Effects of [i]in situ[i] climate warming on monarch caterpillar ([i]Danaus plexippus[i]) development

Climate warming will fundamentally alter basic life history strategies of many ectothermic insects. In the lab, rising temperatures increase growth rates of lepidopteran larvae, but also reduce final pupal mass and increase mortality. Using in situ field warming experiments on their natural host plants, we assessed the impact of climate warming on development of monarch (Danaus plexippus) larvae. Monarchs were reared on Asclepias tuberosa grown under 'Ambient' and 'Warmed' conditions. We quantified time to pupation, final pupal mass, and survivorship. Warming significantly decreased time to pupation, such that an increase of 1˚ C corresponded to a 0.5 day decrease in pupation time. In contrast, survivorship and pupal mass were not affected by warming. Our results indicate that climate warming will speed the developmental rate of monarchs, influencing their ecological and evolutionary dynamics. However, the effects of climate warming on larval development in other monarch populations and at different times of year should be 32 'Ambient' and 'Warmed' conditions. We quantified time to pupation, final pupal mass, and 33 survivorship. Warming significantly decreased time to pupation, such that an increase of 1˚ C 34 corresponded to a 0.5 day decrease in pupation time. In contrast, survivorship and pupal mass 35 were not affected by warming. Our results indicate that climate warming will speed the 36 developmental rate of monarchs, influencing their ecological and evolutionary dynamics. 37 However, the effects of climate warming on larval development in other monarch populations 38 and at different times of year should be investigated. Here, we report results from an in situ warming experiment designed to assess how 81 climate warming influences growth, survival, and development of monarch larvae. We 82 hypothesized that warming would reduce larval development time, as has commonly been 83 reported for monarch larvae , but would also decrease pupal mass and 84 survivorship (Zalucki 1982, York and Oberhauser 2002). In August 2014, monarch eggs and larvae were gathered from A. syriaca within nearby 109 old-growth fields. Eggs and larvae were reared in mesh cages and fed fresh A. syriaca leaves 110 daily until they reached the third instar. Larval development was checked continuously 111 throughout the day. Immediately after molting to the third instar, larvae were randomly assigned 112 to a temperature treatment ('Ambient', 'Warmed') and placed on a single A. tuberosa within a 113 randomly chosen plot (n = 15, n = 18 for 'Ambient' and 'Warmed' treatments, respectively). A 114 mesh bag was placed over the plant to retain the monarch. First or second instar larvae escaped 115 the mesh bags easily and thus were not used. If the monarch consumed the entire host plant, they 116 were transferred to another plant within the same subplot. Time to pupation was recorded as the 117 number of hours between experiment initiation and onset of chrysalis formation, and this number 118 was converted to number of days (development hour / 24). Dead individuals were recorded and 119 removed from the host plant. Chrysalids were carefully transported back to the lab and weighed 120 to obtain final pupal mass. 121 We measured three plant traits (specific leaf area (SLA), water content, and latex 122 production) to determine whether warming effects on monarch development might be mediated 123 through warming effects on plant traits. At the end of the experiment, two newly expanded 124 leaves were collected from each plant. For one leaf, we measured leaf area, obtained a fresh wet 125 mass, and then dried the leaf to obtain a dry mass. We calculated specific leaf area (SLA) as area 126 / dry mass and percent water content as (1 -dry mass (g) / fresh mass (g))*100. Using the second 127 leaf, we determined latex production by cutting the tip of the leaf and blotting all latex onto a 128 dry, pre-weighed piece of filter paper. The filter paper was dried again and latex concentration Although heaters raised temperatures of 'Warmed' plots by ~4˚ C on average, plots 132 varied considerably in temperature due to different light levels across the experimental garden 133 and varying plant biomass within each plot. We therefore measured temperature with a handheld 134 infrared thermometer in each subplot during the night at the end of the experiment. For 135 consistency, we recorded temperature of a white plastic sphere mounted 0.5 m from the ground 136 in the middle of each subplot. We then treated temperature as a quantitative, rather than 137 categorical, variable in all analyses. Note that these measures reflect relative differences in 138 temperature among plots that should be relatively constant over the experiment. 139 We regressed all response variables against night-time temperatures as measured by the 140 IR gun using OLS regressions. We regressed mortality against temperature using logistic 141 regression, where the response variable was dichotomous with survival = 0 and dead = 1. 142 Although monarchs experience mortality as pupae, brief exposure to prolonged temperatures did 143 not alter pupal mortality rates and third instar individuals were the most sensitive to temperature 144 increases (York and Oberhauser 2002). Thus, our experiment likely captured most of the 145 influence of temperature on larval survival. 146 Model assumptions were verified with residual plots where appropriate. All analyses 147 were conducted using Python v2.7 with the 'numpy', 'pandas', and 'statsmodels' modules 148 (McKinney 2010, Seabold and Perktold 2010 151 Time to pupation declined rapidly with increasing temperature (p < 0.001, R 2 = 0.57) 152 169 Warming had no effect on any measured plant trait. SLA (p = 0.940, R 2 = 0), percent 170 water content (p = 0.313, R 2 = 0.05), and latex concentration (p = 0.739, R 2 = 0.01) all did not 171 vary with temperature. Thus, any effects of warming on monarch development time were direct 172 effects of temperature on monarch physiology rather than being mediated through the plant traits 173 we measured. 189 temperatures above 28˚ C induced high mortality rates in monarch larvae (Zalucki 1982, York 190 and Oberhauser 2002). However, these studies used either pulses of extremely high temperatures 191 (i.e. 36˚ C) or held monarchs at a constant temperature (i.e. 28˚ C). Ambient, maximum daytime 192 temperatures averaged 30 ˚C during our experiment; warming increased this maximum to 32-34˚ 193 C. Although these temperatures are above the thermal optimum of monarch survival, we found 194 no effect of in situ warming on either pupal mass or survival. As temperatures exceeded 28˚ C 195 for less than 20% of the full 24 hour day, it is likely that diel and daily temperature fluctuations 196 mitigated the lethality of high temperatures