Discussion
Our population viability assessment indicates that Black Harriers are remarkably susceptible to mortality produced by wind farms. Our baseline population status derived from a decade of bird atlas data revealed an average decline of 1.5% per annum, a trend that alone can result in extinction in just over 370 years. The analysis of Black Harrier population under additional mortality produced by wind farms reveals that time to extinction is significantly reduced with every adult individual removed from the population per year. As such, an additional mortality of only three adult birds killed per year could result in the entire population collapsing in under 100 years (Figure \ref{357766}). If additional mortality increases to five adults killed then we predict the global population will collapse before the turn of the 21st century.
We recognize that the estimates of the initial population of Black Harriers in 2008 is only an approximation based on expert opinion and must be taken with caution \cite{1992}; however, the change is population dynamics should be independent of the initial value, as long as there are no density-dependent parameters. There is no evidence of this type of mechanism acting on Black Harrier populations at present, but population density does play a role among other raptor species \citep{SIMMONS_1993,2002,Elliott_2011}, influencing both productivity and age at first breeding, which may have a significant impact on rare or recovering populations \citep{Morandini_2019}.
We have considered a scenario where rainfall remains at the average levels observed during the study period 2008-2019. Our model suggests an important effect of rainfall on the fecundity of Black Harriers, and therefore, changes in rainfall regimes could impact our predictions. Unfortunately, climate analyses already show a delayed start to the winter rains and a reduction in the length of the winter rainfall season (since 1980) and thus overall rainfall \cite{abrahams} , which would translate in a steeper reduction in population numbers than anticipated by our model. Raw SABAP2 data shows an increase in reporting rates around 2014 followed by a steady decline (figure \ref{747849}). We believe this observation may be an artifact created by a national emphasis to conduct a global population count using bird atlas data. However, this is uncertain and this outlying observation may also correspond to changes in the population produced by variables uncontrolled in the present study. If this was the case, the sharp decline in the Black Harrier population that we observe after 2014 (as shown by a steep decline in reporting rates in figure \ref{747849}) could correct its trajectory under the right circumstances. We have an insufficient duration of appropriate data to clarify any cyclic behaviour of the population or density-dependent life history parameters. Thus, until these questions can be clarified, we suggest taking a precautionary approach and consider that Black Harrier populations could be severely affected by additional mortality, as the current analysis shows.
The latest research indicates that eight Black Harriers have been killed by wind farms in South Africa in the last four years \citep{Perold_2020}. Considering that Black Harriers are rare and have core breeding areas in the south western Cape, where wind farms are proliferating, the mortality figures presented here are not just theoretical possibilities used to populate viability assessments, but are grounded in present rates of harrier mortality and projections of wind farm development in South Africa \citep{Perold_2020}. These relatively high levels of mortality contrast with the few harriers fatalities reported at North American or European wind farms \citep{b2008,Drewitt_2008,Schaub_2020}. While there is evidence showing that some harriers species tend to move away from wind farms \citep{Wilson_2016,Pearce_Higgins_2009,Garvin_2010,Schaub_2020}, breeding Black Harriers do not appear to exhibit this behaviour (RE Simmons and M Martins unpubl data), which may partly explain the difference in observed mortality rates. A study of Black Harriers breeding within an Eastern Cape wind farm revealed that harriers are more susceptible to mortality during breeding, as the proportion of time spent flying within the blade swept area rose from 0% to 45%, and so harrier deaths increased (RE Simmons and M Martins unpubl. data). In addition, the species is known to perform very wide-scale movements and hold large territories, which may expose them to numerous wind farm environments (breeding birds move on average 16.4 km from their nests, with estimates of territory size of 92.7 ± 66.6 km2 and 147.8 ± 205.4 km2 for breeders and non-breeders, respectively \citealt{Garcia_Heras_2019}).
Another possibility explaining the apparent difference in susceptibility to turbine-related deaths between Southern Africa's Black Harrier and harriers in Europe is that deaths are under-reported in the primary literature but are subsumed in internal reports and unpublished data sets . For example, in the comprehensive \citet{brandeburg2020} unpublished report, wind farm fatalities of 3 species of harriers ranked in the top 25% of 31 raptor species killed across 11 countries in Europe. The reasons they don't appear in the primary literature is unknown, but it is clear from these data that harrier deaths are far from uncommon and conservation actions are required to reduce the death rate on both continents.