Table 2. Results for the regression model of coral mortality as a function of different explanatory variables.
| Variables | Coefficients | Average Marginal Effect |
|---|---|---|
| Maximum DHW | 0.304*** | 0.492*** |
| (0.050) | (0.071) | |
| Maximum PAR Anomaly | -0.0073 | 0.0808*** |
| (0.015) | (0.018) | |
| Population | 0.0373 | 0.0679 |
| (0.031) | (0.057) | |
| Wind | 0.00114 | 0.00207 |
| (0.040) | (0.073) | |
| Depth | -0.143*** | -0.0969*** |
| (0.056) | (0.023) | |
| Depth×Maximum PAR Anomaly | 0.00517*** | |
| (0.0012) | ||
| Depth×Maximum DHW | -0.00328 | |
| (0.0040) | ||
| Bermuda | -10.9*** | -93.1*** |
| (0.46) | (5.8) | |
| Bahamian | 1.56** | -6.2* |
| (0.77) | (3.5) | |
| Eastern Caribbean | 0.474 | -4.7*** |
| (0.35) | (1.6) | |
| Greater Antilles | 0.896** | 4.9** |
| (0.42) | (2.3) | |
| Southern Caribbean | -2.08** | -7.4*** |
| (1.03) | (2.6) | |
| Southwestern Caribbean | -0.380 | -19.5*** |
| (0.35) | (2.2) | |
| Floridian | -1.23*** | -12.3*** |
| (0.47) | (2.2) | |
| Constant | -5.74*** | |
| (0.89) | ||
|
Log Likelihood Cross-validated MSE |
-64.34 30.64 |
|
| Cross-validated RMSE | 5.53 | |
| Cross-validated R2 | 0.325 | |
| Number of Observations | 1,045 |
Estimated coefficients and average marginal effects using a fractional logit model, where coral mortality is explained as a function of environmental stressors. Robust standard errors are shown in parentheses (*** p<0.01, ** p<0.05, * p<0.1).