A slow‐growth high‐mortality meta‐analysis for insects: A comment on Chen and Chen

Insect Science recently published a meta-analysis that tried to test support for the slow-growth high-mortality (hereafter SG-HM) hypothesis (Chen & Chen, 2018). Meta-analysis is a powerful statistical tool used to quantitatively compare and summarize multiple studies (Arnqvist & Wooster, 1995). However, results from a meta-analysis must be carefully interpreted when only limited sample sizes are available. Many conclusions made by Chen and Chen (2018) are based on low sample sizes and/or have small fail-safe numbers, which is problematic when making conclusions about whether the SG-HM hypothesis is supported or rejected. The SG-HM hypothesis is based on the idea that development time of an herbivorous insect impacts its survival and thus also its fitness. The longer an insect feeds on a plant, the longer it is exposed to natural enemies (i.e., predators, parasitoids, and pathogens) or adverse climatic events, which increases its risk of mortality. Feeny (1976) first proposed that herbivores take more time to develop when they feed on low-quality plants (e.g., high chemical/physical defenses or low nutritional quality) than when they feed on high-quality plants, which presumably reduces herbivore fitness. Later, Price et al. (1980) predicted that the longer an herbivore in a vulnerable immature stage takes to develop into a reproductive adult, the greater the risk of attack by natural enemies; if an immature insect dies before reaching reproductive maturity, then fitness is reduced to zero. The SG-HM hypothesis has now been tested many times with studies both supporting or rejecting it for various reasons (e.g., Damman, 1987; Haggstrom & Larsson, 1995; Benrey & Denno, 1997; Fordyce & Shapiro, 2003; Lill & Marquis, 2001; Murphy, 2004; Medina et al., 2005; Cornlissen & Stiling, 2006),