Table 1 Summarized effects of predictors on ecosystem services and food web metrics.

Method	Predictor	Response variable	Coeff^b	P
Linear regression	EDS	ES	0.10	0.158
GLM	Gq	ES	−0.13	0.020
GLM	Vq	EDS	0.12	0.007
GLM	NCH	ES	0.25	0.066
GLM	Maize	EDS	0.44	0.061
	NCH	EDS	0.88	0.077
GLM	Intercept^a	Gq	1.26	0.561
GLM	SC	Vq	−3.21	0.034
GLM	Intercept	Parasitoid richness	1.84	<0.001
GLM	Cotton	Parasitoid diversity	−0.80	0.113
	SC	Parasitoid diversity	0.68	0.062
GLM	Intercept	Parasitoid richness	4.12	<0.001
GLM	Intercept	Hyperparasitoid diversity	0.11	<0.001
LMM	Parasitoid richness	ES	−0.06	0.001
	NCH	ES	0.20	0.061
LMM	Vq	EDS	0.11	0.010
	NCH	EDS	0.57	0.076
Path analysis	Parasitoid richness	ES	−0.524	0.043
	Parasitoid diversity	ES	0.072	0.859
	Gq	ES	−0.163	0.705
	Gq	Parasitoid richness	0.693	0.000
	Parasitoid diversity	Gq	0.899	0.000
Path analysis	Vq	EDS	0.514	0.009
	Hyperparasitoid diversity	Vq	0.824	0.000
	SC	Vq	−0.449	0.000
	Hyperparasitoid richness	Hyperparasitoid diversity	0.778	0.000

Analyses include linear regression, generalized linear model (GLM), linear mixed effect model (LMM), and path analysis with structural equation models (SEM). Effects are assessed of multiple predictors on either ecosystem services (ES; parasitism) or disservices (EDS; hyperparasitism). Detailed descriptions of all variables are provided in Supplementary Table 1.
Gq and Vq are the generality and vulnerability of primary-hyperparasitoid food web; NCH is non-crop habitat cover, SC is secondary crop cover.
^aBest model without landscape variables.
^bRegression effect coefficient.

Quick links

Search