Describing rarity: Pattern
There is no universally agreed-upon definition of rarity or what exactly constitutes a rare species. While the term is most commonly used to refer to species characterised by a restricted distribution, low abundance, or both , some authors have incorporated different metrics into their definitions of rarity, including habitat breadth, occupancy, or persistence . As a starting point, we define rarity as being uncommon. While this “uncommonness” can be defined at various scales—e.g., within a particular biotic community, the boundaries of a jurisdiction, or globally—we focus on species that are globally uncommon. Thus, species that are rare only within a certain context, such as a particular jurisdiction or habitat type, are not considered here.
Despite the popularity and evident utility of the Rabinowitz framework, it does have weaknesses. First, the three rarity dimensions (range size, local abundance, and habitat specialisation) represent a combination of pattern and process. Whereas range size and local abundance both describe patterns in species’ occurrence, habitat specialisation is arguably a process (environmental filtering) and it can influence both patterns. Creating a process-based framework for rarity requires a clear distinction between patterns of rarity and the processes that drive them. Second, the Rabinowitz framework captures patterns of occurrence either at very broad scales (i.e., range size) or very fine scales (i.e., local abundance), but lacks a measure of regional-scale rarity. As scale can substantially affect patterns of rarity, along with related issues such as population decline , this omission may create an important gap in assessing rarity.
We propose addressing both these issues by removing habitat specialisation from the Rabinowitz framework, and replacing it with a new dimension, namely occupancy. In some contexts, occupancy is a binary variable that indicates whether a species occurs at a particular site (as in occupancy surveys or models, e.g., . It has also been defined as the proportion of cells in a gridded landscape currently occupied by the species . Here, we define occupancy as the proportion of habitat patches (or sites, grid cells, etc.) within a species’ range that is occupied by the species. While it is somewhat unconventional to define occupancy specifically in relation to a species’ range, by doing so we can capture the density at which the species occurs within the range, which has implications for the persistence of rare species and the type of conservation measures they may require. We discuss these implications inConservation implications (below).
The distinction between occupancy and range size warrants some further precision, as the two are often used interchangeably in the literature, including some studies that have applied the Rabinowitz framework (e.g., . However, as shown in Figure 1, we defined occupancy in a way that is distinct from range size: whereas range size refers to the global limits of the area within which a species occurs (i.e., extent of occurrence), occupancy denotes the density of occurrences within those limits. While occupancy may provide a practical proxy for range size that is suitable in the context of certain research questions, the two are not conceptually equivalent. Furthermore, a species with a small range but high occupancy may require different conservation and management measures compared to a species with a large range but low occupancy (Figure 1); this will be discussed in more detail in Conservation implications . Given that studies of rarity (or rare species) often have conservation objectives, the distinction between range size and occupancy is important to maintain.
There are four major advantages to replacing habitat specificity with occupancy: first, like range size and habitat specificity, occupancy describes a pattern of occurrence. Second, our proposal yields a system that acknowledges the importance of scale in rarity, as the proposed dimensions of range size, occupancy, and local abundance correspond to global, regional, and local scales, respectively (Figure 2), thus filling the regional scale gap in the Rabinowitz framework. Third, as we will argue in more detail in Conservation implications , using occupancy rather than habitat specialisation will allow the framework to more accurately identify species that are truly uncommon, which has important implications when using the classification system for conservation purposes. Finally, because habitat specialisation is among the mechanisms that can influence occupancy , habitat specificity will remain embedded in the framework, but as a causal driver rather than a rarity dimension, as explained in the Explaining rarity: Processsection.