We examined the relationship of post-diapause rearing temperature to developmental rate, survival, and adult body mass of the solitary wasp Isodontia elegans using prepupae from trap-nests. Isodontia elegans is a member of a genus often found in trap-nests in North America and Europe. Trap-nests are commonly used tools for studying cavity-nesting solitary wasps and bees. In temperate zones, progeny in nests usually overwinter as fully prepupae before pupating and emerging as adults. An important aspect of properly using trap-nests is determining temperatures that affect survival and health of developing offspring.  After overwintering >600 cocoons containing prepupae after the summers of 2015 and 2016, we placed cocoons on a laboratory thermal gradient where offspring experienced one of 19 constant temperatures from 643°C; emergence of adults was monitored for 100 days.  Our conservative estimate for the critical thermal minimum for development is 14°C, whereas that for the critical maximum is ~33°C. Prepupae transitioned to adults most rapidly at 29–33°C, but developmental rate was lower for some progeny exposed to temperatures ≥30°C. Offspring successfully reached the adult stage in <100 days at of temperatures of ~19–33°C. Adults from cocoons reared at lower temperatures weighed on average 6-10% more than expected based on their head widths, whereas those reared at higher temperatures weighed 4–10% less than expected. The difference may be due to greater rates of water loss and lipid metabolism during development at higher temperatures. Pre-overwintering cocoon mass was a significant predictor of relative adult body mass, indicating that adult health is partly related to their condition before overwintering. The trends we observed were similar to those for the bee Megachile rotundata, which we previously studied on the same gradient apparatus. However, data are needed on many other species of wasps and bees from a diversity of environments.  

Source of the wasps  

Isodontia elegans cocoons were obtained from trap-nests placed at sites within several km of Hillsboro, Oregon (45°31’ N, 122°59’ W; Spendal et al., 2021). Cocoons were removed from nests in late September in 2015 and early October in 2016. Cocoons were stored at ambient temperature in a barn in Oregon, before transport to Montana State University (Bozeman, MT) in November where they were kept at 8°C, 80% relative humidity until the following April 20 of each year. Cocoons were placed in #00 gelatin capsules, each with its long axis parallel to the capsule; the ends of the capsules had been punctured with pinholes to provide ventilation. In 2016, we measured cocoon mass (MC) prior to overwintering with a Metler AT 250 analytical balance accurate to 0.1 mg, and gave each gelatin capsule a unique numerical code.  

Rearing wasps on Thermal Gradient

On 20 April of 2016 and 2017, we removed gelatin capsules containing I. elegans cocoons from winter storage, at which time all healthy cells would have contained live prepupae. All cocoons were left at room temperature until the following day so that prepupae placed on the coldest and warmest ends of the gradient would not face a rapid temperature change. On 21 April, they were placed on the surface of a laboratory thermal gradient consisting of four parallel aluminum bars each measuring 15 cm wide × 100 cm long × 0.8 cm thick (Fig. 1). The ends of the bars sat upon two 2.5 cm wide, hollow, square cross-section aluminum rods oriented perpendicular to the long axis of the bars. Water from a constant temperature hot-water bath flowed in a loop through the two rods beneath one end of the bars, while another pumped a water/propylene glycol mixture from a constant temperature cold-water bath beneath the other ends of the bars. A 2.5-cm-high clear Plexiglas cover was fitted over each bar to stabilize temperature, but the covers fit loosely enough to allow sufficient air exchange. The entire gradient was enclosed in a larger wooden cover. Covers remained in place throughout the experiment, except for brief intervals when we removed gelatin capsules containing emerged I. elegans or parasitoid adults.

Temperature (T_B) was measured every 10 cm along the bars, from 0 to 100 cm with thermocouples inserted from below into the center of each bar. In 2016, we measured T_B at each position on the bars every two days during the experiment. Because T_B had remained so stable during the 2016 experiment, we measured it every four days in 2017. The mean difference between absolute values of consecutive measurements at each position on the bars was 0.24 ± 0.01°C in 2016 (N = 2200) and 0.32 ± 0.02°C in 2017 (N = 1100).  T_B extremes from the cold to hot end of bars ranged from 3.8–44.6°C, with some variation between bars. In linear regressions of T_B on the position along each bar (all dates combined), r² values ranged from 0.998 to 0.999 among the four bars in 2016 and from 0.994 to 0.999 in 2017. The relationship of T_B to position along the bars was consistent between years. The slope of regressions of T_B on position varied from 0.390–0.402 in both years. Rearing temperatures (T_R) experienced by developing wasps were estimated using a separate set of temperature measurements in which thermocouples were inserted into the center of a cocoon via a small hole in the side of a gelatin capsule. We set the capsule on the surface of the gradient (with the covers in place) and waited until the temperature had equilibrated to the nearest 0.1°C. We did this at all 19 positions while recording T_B at each position when the cocoon temperature had equilibrated. By regressing those temperatures on T_B, we created an equation for converting T_B to T_R. Average T_R experienced by developing wasps within the gelatin capsules at different 5 cm positions on the gradient ranged from 6.2–43.1°C in 2016 and from 6.1–43.2°C in 2017. 

On each gradient bar, we placed 19 rows of eight cocoons each, the rows 5 cm apart and aligned perpendicular to the long axes of the bars with the sides of adjacent capsules touching (Fig. 1). Before placement on the gradient, gelatin capsules were drawn blindly from a container, so that cocoons were randomly distributed. When on the gradient, the long ends of adjacent gelatin capsules in each row were in contact with one another. Each bar had 152 cocoons experiencing 19 temperature treatments. Gelatin capsules were anchored to the bars with double-sided cellophane tape, the center of each cocoon sitting on a line marking a 5-cm interval. The gradient was monitored daily so that adult I. elegans or parasitoid wasps could be placed in a freezer within their gelatin capsules soon after they emerged. In 2016, parasitoids included Melittobia sp. (Eulophidae) and Monodontomerus sp. (Torymidae) whereas in 2017 only Melittobia appeared, and in much lower numbers. Parasitoid-infested cocoons were discarded from analyses.

On 31 July each year, 100 days after cocoons were placed on the gradient, all cocoons from which adults had not emerged were dissected. We recorded the developmental stages of occupants of failed cocoons as 1) prepupa, 2) pupa, or 3) unemerged, fully developed adult. For prepupae, we recorded whether they were pale yellow, soft, and flexible (and thus potentially still living when they were placed in the freezer) vs. desiccated and hard (and thus clearly dead prior to freezing).  For each adult wasp that emerged, we measured its head width (HW) to the nearest ±0.05 mm using a stereomicroscope with an ocular micrometer and its body mass (MA) to the nearest ±1.0 mg to using a Smart Weigh® digital scale accurate to ±1.0 mg.

Data analysis

We characterized the relationship between T_R and the time elapsed from the end of overwintering and emergence of adults in two ways.  First, we used quadratic regressions to model the relationship between days to emergence (D) and TR for each sex and each year. From this, we also estimated for each sex a value for T_OPT, the temperature to the nearest 0.1°C at which wasps emerged most quickly (i.e., optimum temperature for rapid development). The results were also used to estimate temperatures at which male and female emergence dates were most similar. 

Second, to examine the relationship of developmental rate (1/D) to T_R, we used Briere-1 models, which estimate an array of parameters when the relationship of developmental rate to temperature is asymmetrical because it declines relatively abruptly after T_OPT is exceeded and likely mimics the asymmetrical performance curves of enzymes (Briere et al., 1999; Kontodimas et al., 2004; Quinn 2017). Unlike the quadratic model, the Briere-1 model allows estimation of not only T_OPT, but means and standard errors of 1) the lower developmental threshold (T_MIN, the temperature above which development proceeds after diapause) and 2) the upper developmental threshold (T_MAX, the temperature above which development ceases). The Briere-1 model takes the form 1/D = aT_R(T_R – T_MIN)(T_MAX – T_R)0.5, where a, T_MIN, and T_MAX are fitted coefficients. 

We used Kruskal-Wallis tests to test the hypotheses that the number of prepupae, pupae + unemerged adults, and adults at 100 days varied with position on the gradient (as a categorical variable). Each of the four bars was considered a replicate because of variation in the regressions of TB on position among bars. When plotting the numbers of each stage at different positions, we converted position to the mean T_R across all bars at that position during the 100 days so that an easier comparison could be made to the developmental rate plots. We tested the hypothesis that the sex of emerging adults varied with T_R using logistic regressions, coding males as 0 and females as 1. Means are based on data from the four gradient bars, on each of which there were eight cocoons per position at the start of the experiment. The number of cocoons used in the analysis for each bar ranged from 0 to 8 for each position, because the number of unparasitized wasps reaching each stage varied. The numbers were lower in 2016 when 32 positions (of 76 total) on all four bars had ≥2 wasps parasitized. However, just 4 of 76 positions experienced that level of parasitism in 2017, providing us with a clearer view of the effect of temperature on survival.

We tested the hypothesis that M_A, which varied even among wasps with identical head widths, was related to T_R. We began by regressing M_A on HW in separate analyses for each year and sex using linear regressions. We then calculated the body mass residual (M_RES) for each adult as the difference between the observed M_A values and the predicted values from the regression. This gave us an estimate of how much lighter or heavier each wasp was than expected based on its HW. For the 2016 data, we regressed MRES on T_R to test the hypothesis that M_RES varied as a function of rearing temperature (i.e., were wasps reared at higher T_R lighter or heavier than expected?). For the 2017 data, we considered both T_R and M_C as potential explanatory variables in multiple linear regressions. In addition, to estimate the magnitude of the effect of T_R on relative M_A, we divided the T_R range over which adults emerged into thirds. We then compared M_RES values between the coolest and warmest thirds of the T_R range using Mann-Whitney Tests, due to data being not normally distributed. Finally, for wasps in the cooler and warmer T_R groups, we calculated the average percent they were above or below the mean body mass (M_A) for the group. All means are reported ± standard errors (SE) and all statistical analyses were done using SigmaStat® v. 11.0, with the exception of the Briere-1 models, which were produced using TableCurve v. 5.01®. Because temperature and mass were normally distributed, all analyses of those were done on untransformed data.