Introduction

A core aim of conservation management is optimising habitat quality for focal species (Johnson, 2005, McComb, 2016). For management to be truly optimised, a measurable understanding of what constitutes habitat quality is required (Marzluff et al., 2000). The ultimate measure of habitat quality for an individual is the individual’s relative contribution to the growth rate of the population when inhabiting a given habitat (Johnson, 2007). There are two components of this measure: survival and reproduction. By defining habitat quality in terms of population growth rate, habitat quality can be assessed on a continuous temporal scale. For example, habitat quality can be measured instantaneously or as a life-time measure of habitat quality akin to the individual’s fitness. There are many components that combine to influence survival and reproductive output including food availability, predation risk, habitat structure and configuration, and the presence of disturbances (e.g., human foot traffic) (Johnson, 2007).
Quantifying demographic rates (survival and reproductive output) is a challenging task (Stephens et al., 2015), as it requires sustained monitoring of individuals of known-identity. Studies that do achieve this are often conducted either on sessile organisms (e.g., Ma et al., 2014, Wang et al., 2012, Zhao et al., 2006) or large-bodied organisms that are restricted to a small geographic area (e.g., islands: (Kruuk et al., 1999, Richard et al., 2014); natal colony: (Baker and Thompson, 2007, Le Boeuf et al., 2019)). Demographic rates are also financially costly to measure (Knutson et al., 2006, Pidgeon et al., 2006), and the long time-frames for data collection can mean that research extends beyond typical funding cycles and research project lifetimes, particularly for research on long-lived species (Le Boeuf et al., 2019). Despite these challenges, there have been studies that successfully monitor survival (Valdez-Juarez et al., 2019) and reproductive performance (Pérot and Villard, 2009, Pidgeon et al., 2006, Zanette, 2001) of birds in relation to habitat quality. Outputs from these studies are often very applied with actionable recommendations for conservation decision-makers.
Waterbirds are a particularly challenging group to obtain habitat quality estimates for because multiple factors can confound the relationship between site habitat conditions and resultant demographic rates. Many waterbirds are highly dispersive and track ephemeral habitat conditions at local, regional, or even continental scales (Cumming et al., 2012, Pedler et al., 2014, Roshier et al., 2006), creating the potential for mismatches between the scale of monitoring and the scale at which demographic processes are governed. Habitat quality at a particular wetland may be high relative to other points in time, yet waterbirds do not capitalise on these favourable conditions because there are other areas of high quality habitat in the landscape (behavioural choice impacts) (Cumming et al., 2012). Consequently, habitat quality assessments based on abundance, density, or occupancy for the particular site may be decoupled from theoretical predictions if data from the broader landscape are unavailable. The distribution of many waterbird species is also influenced by social attraction (Gawlik and Crozier, 2007). As a result, areas of high quality habitat may go unused because waterbirds newly arriving in an area are drawn to sites with existing waterbird presence (Gawlik and Crozier, 2007).
Many waterbirds are also migratory. Consequently, demographic parameters in one part of the range may be decoupled from the habitat conditions experienced at that time due to carry-over effects from previous seasons (Aharon-Rotman et al., 2016a, Sedinger and Alisauskas, 2014, Swift et al., 2020). For example, survival during the breeding period and breeding success may be higher in individuals that depart their non-breeding grounds in better condition (Swift et al., 2020). Furthermore, breeding performance in one part of the range may influence parameters including abundance and population age structure on the non-breeding grounds, irrespective of the local conditions on the non-breeding grounds (Rogers and Gosbell, 2006). In addition to carryover effects, survival data may be particularly sensitive to pinch points of low-quality habitat along the migratory flyway (Piersma et al., 2016, Studds et al., 2017).
Due to the difficulties of obtaining waterbird demographic data in a given area, an array of methods have been used as proxies to measure habitat quality (Ma et al., 2010). The use of proxies also helps to overcome budget limitations of management agencies by allowing snapshot estimates of habitat quality to be made without the need for extended periods of data collection in space and time (Osborn et al., 2017). However, the many different options available for measuring habitat quality can be bewildering for research scientists and conservation practitioners (Pidgeon et al., 2006). There is little consensus on which method, or combination of methods, produces the most meaningful estimate of waterbird habitat quality, and in some cases, it is unclear as to whether particular proxies meaningfully reflect underlying habitat quality from the perspective of direct impact on population processes (Johnson, 2005, Johnson, 2007, Van Horne, 1983). For example, density of individuals may not reflect underlying habitat quality if the population does not follow the ideal free distribution (Van Horne, 1983), and time spent foraging may not reflect underlying habitat quality if individuals are constrained by prey handling time or digestive bottlenecks (Van Gils et al., 2005). Furthermore, the spatial scale at which proxies are measured may have implications for their relevance to managers (Pidgeon et al., 2006, Stephens et al., 2015).
In this review, we seek to catalogue the methods that have been used to quantify waterbird habitat quality and provide a synthesis of the conditions under which each may provide meaningful measures of habitat quality in future waterbird studies. Outputs from this review are intended to guide environmental managers on the types of data they should be collecting when attempting to quantify waterbird habitat quality. This will ensure that decisions on how to manage habitat to optimise habitat quality are based on meaningful information.