Supplementary Material
Larval performance without competition
Methods
The first experiment compared the performance of larvae without competition (one larva per seed). We placed mated females that emerged within 24 hours into a Petri dish with one layer of beans to obtain seeds bearing one larva (no competition). They were allowed to oviposit for 1 hour and were then removed. Seeds bearing one hatched egg were collected 10 days after oviposition and kept in the growth chamber. Emerged adults were counted daily. Three dependent variables were recorded: the developmental periods until adult emergence, survival to adult emergence and body size as measured by dry weight. We determined the survival rate based on logistic regression analysis with two factors, the host seed and the beetle species. A three-way ANOVA was used to test the effects of the host seeds, the beetle species and the sex on the larval development time and on the adult body size at emergence.Development time was transformed to development rate which was calculated as the inverse of the development time before the analysis because development time had a skewed distribution. Sample sizes for each treatment group ranged from 50 to 103.
Results
Survival of both species from egg hatch to adult emergence was generally high (95%–100%). There was no significant effect of the resource bean (the azuki, V. angularis, or the blackeye bean, V. unguiculata) or beetle species (C. maculatus or C. chinensis) on survival. The development time was significantly shorter in the blackeye bean (P 0.001)(Fig.S1a). The effects of sex and beetle species were also significant (P 0.001 for each factor) because females took slightly longer to develop than males, and C. chinensis developed faster than C. maculatus. Differences in mean development time between the two species tended to be larger in the blackeye bean than in the azuki bean, as indicated by the significant interaction between beetle species and bean type (P 0.01).
Body size of C. chinensis and C. maculatus responded differently to the type of bean (Fig.S1b). Adults of C. chinensis were larger when they developed in the blackeye beans compared with the azuki beans, whereas this was not true for C. maculatus. This was indicated by a significant interaction between the effects of beetle species and bean type (P 0.001). The effect of sex and beetle species and the interaction between beetle species and sex were also significant (P 0.005).
Fig. S1 a) Mean development time ( SE) and b) mean body mass at emergence ( SE) of C. chinensis (Cc) and C. maculatus (Cm). Beetles developed with no competition and were tested on the azuki bean (filled bars) and the blackeye bean (open bars).
Density dependence of fecundity and egg survival to hatched larvae
Methods
The numbers of eggs laid and eggs hatched were fitted to the intraspecific reproductive model as follows:
Ec ecNc / {1 ecNc }bec
Em emNm / {1 emNm }bem
Hc hcNc / {1 hc Nc }bhc
HmhmNm / {1 hmNm }bhm,
where Nc and Nm are the number of introduced adults of C. chinensis and C. maculatus respectively, Ec and Em are the number of eggs, and Hc and Hm are the number of hatched eggs of C. chinensis and C. maculatus. The maximum growth rate is ei for the egg number and hi for the hatched egg numbers of species i. The parameters a and b are the empirical scale parameters.
Results
Figure S2 shows the effect of intraspecific density dependence of four combinations of bean type and beetle species. Regarding oviposition behaviour, the adults of C. chinensis laid fewer eggs than did C. maculatus when they were at high density. The hatchability differed depending on the type of bean used. Both beetle species had a greater scramble tendency on the azuki bean than on the blackeye bean, which was indicated by a larger value of b estimated on the azuki bean. The number of hatched eggs was substantially larger than the number of adults emerged.Thus, larval resource competition can be largely responsible for the outcome of competition for the whole generation.
Fig. S2. Effect of adult density on the total number of eggs laid (), hatched eggs (open circles) and emerging adults (filled circles) for (a)Az-Cc, (b)Be-Cc, (c)Az-Cm, (d)Be-Cm. Estimated parameters for the number of eggs are: (Az–Cc),= 41.1, a = 0.02, b = 0.85; (Be–Cc), = 35.8, a = 0.01, b = 1.02; (Az–Cm), = 47.0, a = 0.48, b = 0.29; (Be–Cm), = 54.3, a = 0.18, b = 0.55. Estimated parameters for the numbers of hatched eggs are: (Az–Cc)= 32.1, a = 0.004, b = 3.15; (Be–Cc)= 31.2, a = 0.01, b = 1.24; (Az–Cm)= 30.0, a = 0.01, b = 2.42; (Be–Cm)= 39.4, a = 0.05, b = 0.78.
1