With the implementation of the Renewable Energy Independent Power Producers Procurement Programme, just under 3.5 gigawatt of wind energy have been installed in South Africa between 2010 and 2020 \citep{IPPPP2019}. Renewable energy development is planned to continue to grow and replace the aging energy production infrastructure in the country. At the same time, South Africa boasts one of world's biodiversity hotspots \citep{Myers_2000} and significant efforts are in place to conserve these natural values while boosting economic development (NBSAP 2015-2025). Among these biodiversity riches are 79 species of long-lived raptors of which 25% are classified as threatened in South Africa \citep{Taylor2015}. We focus our work on the Black Harrier Circus maurus, one of the world's rarest harriers (Simmons and Simmons 2000), that exhibits very low genetic diversity \citep{Fuchs2014}. This species is endemic to Southern Africa, with c. 99% of its population in South Africa \citep{Simmons2005}, and has undergone a long-term reduction in numbers due to habitat loss over the last century \citep{Curtis2004,Taylor2015}. It is therefore important to predict the consequences of impacts associated with wind energy production.
Elsewhere in the world, the effect of wind energy facilities on other harriers (Circus spp.) appears to have been relatively minor. Few fatalities are recorded for North American Northern Harriers Circus hudsonius (7 deaths in 17 years at Altamont: \citep{b2008} and from the three Circus species found in Europe \citep{frank2011,Wilson_2016}. However, in the secondary literature there are over a hundred mortality records of European harriers \citep{2006,a2014,brandeburg2020}, putting them in the top 25% of raptors experiencing wind farm deaths \cite{brandeburg2020}. The reason why relatively few harrier fatalites have been reported could be that they "show particular caution around turbines" \citep{Smallwood_2009,Garvin_2010} or breeding birds tend to move away from operational wind farms \citep{Dohm_2019}. Hen Harriers breeding in Ireland tended to move away from operational turbines and few fatalities were recorded over a decade long study \citep{Wilson_2016}. The same was found with Montagu's Harriers Circus pygargus followed with GPS trackers breeding near wind farms which showed significant macro-avoidance of turbines \citep{Schaub_2020}. With its small, range-restricted population, the Black Harrier could be critically impacted by both mortality and displacement.
To investigate the magnitude of these impacts and to affect appropriate conservation measures, we conduct a population viability assessment for the Black Harrier integrating the current population trend with the life history parameters of the species. We use population matrices, which are widely used to predict the evolution of populations under different scenarios \citep{Boyce_1992,May_2019}. Integrated population modelling allows estimation of life history parameters and population numbers from different data sets simultaneously in the same model. This provides important benefits in terms of the number of parameters that can be estimated, as well as in the precision of the estimates \citep{Schaub_2010}. However, one of the main obstacles when conducting a population viability assessment is the availability of detailed demographic data for the species of interest, which could compromise the quality of the assessment \citep{Brook_2000,Coulson_2001}. We overcame the need to have detailed count data to estimate different life history parameters by using presence/absence data collected during the South African Bird Atlas Project (SABAP2, \citealt{Brooks2020}). We fit a Bayesian dynamic occupancy model, although our main objective is not to estimate occupancy probabilities \citep[see][]{Royle_2007}, but to investigate changes in the population underlying occupancy, and how these changes relate to specific life history parameters \citep{Royle2003,Rossman_2016}. We allow the model to simultaneously estimate population size and life history parameters, and use information published on the breeding ecology and satellite-tracked movements of the Black Harrier \citep{Curtis2004,Simmons2005,Garcia_Heras_2016,Garcia_Heras_2017,Garcia_Heras_2019} to define sensible priors for the model parameters. With a model for the population dynamics, we undertake a population viability assessment for the species using Monte Carlo simulations to forecast scenarios under different levels of added mortality produced by wind farms.