Spider mites avoid caterpillar traces to prevent intraguild predation

All the materials followed relevant institutional and national guidelines and legislation.

Mites

We used a T. kanzawai population collected from trifoliate orange trees (Poncirus trifoliata [L.] Raf.) in 2018 in Kyoto, Japan, and a T. urticae population collected from chrysanthemum plants (Chrysanthemum morifolium Ramat.) in 1998 in Nara, Japan. These populations were reared on adaxial surfaces of kidney bean (Phaseolus vulgaris L.) primary leaves, which were pressed onto water-saturated cotton in Petri dishes (90 mm diameter, 14 mm depth). The water-saturated cotton served as a barrier to prevent mites from escaping. The dishes were maintained at 25 °C, 50% relative humidity, and a 16L:8D photoperiod. All experiments were conducted under these conditions. We only used mated adult females (i.e., the dispersal stage) of T. kanzawai or T. urticae mites.

Caterpillars

We used caterpillars of four lepidopteran species: Bombyx mori L., P. Xuthus, Spodoptera litura Fabricius and T. oldenlandiae. We collected eggs and larvae of T. oldenlandiae from C. japonica in 2021 in Kyoto, Japan, and reared them on C. japonica leaves until pupation. Theretra oldenlandiae shares Vitaceae host plants with T. kanzawai and T. urticae^8,15. We collected eggs and larvae of P. xuthus from Ptelea trifoliata in 2021 in Kyoto, Japan, and reared them on Citrus unshiu Markov. leaves until pupation. Papilio. xuthus and T. kanzawai share P. trifoliata as a host plant in Kyoto (Kinto, personal observation).

We obtained commercial populations of the B. mori Kinshu × Showa strain (Ueda-sanshu Co., Ltd, Nagano, Japan) or the w1-pnd strain. We reared B. mori larvae on an artificial diet produced at the Kyoto Institute of Technology. Although T. kanzawai use Morus alba, a food plant for the B. mori strain, the mite and the strain never encounter one another in the wild, because the B. mori strain has been domesticated for hundreds of years.

We obtained a sub-cultured population of S. litura from the Kyoto Institute of Technology. We reared first to fourth instars of S. litura on an artificial diet (Insecta LFM, Nosan Insect Materials, Kanagawa, Japan), while final instars were fed P. vulgaris leaves. Because S. litura feeds on various wild and cultivated plants^22,23, it may share some host plants with T. kanzawai and T. urticae, both of which also feed on many host plant species^8,9,10.

We reared caterpillars of T. oldenlandiae, P. xuthus, and S. litura in 900 mL transparent plastic cups and caterpillars of B. mori in transparent plastic containers (140 × 220 × 35 mm). All caterpillars were maintained under the same laboratory conditions described above.

Plants

We used several parts of P. vulgaris plants in the following experiments. This species is a preferred food for both mite species^16,17 and S. litura²⁴, but the other three caterpillar species do not feed on it (Kinto, personal observation). We thus used P. vulgaris rather than shared host plants, because some caterpillars and mites (T. urticae and P. xuthus, for example) do not share any host plant.

Avoidance of caterpillar traces on leaf surfaces by spider mites

To examine whether spider mites avoid settling on host plant surfaces bearing caterpillar traces, we conducted dual-choice tests using paired adjacent leaf squares with and without caterpillar traces. We did not use whole plants because, in practice, it was difficult to induce caterpillar traces on whole plants. We used two spider mite species (T. kanzawai and T. urticae) and four caterpillar species (T. oldenlandiae, P. xuthus, B. mori, and S. litura). We cut a 10 × 20 mm leaf piece from a fully expanded primary kidney bean leaf and then cut the piece into two equal squares (10 × 10 mm). To introduce caterpillar traces to one square, we arranged them on a separate piece of paper towel on water-saturated cotton. This procedure was necessary because the caterpillars used were larger than individual leaf squares. Then we placed a fourth or final instar caterpillar on the squares and induced the caterpillar to walk across every leaf square three times (Fig. 1a). We carefully removed all caterpillar-produced silk threads from the squares. Within 30 min, we arranged the square (trace +) to touch against the other square (trace −) on water-saturated cotton in a Petri dish. Subsequently, a 2- to 4-day-old mated adult female of T. kanzawai or T. urticae was introduced onto a pointed piece of Parafilm in contact with both leaf edges using a fine brush (Fig. 1a). We recorded the leaf square onto which the mite had settled at 2 h after its introduction, as preliminary observations confirmed that all females would settle on a particular leaf within that period. Each female mite and pair of leaf squares were used only once. All tests described below were conducted between 13:00 and 17:00 h, when adult female spider mites actively disperse by walking. There were 14 replicates using traces of T. oldenlandiae, 48 of P. xuthus, 20 of B. mori, and 26 of S. litura for T. kanzawai, as well as 18, 32, 16, and 47, respectively, for T. urticae. Data were subjected to two-tailed binomial tests with the common null hypothesis that a spider mite would settle on the two squares with equal probability (i.e., 0.5).

Duration of B. mori trace avoidance by T. kanzawai

To examine whether the effects of caterpillar traces on spider mite avoidance decline over time, we used T. kanzawai mites and B. mori caterpillars. We used B. mori because populations can be easily maintained over many generations. We prepared bean leaf squares with B. mori traces in the same manner descried above and preserved the traced square on water-saturated cotton for 0 h (n = 30), 24 h (n = 29), 48 h (n = 28), or 72 h (n = 28). Then we arranged the square (trace +) to lie in close proximity to the control square (trace −) that had been preserved for the same periods of time. Then we compared the avoidance response of T. kanzawai females in the same manner described above.

Avoidance of B. mori traces on plant stems by T. kanzawai

To examine whether T. kanzawai females avoid walking along plant stems bearing caterpillar traces, we used Y-shaped kidney bean stems (Fig. 1b). We cut symmetric bean plants ca. 15 days after sowing from their base and inserted them perpendicularly into a 5 mL glass bottle filled with water and wet cotton. To induce caterpillar traces on one branch of the stem, we allowed a silkworm to crawl from the branching point to the far end of one branch three times for each stem (n = 20). Then we introduced a T. kanzawai adult female at a release point 35 mm below the branch point (Fig. 1b). We recorded the branch along which the female walked to the far end. Each female mite and each Y-shaped stem were used only once. The numbers of females were compared using binomial tests in the same manner described above.

Avoidance of B. mori trace extracts by T. kanzawai

To extract chemical traces of caterpillar, we introduced 10 third instar B. mori to a glass Petri dish (120 mm diameter, 60 mm depth). After 1 h, we removed all caterpillars and washed the inside bottom of the dish with 1.0 mL acetone. We replicated the procedure twice using different individuals to combine all extracts and to acquire enough extract for the following experiment.

To examine avoidance of B. mori trace extracts by T. kanzawai females, we conducted dual-choice experiments using T-shaped pathways of filter paper (35 × 35 mm; width, 2 mm; Fig. 1c). Using disposable micropipettes (Drummond Scientific Co., PA, USA), 1.75 caterpillar equivalents (i.e., 60 µL) of acetone extract were applied to an alternately selected branch (17.5 mm long) of each pathway (i.e., 0.10 caterpillar equivalent/mm), with control acetone applied to the other branch. We applied each solution dropwise at the junction point to minimize mixing. After evaporating the solvent from those pathways, we perpendicularly suspended them (Fig. 1c) and introduced an adult female mite at 2 days post-maturation onto the bottom of each pathway using a fine brush and recorded the branch along which the female first walked to the far end. Each female mite and each T-shaped filter paper were used only once, with 19 replicates. Each female mite made a choice within 10 min. The avoidance response of T. kanzawai was analysed in the same manner described above.

Indirect effects of B. mori traces on T. kanzawai via plants

To determine whether B. mori traces on plants indirectly affect the performance of T. kanzawai on plants, we introduced 70–80 randomly selected quiescent female deutonymphs of T. kanzawai onto kidney bean leaf disks. Immediately after synchronized adult emergence, we introduced the same number of adult males to allow mating; the detailed procedure is described elsewhere²⁵. After 24 h, we transferred the females singly onto 10 × 10 mm bean leaf squares with or without B. mori traces prepared as described above. Because the number of eggs laid within a certain period is considered the most sensitive performance index of spider mite females^26,27, any plant-mediated indirect interaction, such as defence induction in response to caterpillar traces, should result in lower egg numbers laid by the test females. We counted the eggs laid on the leaf squares 24 h after their introduction. One female that laid no eggs during the 24 h period (n = 1, trace +) was excluded from the analysis. We obtained 33 and 36 replicates for the trail+ and trail– conditions, respectively. We compared the numbers of eggs laid on leaves with and without B. mori traces using a generalized linear model with a Poisson error distribution using the SAS 9.22 software (SAS Institute Inc., Cary, NC, USA).

Ethics

This article does not contain any studies with human participants or animals.

Source: Ecology - nature.com