The experiment was carried out at the experimental farm of Wageningen University & Research from 15 June to 2 October in 2021 (51°59'47" N, 5°39'36" E, Wageningen, the Netherlands). The objective of the experiment was to examine how natural enemy assemblages with different evenness but the same species richness and abundance affect the efficiency of pest predation. To this end, we used the aphid pest species S. avenae on wheat (T. aestivum) exposed to four species of arthropod predatory natural enemies (C. septempunctata, M. pygmaeus, P. cupreus and T. ruricola) as our model system.

We applied four evenness treatments and a control treatment in which no arthropod predators were present . In all four evenness treatments, we introduced four species of arthropod predators amounting to 16 individuals in total. The evenness treatments varied from highly uneven to completely even and consisted of (E1) one, two, three and ten individuals of the four species, respectively; (E2) one, three, five and seven individuals of the four species, respectively; (E3) three individuals of two species each and five individuals of the other two species each, and (E4) four individuals of each of the four species. We randomly assigned each of the five treatments (four evenness treatments and one control treatment) to four cages. To be able to examine the effects of species identity independently from the evenness treatment, we carried out the experiment for four rounds. In each round the total number of individuals, species and evenness treatments were the same but a different species was the most abundant, the second most abundant, the third most abundant and the least abundant species in all the treatments of that round. The total sample size of our study was therefore 80.

In each cage and each round, we introduced approximately 600 adult apterous S. avenae on the wheat plants of the central four pots (i.e., about 150 aphids per pot). Aphids were inoculated on the top three leaves of the plants using a fine hair brush. We then closed the cage and allowed the aphids to establish. After two days we recorded all observed live aphids on the wheat plants (the first aphid count). In each of the evenness treatments, the arthropod predators were introduced immediately after the first aphid count by randomly placing them in the outer twelve pots in each cage to avoid initial intraguild predation. After the aphids had been recorded and the arthropod predators had been introduced in all cages, they were left undisturbed until the second aphid count on the sixth day. We only counted the aphids that were on plants as the dropped live aphids can quickly return to plants canopy if they were not predated by ground-dwelling arthropod predators. Thus abundance of the aphids on plants can also reflect predation by ground-dwelling arthropod predators. Aphid growth rate was calculated for each cage as the number of aphids of the second count divided by the number of aphids of the first count

Explanation of table headings
"Cage.no" refers to the cages, different cages were assigned with different treatments. 
"Treatment" refers to different evenness or control treatments (detailed treatment can be found in Fig. S1). 
"Round" refers to four different round. "Count" refers to the first or second count. "Count.date" refers to the date which the aphids count was carried out. 
"Pot 1" to "Pot 16" refer to the individuals of aphids of each pot. 
"Carabids" to "Mirid bugs" refer to the the indivuduals of arthropod predators, for the date in the "1st count", they indicate the individuals of arthropod predators introduced as designed; for the date in the "2nd count", they indicate the recollected individuals of arthropod predators. 
"Non-target predator" indicates if the non-target predators were collected after each round of experiment. "1" in this column only indicates that we collected non-target predator, but not the exact number of the non-target predator we collected.

Cage no. 8, 12, 17, 20 in the 1st round; 3, 8 in the 2nd round; 1, 3, 14, 18 in the 3rd round and 8 in the 4th round were excluded from analysis as we collected non-target predator or recollected more arthropod predators than we had introduced. 