Drosophila suzukii, a widely distributed pest globally, causes significant economic losses to the fruit industry. The body color of D. suzukii exhibits plasticity, with the color changing according to the season: it is lighter in summer and darker in winter. Currently, research on the body color of D. suzukii mainly focuses on its functional aspects. However, a notable deficiency exists in current understanding of how plastic body color traits in D. suzukii are formed, highlighting a significant area for future research. Given the important roles of natural and sexual selection in the formation of insect traits, we deduced that both natural and sexual selection play key roles in the development of body coloration in D. suzukii. In order to test this hypothesis, we measured the dorsal abdominal melanization rate, chill coma duration, and mating success rate in D. suzukii and we also calculated the contributions of natural selection and sexual selection to the formation of plastic body color traits. Our results showed that the melanization rate of D. suzukii developed under 15°C temperature conditions is significantly higher than that of individuals developed under 25°C temperature conditions; the chill coma recovery duration of D. suzukii developed under 15°C temperature conditions is significantly shorter than that of individuals developed under 25°C temperature conditions; Natural selection and sexual selection accounted for 89.02% and 10.92% of the total selection differential, respectively. Thus, natural selection, rather than sexual selection, plays a dominant role in the formation of body coloration in D. suzukii. Our study is the first to quantitatively elucidate the relative contributions of natural and sexual selection to the formation of the plastic body color phenotype in D. suzukii. Future research could further investigate how natural selection influences the allele frequencies of body color-related genes in D. suzukii.

Fly stocks

Adult D. suzukii were collected from a Hui Wang blueberry orchard in Nanling, Anhui Province, China, in 2022 (30.92°N, 118.34°E). And all D. suzukii used in this study were collected from the wild and used in experiments within five generations. Under laboratory conditions, adults D. suzukii were independently reared in nylon cube cages (35×35×35 cm) at artificial climate chambers with 65% humidity. Three different temperature and photoperiod conditions were used as the experimental subjects: (1) 15°C with L:D =10h:14h, (2) 20°C with L:D (Light: Dark) =12h:12h, (3) 25°C with L:D =14h:10h. D. suzukii were fed a 240-mL plastic cup containing 80 mL of solid food (water: 1000.00 mL, corn flour: 44.67 g, sucrose: 71.33 g, agar: 16.00 g, yeast: 25.00 g, propionic acid: 6.30 mL, 100% ethanol: 22.30 mL) (Gonzalez-Cabrera et al. 2018), which was changed daily.

Dorsal abdominal melanization measurement

Thirty 3-day-old male and Thirty 3-day-old female D. suzukii were collected from each of the three temperature and photoperiod rearing artificial climate chambers. The dorsal side of the abdomen of D. suzukii was photographed under a microscope (VHX-5000, Osaka Japan, Keyence Corporation) at 30x magnification, then, all images were converted to grayscale using ImageJ (version 1.54g) software, where each pixel is represented by 8 bits, indicating a grayscale level from 0 to 255. We used the “Subtract Background” function to adjust the images and invert the brightness and darkness. At last, we set the threshold at 97 to mark the melanization area of the abdomen, so that the software could automatically calculate the abdominal melanization rate. All D. suzukii were collected and subsequently reared in artificial climate chambers after the melanization rate measurement to prepare for the chill coma recovery experiment.

Chill coma recovery experiment

4-day-old male and female D. suzukii adults were collected from each of the three temperature and photoperiod rearing chambers, with each individual placed separately in a 1.5 mL EP tube. And then, the EP tubes were placed in the - 20°C freezer (BCD-310, Qingdao China, Haier Corporation) compartment for 15 min and 30 sec (a duration that ensures cold-induced immobilization without causing death). Afterward, the EP tubes were quickly removed from the freezer and placed at room temperature, and the time from immobilization to standing was recorded for each D. suzukii. One hundred male and female D. suzukii from each of the three temperature and photoperiod rearing conditions were used in this experiment. Data from D. suzukii that did not recover from chill coma in this experiment will not be included in the analysis. All surviving D. suzukii were collected and reared in artificial climate chambers for 24 hr to prepare for the mate choice preference experiment.

D. suzukii mate choice preference experiment

6-day-old male and female D. suzukii adults were collected from each of the three temperature and photoperiod rearing chambers, with each individual placed separately in a 1.5 mL EP tube. We used non-choice experiments to test the mating preference of female D. suzukii for male D. suzukii. The non-choice experimental groups are as follows: (1) Randomly select 3 females reared at 15°C, pairing each with a male reared at 15°C, 20°C, or 25°C, respectively; (2) Randomly select 3 females reared at 20°C, pairing each with a male reared at 15°C, 20°C, or 25°C, respectively; (3) Randomly select 3 females reared at 25°C, pairing each with a male reared at 15°C, 20°C, or 25°C, respectively. Each pair is placed in a separate culture dish (35×10 mm). For each of the above three groups, 10 replicates were performed. As the morning is the peak mating period for D. suzukii (Revadi et al. 2015), the sexual selection experiments were conducted from 7:00 AM to 11:00 AM, during which the mate choice behavior of females was recorded over a 50-minute period.

Statistical analyses

The Kolmogorov-Smirnov test was used to assess the normality of the obtained melanization rates and chill coma recovery data. Subsequently, a one-way analysis of variance (ANOVA) was performed. To evaluate natural and sexual selection on the character of abdominal melanization rates, we employed the method proposed by Arnold and Wade (1984) (Arnold and Wade 1984). According to the theory of measuring natural and sexual selection, assessing selection on phenotypic characters does not require knowledge of their genetic background. The statistical relationship between fitness and phenotypic characters can be deduced from the change that occur within a generation in the mean, variances, and covariances of phenotypic traits (Arnold and Wade 1984). In this research, we view sexual selection as selection arising from variance in copulation success and natural selection as selection arising from variance in cold tolerance. Since natural selection and sexual selection act on different life stages of D. suzukii, so that whole selection procedure can be divided into two selection episodes. We regarded cold tolerance as the first selection force and copulation success as the second, because if D. suzukii is unable to survive cold environments, it would result in the loss of copulating opportunities.

The parameters and formulas used in this study are as follows (Table1, Table3):

(1) p, initial frequency distribution of the phenotypes, changes as a result of each episode: from p to p1, p1= pw1 after the first episode, to p2=p1w2 after the second episode and so forth;

(2) , the mean of phenotypic character before selection;  is a column vector representing the change in the mean of each phenotypic character from one generation to the next generation;

(3) S, selection differential, covariance between relative fitness and character, S = m-, m is the mean after m episodes of selection;

(4) Wk, the absolute fitness at episode k of selection;

(5) k , the average absolute fitness at the kth epiaode, k = pk-1Wk;

(6) wk, the relative fitness at episode k of selection, wk = Wk/k;

(7) , univariate selection gradient, partial regression of relative fitness on character;

(8) P, the phenotypic variance-covariance matrix assessed before selection, P = S/;

(9) , the standardized selection differential (intensity of selection),  = S/P1/2