The vulnerability of overwintering insects to loss of the subnivium

Aim: Winter climate change threatens the subnivium (i.e., the microhabitat that exists between the snowpack and the ground), and the community of species that depends on it for overwintering survival. One group of species that will likely exhibit an array of responses to subnivium loss are overwintering insects because they vary in their cold tolerance strategies and lower thermal limits. For an assemblage of eight insect species that range in their cold tolerance strategies and include both pollinators and pests, we investigated species-specific vulnerabilities to shifting subnivium conditions.

Location: Great Lakes region of North America

Methods: We applied information on each insect’s supercooling point to spatially- and temporally-explicit models of minimum subnivium temperatures generated from active-warming experiments and comprising three scenarios: current conditions (i.e., control), +3°C, and +5°C.

Results: Although species varied in their vulnerabilities, our predictions indicated that exposure to lethal temperatures generally decreased under warming of 3°C, but increased under warming of 5°C, indicating that once enough warming happens, a tipping point is reached. We also found that freeze-tolerant species (i.e., species that can survive at temperatures below their supercooling point) possess a more cryptic vulnerability to winter climate change because sustained below-freezing temperatures were sufficient to induce vulnerability (i.e., predicted mortality), even when temperatures were above the supercooling point.

Main conclusions: This work provides a better understanding of the vulnerability of different insect species to winter climate change, which is critical because overwintering survival and the fitness consequences incurred during overwintering likely represent important bottlenecks for the population dynamics of subnivium-dependent species.

Methods
Active-warming Experiments

Experimental Design

In the fall of 2016-17, we installed three micro-greenhouses (2.5 m x 2.5 m x 3 m) at each study site for a total of 27 greenhouses. Each greenhouse had an aluminum frame with walls of corrugated plastic, and was equipped with a pair of heaters and vents positioned on opposite walls. At each site, greenhouses were separated by a minimum of 4.5 m, and consisted of three temperature treatments: control (GHcontrol, internal temperature = ambient temperature), 3°C warmer than ambient (GH+3°C), and 5°C warmer than ambient (GH+5°C). We also monitored the environment external to all greenhouses to capture current conditions (hereafter external).

Inside each greenhouse, we attached a temperature probe within a radiation shield (Davis Instruments Corp External Temperature Sensor), an anemometer (Davis Instruments Corp Anemometer), and a snow depth sensor (HRXL-Max Sonar WRS Series Ultrasonic Snow Depth Sensor, typical accuracy of 1%). Temperature (°C) and wind speed (m/s) were measured every minute, while snow depth (cm) was measured every five minutes. We also established three weather stations at each site and paired each station with a greenhouse. Weather stations had a heated rain gauge for measuring liquid precipitation (mm, Davis Instruments Corp Rain Collector and Rain Collector Heater), an anemometer (m/s, Davis Instruments Corp Anemometer, mounted at mean height of 1.8 ± 0.3 meters), and a temperature probe within a radiation shield (°C, Davis Instruments Corp External Temperature Sensor, mounted at mean height of 1.6 ± 0.2 meters). Measurements from these instruments were also recorded every minute. At one of the three weather stations at each site we also attached a snow depth sensor to measure the snow depth (cm) external to the greenhouses, at five minute intervals. All instruments recorded data from December 2016 through March 2017.

Temperature and Precipitation Regulation

All instruments on the weather station and inside the greenhouse were connected to a central control box. This box monitored temperatures from the greenhouse (i.e., experimental temperature) and the weather station (i.e., ambient temperature) every minute and increased heating or venting within the greenhouse to maintain a set experimental temperature relative to ambient conditions. In addition to temperature controls, greenhouse automation included a retractable roof to capture all precipitation events during the winter season. Within the rain gauges at each weather station, a wetness sensor logged voltage measurements at 1-minute intervals. Once a positive voltage was recorded, the roof of the greenhouse opened to allow precipitation to fall inside. When the voltage measurement returned to zero, indicating that there was no more precipitation, the roof closed.

Subnivium Temperature     

We positioned temperature strings containing 20 individual temperature probes at each greenhouse to measure subnivium temperatures. Each probe was separated by 0.3 m, and was staked to be flush with the ground. We arranged the temperature strings so that 16 probes were fastened to the ground inside of each greenhouse, while 4 probes were fastened to the ground outside of each greenhouse, leading to a total of 16 subnivium probes for each of the greenhouse treatments and 12 probes for external subnivium conditions. Subnivium temperatures were recorded at five-minute intervals from December 2016 through March 2017.

To derive a daily subnivium temperature for each treatment, we extracted the daily minimum ground temperature from each sensor for the period of December 1, 2016, to March 31, 2017 and then calculated the mean of those minimum temperatures for each treatment (environmental control, n = 12; greenhouse treatments, n = 16) and each day.