Parasite Intensity
An REM of intensity effect sizes did not detect a statistically significant effect of increased predation on parasite intensity in prey (z = 0.829, p = 0.407; Figure 1b), with a large amount of true heterogeneity between studies (I2 = 75.93%), and no evidence of publication bias (𝞽 = -0.106, p =0.231). The single most important moderator was interaction type (Importance = 0.775; Figure 2b, Table 2), and this was the only variable identified as a significant moderator in subsequent univariate analyses (QM1 = 5.848, p = 0.016), despite significant residual heterogeneity (QE59 =182.050, p < 0.001). Non-consumptive interactions had more positive effects than all interactions (Figure 3).