Modelling biodiversity and marine conservation priorities in Mozambique
A carregar...
Autores
Supervisores
Data
2024-02
Título da revista
ISSN da revista
Título do Volume
Editora
Resumo
Descrição
The long coastline of Mozambique transitions from tropical to temperate regions and thus provides a good location to evaluate the efficacy of modelling approaches to conservation decision-making. This report describes a process for modelling marine species diversity in Mozambique with the aim of identifying where the key coral reef biodiversity areas are located and how these compare to both existing and proposed marine protected areas. To achieve this objective, we initially developed 5 different predictive models for two diverse taxa groups (fish and coral) based on field data from regional (WIO marine province) and national (Mozambique) sources combined with relevant environmental variables from satellites and shipboard measurements. First, this included two provincial models with data from the entire WIO region (n~1000 sites) corresponding to a unique set of selected environmental variables. Secondly, we evaluated three Mozambique data-only models (n<113 sites). Mozambique models 1 and 2 used the same environmental variables as selected in the WIO model but based only on fits to data from the Mozambique sites (n<113 sites). Model 3 selected variables independently of the WIO model and fits to Mozambique data only (filtered model). This report focuses on the results of the three Mozambique data-only models. Rankings and responses of the strengths and relationships of the environmental model variables were compared and used to map numbers of taxa in all 1180 cells of a Mozambique grid-map, including those without field data. These procedures and between-model comparisons allowed us to establish a measure of confidence in the number of taxa predictions. The models differed in their ranking and strengths of the specific variables and their responses to environmental conditions, producing somewhat different predictions at the reef cell level and subsequent maps. For instance, the Mozambique data-only models predicted more fish but less coral taxa than the WIO models, however, between-models variation in prediction were generally low. The largest between-model differences occurred along nearshore reefs in the Quirimbas Archipelago, reefs between the Cabo Delgado and Nampula provinces, and for numbers of fish in the southern Primeiras and Segundas Archipelago. The higher variation here may arise from the fact that Mozambique represents a transitional zone and that environmental variability is higher in these particular areas making these locations priorities for ground-truthing field studies. Additionally, further increasing the number of field studies is expected to improve the predictive strengths of these models, as indicated by the higher fits in the provincial models with more data (R2~0.55-65%). Overall, our findings indicated 19 priority locations for coral reefs and fish that were mostly concentrated in Cabo Delgado and Nampula provinces. Many of the identified priority locations overlapped with existing or proposed marine protected areas. Nevertheless, our approach suggests there is a need for more conservation efforts focused on northern Mozambique. This study also aimed at identifying areas with the highest potential to trigger Key Biodiversity Area (KBA1) status and where formal assessment should thus be prioritized. Considering that our models predict higher biodiversity in northern Mozambique, there is also a higher probability of including threatened species and increasing the potential to trigger KBA status. The maps produced by the models provide sufficiently low variation among models and thus clearly show where to focus future effort.