As renewable energy becomes more prevalent in coastal environments, research by Julie Miller and colleagues provides important insights into the effects of anthropogenic influences on bird populations; both the risks and how these can be mitigated. Associate Editor Des Thompson and Scottish Natural Heritage ornithologist Andy Douse discuss issue 56:9‘s Editor’s Choice article.
Of the globe’s birds, seabird populations are arguably among the most sensitive to environmental change. Climate change is a key driver of oceanographic change, with effects both direct and indirect through marine food webs. In concert with this, there are direct influences of competition from conspecifics and weather on nest site availability, breeding success and mortality. Human-related factors add to the complexity, notably through influences from by-catch, oiling, other pollutants, harvesting, and the spread of non-native invasive species. Supplement this with the potential for mortality from offshore renewable energy development, and you appreciate the sheer sensitivity of seabird populations. In his award-winning book Bird Populations (Collins New Naturalist Library, 2013), Ian Newton commented in relation to seabird numbers, ‘The recent declines have been better studied than the earlier increases, and have involved a huge research effort.’ , and that is the problem in looking ahead at how seabird populations may change.
The UK coastline, characterised by its myriad islands and spectacular sea cliffs, hosts globally significant concentrations of seabirds, and for some species, such as the Northern gannet, the largest single colonies (Bass Rock). However, much of this land area and the wider seas are attractive not only for seabirds, but also for renewable energy development and deployment, particularly harnessing offshore wind. The UK is a world leader in offshore energy installations and also in effective adherence to legislation for wildlife protection and standards for development installation and operation. Whilst renewable energy technologies offer to offset some of the effects of human-induced climate change, the potential impacts of structures to wild bird and other animal populations requires careful consideration.
Assessing the risks to seabird populations of such renewable energy developments is challenging, due in part to seabird life histories and the difficulties in observing direct impacts offshore. This includes not being able to observe or quantify seabird demographic processes and collision risk mortality, and not having a firm handle on how uncertainty confounds confidence we might have in environmental impact assessments. This gives rise to significant risks for both industry and conservation. In particular, uncertainty in our estimates may both unnecessarily limit renewable energy development expansion, to the detriment of reducing carbon-based power, and misidentify risks to seabirds, threatening population persistence.
Julie Miller and colleagues’ paper is therefore both highly important and welcome, not least as it devises for renewable energy developments a new approach in developing Population Viability Analyses (PVAs) to help us understand the vulnerability of seabird populations to anthropogenic mortality. Significantly, they have devised an approach which allows assessment of population response to mortality which is sensitive to empirically derived ranges of density-dependence and environmental stochasticity in closed and connected (‘re-seeded’) populations.
Think of a nationally important seabird population, such as a massive sea cliff colony, which fluctuates over time, and is subjected to a development pressure. What are the predominant processes regulating this population? Is it competition from conspecifics? Is it a range of environmental factors? How does the population respond to these regulators in the absence of additionally mortality? How does it respond to development-induced mortality? Common practice in impact assessment involves the use of precautionary models which do not assume density-dependence, and assume a closed population dynamic with no net immigration. However, if there are in fact regulatory and/or connectivity processes at play, and these feed through directly to drive population change, these could have a significant bearing on responses to development mortality.
This is where this paper is so special. Using Bayesian state-space models fitted to population time series data for three sympatric seabird populations, selected for varied life histories, Miller and colleagues were able to show, in subsequent PVA, how population dynamics are driven by environmental stochasticity and density-dependence in both closed and re-seeded populations.
First, by using the Bayesian approach in an estimation of regulation they captured a range of estimates of density-dependence and environmental stochasticity. Running these in a subsequent PVA offered a broad scenario response to all possible combinations of density and environment. From this, a baseline understanding of the effect of density-dependent regulation and environmental stochasticity shaping the population was revealed. By adding human-related mortality scenarios and a ‘rescue-effect’ to simulate the potential for connectivity, they have shown how populations may respond to human induced mortality, and response to model variants incorporating a theoretical, conservative connectivity.
In their assessments, they also apply a previously popular assessment tool, ‘Potential Biological Removal’ (PBR; Wade 1998; Dillingham and Fletcher, 2008; Dillingham and Fletcher 2011). This is used to assess impacts of additional mortality on marine mammal and seabird populations, especially where these are data-deficient. For two of the three species studied, black-legged kittiwake and common guillemot, they found environmental parameters regulate the baseline population. Moreover, there were real risks of decline for all three species under even modest risks of additional mortality, agreeing with recent evidence that PBR is an inappropriate tool for seabird population assessments. However, for demographic data-deficient taxa, PBR remains useful.
What does this mean? Effectively, we need more empirical studies of seabird populations to understand how birth and death rates, and immigration and emigration, vary in relation to underlying regulatory processes. In the presence of knowledge gaps in seabird population dynamics, advanced computational methods exist to assist us in deriving estimates for demographic processes mathematically, and then exploring subsequent population dynamics and risk. By adding realism with empirical evidence and advanced modelling methodologies we can be more confident in predicting how environmental fluctuations, density-dependence and off-shore renewable energy developments can influence seabird populations.
As Newton concludes in his almost 600 page treatise on bird populations:
Not all factors that influence bird numbers act in a density-dependent manner. Some act independently of density, while others act in an inversely density-dependent manner, affecting a decreasing proportion of individuals as their numbers grow, and thereby accentuating fluctuations in numbers. Yet others act in a delayed density-dependent manner, acting to promote cyclic fluctuations in numbers… Bird numbers fluctuate because of all the factors that act on them, however they relate to density, but it is those that act in a direct density-dependent manner that serve to keep numbers within limits, acting to resist boundless increase or extinction. (Newton, Bird Populations, 2013).
Mindful of this, we must endeavour to develop models which build on the highly important work of Miller and her colleagues, and heed the advice of a former President of the British Ecological Society. If we do not, we will superficially assume that marine developments will not adversely affect seabird populations, when in fact they could be catastrophic tippers of change.
Read the full Editor’s Choice article, The sensitivity of seabird populations to density-dependence, environmental stochasticity and anthropogenic mortality, in issue 56:9 of Journal of Applied Ecology.