Table 2.
Model output variables and descriptions used in the sensitivity analysis
| Variable | Description |
|---|---|
| Per reef patch | |
| I. Local settlement | |
| LRa | Local retention |
| SR | Self-recruitment (with eight-patch metapopulation) |
| H’ | Shannon index of diversity of settlers (sensitive to weak connections) |
| II. Downstream connectivity | |
| mdGa | Median geographic distance of downstream connections |
| mxG | Maximum distance of downstream connections |
| S | Total proportion of larvae that settle downstream |
| dC | Out-degree, total number of downstream connections |
| Cw a | Weighted degree centrality as dC(1-α) x Sα; dp is d as proportion of total possible connections; α = 0.5 |
| III. Metapopulation consequences | |
| λM a | Metapopulation growth rate with variable population sizes, fecundity, & survival [6] |
| λmax | Metapopulation capacity [5] |
Selected parameters (marked with a) are presented in the Figures, with the remaining in the Additional file 1. The intrinsic and extrinsic drivers of larval dispersal Fig. 1, (Table 1) can influence population connectivity at three different scales. At a local scale, the magnitude of local settlement will depend on: (1) what proportion of locally spawned larvae are retained and settle to their natal reef (local retention); (2) what proportion of settling larva were spawned locally (self-recruitment); and (3) whether dispersing larvae come from a diversity of sources (Shannon H’). At a regional scale, how strongly connected populations are by larval dispersal will depend on: the distribution of dispersal distances (mdG and mxG), what proportion of spawned larvae survive to settle to another reef (S), how many downstream reefs receive these larvae (dC) and whether the strengths of these connections are even or skewed (Cw). At a metapopulation scale, connectivity patterns have important consequences for rates of replenishment (i.e., growth) across all patches (λM) and the ability of a species to persist in the landscape/seascape (λmax)