The role of habitat quality in fragmented landscapes: a conceptual overview and prospectus for future research

Abstract

There is increasing empirical evidence that the quality of habitat patches (determined by either habitat degradation or natural heterogeneity in the quality of habitat) plays an important role in determining species distribution patterns and in regulating spatial dynamics in fragmented landscapes. However, to date, most of the debate has focused on whether or not to include habitat variables in fragmentation studies, and we still lack general conclusions as well as standard and robust research approaches. In this paper we show how a weak conceptualization of “patch quality” and the inappropriate choice of target surrogate variables (e.g., density is often used as an indicator of patch quality) have mainly produced case-specific results, rather than general conclusions. We then identify weaknesses in the inclusion of habitat quality measurements within fragmentation studies. In particular, we focus on: (1) the lack of appropriate experimental design, outlining how few studies have actually included a gradient of habitat quality in their sample; (2) the lack of fundamental information provided (e.g., lack of standard outputs), which in turn hampers the possibility of carrying out meta-analyses. We finally synthesize available knowledge from empirical studies and highlight the different conceptual frameworks needed for patch occupancy versus patch use studies.

Appendix 1: Quantitative definitions of habitat quality and source–sink habitats

• 1. Definition of the quality of a given habitat by Van Horne (1983):

  $$ Q_{{\text{j}}} = {\frac{{\left\{ {\left[ {\sum_{{x{\text{j}}}} \, n_{{x{\text{j}}}} \left( {l_{{\alpha {\text{j}}}} B_{{x{\text{j}}}} + \,P_{{x{\text{j}}}} } \right)} \right]/\sum_{{n_{{x{\text{j}}}} }} } \right\}\left( {1/a_{{\text{j}}} } \right)}}{{\sum_{{xi}} \left\{ {\left\{ {\left[ {\sum_{{xi}} \, n_{{xi}} \left( {l_{{\alpha I}} B_{{xi}} \, + \,P_{{xi}} } \right)} \right]/\sum_{{n_{{xi}} }} } \right\}\left( {1/a_{\sc{i}} } \right)} \right\}}}} $$

  (1)

Q_j is the relative quality of habitat for a given species, B _x is the fecundity of a x-year old, and l _α is the probability that the offspring will survive to α, α being the first age of breeding. P is the probability of surviving from age x to age x + 1, n is the number of individuals in each of the i habitats being compared, and a is the area that includes all sampled individuals in the ith habitat.

• 2. Definition of source and sink habitats by Pulliam (1988)

$$ \lambda = P_{\text{A}} + P_{\text{j}} \beta ; $$

(2)

λ is the population growth rate of a given population in a given habitat; P_A and P_j are the habitat specific survival rates of adults (A) and juveniles (J); β is the habitat specific per capita reproductive rate. In a source habitat λ > 1, while in a sink λ < 1. This definition assumes that populations are at equilibrium and that survival is measured from pre-breeding to pre-dispersal (not annually).

• 3. Definition of the contribution of a local subpopulation to the whole spatially structured population by Runge et al. (2006):

$$ C^{r} = \Upphi_{\text{A}}^{rr} + \sum_{r \ne s} \Upphi_{\text{A}}^{rs} + \beta^{r} \left( {\Upphi_{\text{j}}^{rr} + \sum \Upphi_{i}^{rs} } \right) $$

(3)

C^r is the contribution of a member of the focal subpopulation “r” to the spatially structured population; Φ ^rr_A is the apparent survival of adults that remain in the subpopulation r; Φ ^rs_A is the apparent survival of adults that emigrate to other subpopulations; β^r is the number of juveniles per adult; Φ ^rr_j is the apparent survival of juveniles that remain the subpopulation r, and Φ ^rs_j is the apparent survival of juveniles that emigrate to other subpopulations. If C^r > 1, the focal subpopulation contributes more individuals than it loses via mortality and is a source. If C^r < 1, the focal subpopulation loses more animals to mortality than it contributes and is a sink.