Low breeding performance of the Steller’s sea eagle (Haliaeetus pelagicus) causes the populations to decline

Abstract

Adult survival has long been believed to be the most important demographic parameter in long-lived species such as large raptors, with productivity being of secondary importance. Our studies of the Steller’s sea eagle (Haliaeetus pelagicus) populations on the north-east of Sakhalin Island and in the Amur region (Russian Far East) in 2004–2017 have revealed highly unstable productivity that averages 0.53 and 0.61 fledglings per occupied territory per year, respectively, which is fairly low in comparison to previous decades (about 1 fledgling per occupied territory per year). The productivity varied between years within a 2.7-fold span, being relatively low most of the time. On Sakhalin Island, the most serious hazard influencing eagle productivity is nestling predation by brown bears. Among other negative factors are adverse weather and food conditions, habitat alteration, disturbance at nest sites, and (on the mainland) the hydrological regime of the Amur River.To assess the sustainability of both populations, we have developed two matrix population models of the Lefkovitch type. The models predict a decline of both island and mainland populations, these being halved in 115 and 72 years respectively. Apart from the main scenario (1), we considered two additional scenarios to stabilize the populations: (2) cessation of bear predation and (3) enabling population regulation by involvement of floaters (non-territorial adults) into the reproduction. On Sakhalin Island, increasing the fecundity under Scenarios 2 and 3 could theoretically stabilize the population. On the mainland, the stabilization cannot be realized under either scenario. Stochastic modelling supports the main conclusion of population decline, but generally its predictions are more optimistic, some scenarios leading to different conclusions. Therefore, though admitting the secondary importance of fecundity in large raptors as compared to adult survival, we suggest that the fecundity can vary in broader limits and, if at a poor level, can alone lead to population decline. On the other hand, this parameter is more amenable to management, so that efforts should be focused on raising it to higher levels. Stochasticity can have major effects on the population dynamics and should be accounted for in models.