid regions. The final map is composed by a total of 18 classes, and might be used for mapping local biodiversity distribution and for improving the effectiveness of local conservation actions and management of natural resources for local communities.

Two works are developed to accomplish the last goal. Firstly, an update of the distribution, occupied habitats, population size, and threat factors of C. suchus in Mauritania is conducted. Data collected during five field expeditions to Mauritania (2011-2016), allowed the detection of several new crocodile localities (N=26), increasing by 27% the current number of all known locations. Several threat factors affecting crocodile populations and associated habitats were identified (N=11). Droughts and temperature extremes (100% localities affected) and water extraction for domestic use and nomadic grazing (94%) were the most frequently detected. Isolated crocodile populations are apparently vulnerable, especially in the northernmost areas of Mauritania (Tagant mountain), and future local conservation strategies are needed to assure the persistence of its fragile populations. Secondly, an evaluation of the role of climatic oscillations on the spatial and temporal landscape connectivity, genetic patterns and dispersal dynamics of C. suchus populations is performed. Genetic diversity, population structure and gene flow of crocodiles are measured, and historical (1980s) and contemporary (2010s) scenarios of landscape connectivity are evaluated, in order to verify if genetic patterns of crocodile populations are correlated with climate-driven patterns of connectivity. Crocodiles display higher levels of genetic isolation in the north of Mauritania, in comparison to the southern populations, indicating effects of isolation by distance. The signs of population connectivity exhibited by juvenile crocodiles are probably related with contemporary dispersal dynamics.
This thesis successfully demonstrates the importance of landscape connectivity for desert organisms and how assessments of biodiversity patterns in desert environments can be improved using RS tools. Interdisciplinary approaches, such as the example presented in this thesis, in which landscape connectivity methods/RS techniques/genetic analyses are combined for assessing local population connectivity of West African crocodiles, are strongly recommended for supporting solid conservation frameworks, principally in overlooked and biodiversity rich ecosystems like deserts environments.