2.1 Insect rearing and pupae preparation
Populations of L. trifolii and L. sativae were collected from field vegetable plants in Sanya, Hainan Province, China (18°14'52"N; 109°30'29"E). They were subsequently cultured on kidney bean plants (Phaseolus vulgaris L.) at 26±1°C, humidity at 75% with a 14:10 light: dark photoperiod. The two species were reared for more than 3 generations to ensure the homogeneity of each population before the two species were used in the experiments.
We used pupae to investigate the overwinter potential and cold tolerance of these two leafminer species. One-day-old pupae were used in the following experiments because this stage was developmentally mature and most sensitive to thermal stress (Parrella 1987; Zhao and Kang 2000).
2.2 Overwintering capacity along latitudinal gradients
To determine the overwintering capacity and overwintering boundary of the two leafminer species in China, we conducted overwintering experiments from November 2019 to April 2020 at 16 locations spanning 7 degrees of latitude from 30.23° to 37.43°N in Central and North China (Fig. 1). We used a specialized bottle (Fig. 2) to perform the overwintering experiments: we put the bottles into the topsoil (Fig. 2) of each location (Linden 1993). Each bottle contained 100 pupae and was treated as one replicate. Each location was replicated 5 times. We started the experiments in the northernmost location (Hebei Province) in early November 2019 and ended in the southernmost location (Zhejiang Province) in late November 2019. In April 2020, we collected pupae from north to south and placed them into a climate chamber at 26±1°C and humidity: 75% with a 14:10 light: dark photoperiod. We checked the survival rate of each replicate.
2.3 Overwintering capacity in common garden experiments
To compare the overwintering ability of the two species in northern China, we carried out common garden experiments under greenhouse and field conditions at the Langfang Experimental Station, Institute of Plant Protection (IPP), Chinese Academy of Agricultural Science (CAAS) (39°30'38”N; 116°36'2”E). We put bottles (Fig. 2) into the topsoil of experimental fields and greenhouses (Linden 1993). Each bottle contained 100 pupae and was treated as one replicate. We started the experiments at late December 2019. We collected the pupae of each replicate in fields and greenhouses every 10 days, and we ended the experiments after 100 days of exposure (early April 2020). Survival rate of each sampling was checked in laboratory. The combination of each species and duration of exposure was considered one treatment. Each treatment was replicated 5 times.
2.4 Cold tolerance capacity
To determine the cold tolerance capacities of the two leafminer species, their pupae were exposed to 5℃ and 0℃ for 1, 2, 3, 4, 5 and 6 days and exposed to -10℃ and -20℃ for 10, 20, 30, 40, 50 and 60 minutes. We considered the combination of each temperature and duration of exposure as one treatment. There were 5 replicates of each treatment, and each replicate contained 100 pupae. We checked the survival rate of each treatment under laboratory conditions.
2.5 Supercooling point determination
The supercooling points (SCPs) of the pupae of the two leafminer species were determined following the method described by Zhao and Kang (2000). We used a thermocouple (Testo, model 177-T4, Germany) to determine their SCPs. The individuals of the two species were fixed to the end of the sensor (nickel-worm) of the thermocouple by white petrolatum. Individuals were cooled at a rate of 0.5°C per minute in a programmable refrigerated test chamber. Any decrease in the temperature was recorded. The SCPs were defined as the lowest temperature before an exothermic reaction, as indicated by a sudden temperature increase. Each species was tested for 50 pupae. The SCP data were read using Comsoft3 Software.
2.6 Statistical analysis
We fitted a linear mixed model (LMM) with the lme function from the lme4 package (Bates et al. 2015) to compare the survival rate of overwintering pupae in the field and common garden experiments. To evaluate the overwintering capacity along latitudinal gradients, we fitted a linear mixed model with the survival rate as the response variable, species and region as fixed factors, and latitude as a random factor. To evaluate the overwintering capacity in the common garden experiments, we fitted linear mixed models with the survival rate as the response variable, species as a fixed factor, and duration of exposure as a random factor. In addition, we fitted linear models with the survival rate as the response variable and latitude as the explanatory variable.
To evaluate their cold tolerance capacities, we fitted linear mixed models with the survival rate as the response variable, species as a fixed factor, and duration of exposure as a random factor. All linear mixed model results were summarized with ANOVA. The SCP data were read using Comsoft3 Software. The independent-sample Student’s t-test was used for comparison of SCPs and for comparison of the survival rate of each experiment between L. trifolii and L. sativae.
All statistical analyses were conducted using R (version 3.6.3). We performed a normality test and residual test for all linear mixed models in R (package 'DHARMa', Hartig 2019). The results showed that all the data met the conditions for fitting the models. We also checked the SCP data for normality and equality of variances prior to statistical analysis.