Death feigning is fairly common in a number of taxa, but the adaptive significance of this behaviour is still unclear and has seldom been tested. To date, all proposed hypotheses have assumed that prey manage to escape predation by sending a death-mimicking signal, although death-feigning postures are markedly different from those of dead animals. Moreover, the efficacy of this technique may largely depend on the foraging mode of the predator; death feigning seldom works with sit-and-wait predators that make the decision to attack and consume prey within a very brief time. We examined whether death feigning in the pygmy grasshopper Criotettix japonicus Haan was an inducible defence behaviour against the frog Rana nigromaculata, a sit-and-wait, gape-limited predator. The characteristic posture assumed by the grasshopper during death feigning enlarges its functional body size by stretching each of three body parts (pronotum, hind legs and lateral spines) in three different directions, thereby making it difficult for the predator to swallow the prey. Our result is the first consistent explanation for why death-mimicking animals do not always mimic the posture of dead animals.

The population dynamics of two major rice bug species in the fields of Italian ryegrass and gramineous weeds were compared in southern Ibaraki Prefecture, Japan from 2002 to 2004. Overwintered adults of Leptocorisa chinensis appeared in grass fields from late June to early July. Their progeny appeared soon, and the population fluctuation until October was bimodal. In contrast, overwintered adults of Cletus punctiger appeared much earlier, in early May, while the next-generation nymphs were seldom detected until late June. Based on the immigration dates of overwintered adults and the cumulative effective temperature for development and ovarian maturation, it was confirmed that L. chinensis is bivoltine in the study area. The first appearance date of first-generation adults of L. chinensis in grass fields did not coincide with the heading date of mid-season rice cultivated most widely in southern Ibaraki Prefecture. This reveals that current rice cultivation practices in this area are suitable to prevent the massive invasion of L. chinensis.

To clarify the relationship between the occurrence patterns of three species of rice bugs and the developmental stage of panicles in rice fields, we carried out two experiments. In the first experiment, we observed the development of spikelets in three paddy fields (Plots A, B, and C). The rice variety 'Sainohana' was used in Plot A and 'Koshihikari' was used in Plots B and C. Dates of transplanting were on April 25, May 8, and June 19; for Plots A, B, and C, respectively. Spikelets were classified into three stages based on ovarian development: Stage I (initial), Stage II (middle), and Stage III (fullness). After the initial heading stage, changes in the average number of each stage of spikelet development were expressed in days or cumulative degree-days above a base of 10°C. In the second experiment, the incidence of each rice bug species was monitored in rice fields. Then, the relationship between rice bug incidence and cumulative degree-days was determined. Variations in the average number of spikelet development stages among the three plots expressed in cumulative degree-days were smaller than those expressed in days after the initial heading stage. The peaks of incidence of adult rice bugs were near the peak of Stage I at 50-150 degree-days. The incidence of nymphs began to increase near the peak of Stage II at 200-250 degree-days. The occurrence patterns of the rice bugs were found to be closely related to the abundance of the three stages of spikelets. Cumulative degree-days can be used to uniformly express the changes in the incidence of rice bugs among different rice varieties, seasons, and fields.