4.3 Inter-colony variation in breeding probability
The recent population trend for African penguins at Stony Point has generally been positive while that at Robben Island has been negative, but the drivers of this difference have not been fully explored (Sherleyet al ., 2020). We indentified clear differences in survival and breeding propensity between the two colonies, drivers that likely underpin these divergent population trends. Overall, individuals at Stony Point had higher adult survival and a higher probability of breeding than those at Robben Island, with breeders at Stony Point also more likely to remain in a breeding state and nonbreeding individuals (including prebreeders) more likely to transition into a breeding state (Figure 5). Trends over time indicate this difference is growing, with an increasing presence of reproductive skipping behaviour over time at Robben Island (Figure 4a). Notably the one occasion at Robben Island where breeding individuals were more likely to skip reproduction the following year than to remain breeders (Figure 4a) coincided with the lower encounter rates in 2020 (Figure 2), likely due to reduced monitoring during the Covid-19 pandemic; this may have led to an overestimation of reproductive skipping rates in this year.
The presence of variation in breeding propensity over time suggests reproductive skipping here is not purely driven by the individual constraints of lower quality individuals, but implicates external drivers that differ between colonies. In other seabirds, food availability is a prominent driver of reproductive skipping (Gauthier-Clerc et al. 2001). Despite this, and previous work linking food availability to lower reproductive output in African penguins (Sherley et al. 2013, 2018, Campbell et al . 2019), this analysis did not find support for food abundance as a driver of reproductive skipping in African penguins. As previously discussed, this may be explained by colony specific food availability not being fully accounted for in this study, with colonies not responding as expected based on the more general prey abundance estimates (e.g., Sherley et al . 2013). Alternatively, food availability may only impact African penguin breeding after they have made the decision to breed (e.g., impacting offspring survival, but not the likelihood of attempting to breed). However, disentangling this relationship requires future study with colony-specific measures of prey availability, along with additional years of monitoring to improve the estimates of the transition parameters and better capture the full variation in breeding decisions and how they relate to food availability.
Nevertheless, recent research highlighting inter-colony variation in African penguin metrics – such as the higher adult body mass at Stony Point than nearby Western Cape colonies (Espinaze et al. 2020) – may help us understand these differences in breeding propensity. In king penguins, individuals will abandon breeding attempts when body mass drops below a certain threshold (Gauthier-Clerc et al. 2001), highlighting this as a potential driver of higher breeding propensity at Stony Point. Penguin chicks at Stony Point have also been found to have lower levels of corticosterone (a stress hormone) than chicks at Robben Island (Scheun et al. 2021). If this is also true of the adults, this could be a key proximate-level driver of reproductive skipping, with elevated stress due to factors like low food availability causing disruption of the endocrine control of breeding (Kitaysky et al.2007; Bókony et al. 2009; Crossin et al. 2013). These inter-colony differences in stress have been linked to an irregular anthropogenic presence at Robben Island (e.g., researcher presence, some limited tourism, and anthropogenic noise) compared to Stony Point (e.g., regular, high volumes of tourists confined to boardwalks; Scheunet al. 2021). Whether variations in localised food availability also contribute to these differences is as yet unclear; further monitoring of stress hormones in relation to external factors (e.g., prey availability) and reproductive skipping could confirm this to better inform our understanding of inter-colony differences and improve population management.
In conclusion, we reveal key insights into African penguin demography, providing the first evidence of reproductive skipping in this species. Penguins are breeding less than theoretically possible, with reduced breeding propensity driven predominantly by individual-specific constraints. We also present clear inter-colony differences: individuals at Robben Island are responding more negatively to declining food availability and are characterised by lower survival and a lower breeding propensity than those at Stony Point. These differences highlight a need for a more detailed understanding of the localised drivers of these differences in population dynamics and imply a greater need for conservation action at Robben Island, beginning with actions to improve access to prey.