Marking the larval reserves

Third instar hosts were extracted from the beans and injected with either 1µl of a 3Hmarked amino acid mixture (37MBq ml-1 , ICN Pharmaceuticals) or with 2µl of a 14C-marked amino acid mixture (3.7MBq ml-1 , ICN Pharmaceuticals). The amino acid mixtures had been previously diluted with Ringer’s solution to a total activity of 7kBq ml-1 (for the 3H mixture) or 3.5kBq (for the 14C mixture) so that the total activity injected in each larva was in both cases 7kBq. Injections were carried out using a graduated micro capillary connected to a manual pump and with the aid of a binocular microscope. Injections took place on the midlateral side of the host’s body which, our own trials showed, eliminated almost entirely the loss of bodily fluids through the wound. Larvae that “bled” profusely after the injection were immediately discarded. Injected larvae were kept at room temperature for a minimum of two hours to allow the wound to scar over and the distribution of the radioactivity within the body. Parasitoid eggs were obtained by individually placing D. basalis females in small Petri dishes (diameter 5.5 cm) with two non-radioactive third-instar larvae of C. maculatus placed within an artificial bean made from a gelatine capsule (for details of how hosts are prepared inside the capsule see Gauthier & Monge 1999). Females were left to oviposit on the hosts for 24 h after which time one egg of each female was collected and transferred onto a randomly chosen 3H- or 14C-injected larvae. The rest of the eggs were used as a control for background radiation (see below). Although parasitoids will readily oviposit directly on injected larvae, this technique was preferred because it reduced considerably the manipulation of the injected larvae; parasitoids usually lay 3-4 eggs per host in a 24 h period, often on the underside of the larva, which therefore needs to be taken off the gelatine capsule and turned in order to completely eliminate any eggs in excess of one. Once the egg had been transferred onto the surface of the injected larva, the larva was placed inside an artificial bean (as above) and the bean was kept inside a small Petri dish at 13L:11D photoperiod, 33:23oC temperature and 75% humidity until the emergence of the parasitoid. The time taken by D. basalis to develop from egg to adult inside the gelatine capsule is the same as the normal developmental time inside a real bean, 15-16 days. The adult parasitoid breaks free from the artificial bean and into the Petri dish by biting off the 30 Chapitre 2 gelatine capsule. A total of 6 female parasitoids with its larval reserves marked with 3H and 8 female parasitoids with its larval reserves marked with 14C were obtained in this manner. 

Marking the adult reserves

 One day after emergence, the parasitoids with their larva l reserves marked were individually weighed inside a gelatine capsule and then placed in a Petri dish (5.5 cm) with an artificial bean containing a third instar host under the above-mentioned temperature, humidity and photoperiod. The hosts had been previously injected with either 3H or 14C amino acid mixture (as above). Females that had their larval reserves marked with 3H were provided with 14C-injected hosts and vice versa. Preliminary trials had shown that females both feed and oviposit on these injected hosts. Egg laying and host feeding cannot be decoupled in this species, and although females seem to make feeding tubes at least some of the times, these feeding tubes are easily broken and thus frequency of host feeding cannot be estimated. Each day the host was removed, the eggs were extracted and the female provided with a new injected host. This process was repeated daily until the death of the parasitoid. The gelatine capsules containing the eggs were stored on a daily basis at -80oC for later analysis. At the end of the experiment the female was also stored at -80oC to analyse the radioactivity remaining in the body. 

Sample preparation and radioactivity quantification 

Quantification of the amount of each isotope incorporated into the eggs was carried out using a liquid scintillation analyser (LSA, TriCarb1900, Packard Instruments). The batch of eggs laid by each female on each day were prepared by crushing them together in a liquid scintillation tube and immediately adding 100µl of a tissue solvent (Solueneâ-350, Packard). After 30 min, 1ml of the liquid scintillation cocktail (Hionic-Fluorä, Packard) was added. In order to control for background radiation in the environment, for each batch of eggs prepared one extra tube containing the tissue solvent and the liquid scintillation cocktail but no egg was used. After one hour, all tubes were read in the LSA for 15 min each and the number of desintegrations per minute (DPMs) on each side of the energy spectrum (0-15 KeV for 3H and 16-156 KeV for 14C) obtained. An external standard was used as a quench-indicating parameter (tSIE: the transformed Spectral Index of the External standard). In addition, an automatic efficiency control function (AEC) that makes use of the tSIE values to adjust the lower and upper energy boundaries of each isotope to compensate for differences in quench  levels between the isotopes was selected. Since larval and adult resources are marked in each female with a different isotope, each with its own specific activity and metabolic pathway, it is not possible to calculate relative investment of larval and adult food sources into eggs in terms of proportions. DPM’s for 3H and 14C are thus reported separately. Further, in order to control for the presence of natural radiation in the eggs, a series of control tubes were prepared by crushing unmarked eggs either singly, or in batches ranging from 2 to 7 eggs, and analysing them separately in the same way as the experimental eggs. The mean radioactivity in these unmarked eggs was very low (1.52 ± 0.69 DPMs for 3H and 5.56 ± 1.06 DPMs for 14C after correction for environmental radioactivity, n=7 in both cases), and was independent of the number of eggs in the batch (Kruskal Wallis test c 2 6 = 7.0, NS for 3H and c 2 6 = 2.3, NS for 14C). This lack of correlation between DPM and the number of unmarked eggs in the batch suggests that in fact the DPM readings are precision errors in the radioactivity measures of the LSA rather than natural radiation associated to parasitoid eggs. However, in order to obtain a stringent measure of what was the radioactivity incorporated into the eggs as a result of our experiment, as opposed to background noise or simply error detection in the LSA, these measures were taken into account. For each isotope, the radioactivity incorporated per egg as a result of our experimental treatments was calculated by dividing the DPM of the egg batch, as calculated by the LSA, by the number of eggs in the batch, and subtracting the mean DPM of unmarked eggs (either 7.0 DPMs or 2.3 DPMs depending on the isotope). The females were analysed to determine the amount of resources from larval and adult signature not invested in eggs and remaining in the body at the end of the experiment. For this purpose each female was separated into two fragments, the abdomen and the rest of the body, which were crushed, prepared and read in the LSA separately following the same exact procedure as for the eggs. In order to avoid loosing radioactivity from the body as a result of a dissection, the mature eggs remaining in the ovaries were not dissected out of the abdomen before this was crushed and read in the LSA. The values for the amount of nutrients remaining in the abdomen tissue and fat body may thus be slightly overestimated. In addition, in order to estimate the amount of larval reserves present in the body of the females at emergence, an additional group of females with the larval reserves marked with 3H (n=10) and 14C (n=10) was obtained as above. These females were killed on the day of emergence and the radioactivity present in the two body fragments (head-thorax and abdomen) quantified using the same procedure. The hind tibia length of all females involved in the experiment was 32 Chapitre 2 measured in order to control for size-related differences in resource acquisition or allocation. Non-parametric tests reported were carried out using SPSS.

Results

The mean weight and tibia length of females emerging from hosts injected with 3H and 14C did not differ significantly (mean ± s.e. 1.27 ± 0.07 mg and 1.25 ± 0.06 mg weight respectively, Mann-Whitney U=15.0, NS and 0.74 ± 0.014 mm and 0.76 ± 0.004 mm respectively, Mann-Whitney U=10.0, NS). Females also laid a similar mean total number of eggs in both treatments (mean ± s.e. 64.38 ± 6.62 and 75.00 ± 6.08 for females laying eggs and host feeding in 3H- and 14C-injected hosts respectively Mann-Whitney U= 13.5, NS). The mean number of eggs laid per day varied from 4.80 ± 0.37 and 3.89 ± 0.35 (mean ± s.e. for females laying eggs and host feeding in 3H- and 14C-injected hosts respectively) on the first day, to a maximum on day 5 (8.20 ± 0.37 and 8.25 ± 1.03 respectively). Egg laying decreased monotonically from day 5 onwards so that towards the end of the experiment (day 14) the mean number of eggs laid was 4.67 ± 0.67 and 3.50 ± 0.50 (for females laying in 3H and 14C larvae).