Limited plasticity of Drosophila suzukii wing spot size in 3 response to developmental temperature 4

: 24 Phenotypic plasticity is an important mechanism allowing adaptation to new environments and as such 25 it has been suggested to facilitate biological invasions. Under this assumption, invasive populations 26 are predicted to exhibit stronger plastic responses than native populations. Drosophila suzukii is an 27 invasive species whose males harbor a spot on the wing tip. In this study, by manipulating 28 developmental temperature, we compare the phenotypic plasticity of wing spot size of two invasive 29 populations with that of a native population. We then compare the results with data obtained from wild- 30 caught flies from different natural populations. While both wing size and spot size are plastic to 31 temperature, no difference in plasticity was detected between native and invasive populations, 32 rejecting the hypothesis of a special role of thermal plasticity in the invasion success. In contrast we 33 observed a remarkable stability in the spot-to-wing ratio across temperatures, as well as among 34 geographic populations. This stability suggests either that the spot relative size is under stabilizing 35 selection, or that its variation might be constrained by a tight developmental correlation between spot 36 size and wing size. Our data show that this correlation was lost at high temperature, leading to an 37 increased variation in the relative spot size, particularly marked in the two invasive populations. This 38 suggests (I) that D. suzukii ’s development is impaired by hot temperatures, in agreement with the cold- 39 adapted status of this species; (ii) that the spot size can be decoupled from wing size, rejecting the 40 hypothesis of an absolute constraint and suggesting that the wing color pattern might be under 41 stabilizing (sexual) selection; (iii) that such sexual selection might be relaxed in the invasive 42 populations. Finally, a subtle but consistent directional asymmetry in spot size was detected in favor of 43 the right side in all populations and temperatures, possibly indicative of a lateralized sexual behavior. 44 45 46


24
Phenotypic plasticity is an important mechanism allowing adaptation to new environments and as such 25 it has been suggested to facilitate biological invasions. Under this assumption, invasive populations 26 are predicted to exhibit stronger plastic responses than native populations. Drosophila suzukii is an 27 invasive species whose males harbor a spot on the wing tip. In this study, by manipulating 28 developmental temperature, we compare the phenotypic plasticity of wing spot size of two invasive 29 populations with that of a native population. We then compare the results with data obtained from wild-30 caught flies from different natural populations. While both wing size and spot size are plastic to 31 temperature, no difference in plasticity was detected between native and invasive populations, 32 rejecting the hypothesis of a special role of thermal plasticity in the invasion success. In contrast we 33 observed a remarkable stability in the spot-to-wing ratio across temperatures, as well as among 34 geographic populations. This stability suggests either that the spot relative size is under stabilizing 35 selection, or that its variation might be constrained by a tight developmental correlation between spot 36 size and wing size. Our data show that this correlation was lost at high temperature, leading to an 37 increased variation in the relative spot size, particularly marked in the two invasive populations. This 38 suggests (I) that D. suzukii's development is impaired by hot temperatures, in agreement with the cold-39 adapted status of this species; (ii) that the spot size can be decoupled from wing size, rejecting the 40 hypothesis of an absolute constraint and suggesting that the wing color pattern might be under 41 stabilizing (sexual) selection; (iii) that such sexual selection might be relaxed in the invasive

52
Phenotypic plasticity often plays an important role in the adaptation to new environments (Lande, 53 2015;West-Eberhard, 1989). In particular, it has been repeatedly suggested to facilitate invasions 54 (Chown Steven L et al., 2007;Gibert et al., 2016;Richards et al., 2006), as genetic variation tends to 55 be limited during the first stages of colonization (Geng et al., 2007;Geng et al., 2016). However, this 56 possibility has received little empirical support so far (Arnold et al., 2019;Fox Rebecca J. et al., 2019).

57
Invasions provide a particularly favorable context for the study of the evolutionary role of plasticity as 58 we can compare derived populations repeatedly confronted to new environments with populations 59 from the native range, arguably closer from the ancestral state. Differences in reaction norms are 60 expected across populations (Pigliucci, 2005): native populations being predicted to be overall less 61 plastic than invasive populations (Davidson et al., 2011;Lande, 2015;Lee and Gelembiuk, 2008; 62 Richards et al., 2006;Yeh and Price, 2004).  The effect of sexual selection on phenotypic plasticity is controversial (Greenfield and 80 Rodriguez, 2004;Møller and Alatalo, 1999;Rowe and Houle, 1996). Although sexually selected traits 81 are often considered to be plastic (Price, 2006), an increased robustness might evolve through 82 stabilizing or directional selection (see Fierst (2013) for a theoretical treatment, and Nieberding et al.

83
(2018) for a case study in B. anynana). Stabilizing selection would limit variation around the values 84 preferred by females (Mead and Arnold, 2004), while directional selection would limit variation on one 85 side of the distribution due to female preferences and on the other side due to developmental or 86 physiological constraints (Wiernasz, 1989). We reasoned that D. suzukii's spot might be more plastic, 87 or at least less canalized and thus more variable in the invasive populations compared to native 88 populations, owing to a hypothetical reduced choosiness in femalesand thus reduced sexual

90
Here we quantify the phenotypic variation of the wing spot size in several natural populations 91 of Drosophila suzukii and assess its plasticity to temperature using samples from three geographic 92 populations reared in the lab. Temperature is one of the main environmental drivers of life history and 93 morphological evolution of drosophilids in particular and ectotherms in general (Atkinson, 1994;Crill et 94 al., 1996;David et al., 1997;Gibert et al., 1996;Gibert et al., 2007)

136
Note: The first two columns show the geographic origin of each population (first) and the temperature 137 set during their development in-vitro (second). Then, the total number of wings used to analyze overall 138 variation and the number of complete individuals used to analyze asymmetric variation.

140
Phenotypic data 141 Two phenotypic traits were directly inferred from the pictures: wing size and spot size ( Figure 1).

142
Because the potential visual effect of the wing spot during courtship may be associated to the 143 proportion of the wing occupied by the spot and not its absolute size, we also estimated the relative 144 spot size as the ratio between spot size and wing size for each individual.

145
In order to remove the proximal part of the wing, which is usually deteriorated during the

196
To explore the hypothesis that temperature affects asymmetry, we measured the difference 197 between right and and left values for each of the three phenotypic traits (Table 1). Then, a two-way 198 mixed model ANOVA was applied to each trait (Palmer and Strobeck, 1986;Palmer 1994). "Individual" 199 was entered the model as a random effect, and tested for the variation among individuals; "side" was 200 treated as a fixed effect, and tested for a systematic deviation from perfect symmetry (directional 201 asymmetry, DA); the interaction "side x individual" tested for the significance of non-DA relative to 202 measurement error (Fluctuating asymmetry, FA). The replicated measurements were included in the 203 error term. Once the presence or absence of significant DA and FA were assessed in each particular 204 group, we assessed the effect of temperature and geography on these two traits with an ANOVA 205 (effect on DA). Differences in the slope of the relationship temperature-ratio asymmetry among 206 populations with three pairwise t-test. Finally, Levene's tests were applied on the asymmetry 207 measures to assess FA.

208
As effect sizes, Cohen's d for differences among geographic populations and r 2 for 209 temperature effects were reported due to their simplicity (Rosenthal et al., 1994). All data, scripts and        The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/800417 doi: bioRxiv preprint

257
Within temperature variation tended to be higher at 28°C for both the spot size and the ratio.

284
No DA was detected for wing size. In contrast, significant DA was detected for both absolute and ratio,

301
Our results revealed the existence of a relatively large natural variation in wing and spot sizes and 302 temperature may explain a substantial part of it. In contrast, the spot-to-wing ratio was quite stable in 303 nature and very robust to developmental temperature in the lab. This stability might either reflect a 304 tight structural connection between the development of the spot and that of the wing, imposing a 305 strong correlation (i.e. constraint hypothesis), or a history of stabilizing selection on the relative spot 306 size (i.e. adaptive hypothesis). All three populations exhibited an increase in relative spot size variation 307 at 28ºC, indicative of a lesser developmental robustness at high temperature, in agreement with the 308 status of cold-adapted species of D. suzukii (Enriquez et al., 2018;Jakobs et al., 2015;Shearer et al., 309 2016;Stephens et al., 2015). This increased variation originating from a de-correlation of spot size and 310 . CC-BY-ND 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/800417 doi: bioRxiv preprint wing size at 28°C suggests that the stability of the relative spot size is not an absolute structural 311 constraint, and therefore points at the adaptive hypothesis. D. suzukii males wing color pattern might 312 thus be under stabilizing (sexual?) selection. Native and invasive populations showed very similar 313 reaction norms, suggesting that spot plasticity is not affected byand played no role in the success of 314 the invasion history. Relative spot size variability was higher in invasive populations compared to the 315 native population, possibly pointing at a relaxed sexual selection during the invasion. Finally, both 316 natural and laboratory samples exhibited a slight but significant DA of spot size (both absolute and 317 relative), consistently in favor of the right side.

319
Relative spot size robustness: selection or constraint? 320 Although our experimental populations show a slight but significant decrease in the ratio at both high 321 and low temperatures, the overall pattern reflects a global stability to developmental temperature, 322 contrasting with both wing size and the absolute spot size that are both strongly plastic to temperature.

323
Such robustness might suggest either that spot size is fully determined by wing size, owing to 324 the tight connection between wing development and the spot formationan example of 325 developmental constraintor alternatively, that stabilizing selection is acting on D. suzukii males color 326 pattern, removing individuals with extreme relative spot sizes.

327
The male-limited presence of the wing spot in all species has been interpreted as a mark of 328 sexual selection (Kopp and True 2002). The presence of a spot is indeed phylogenetically tightly 329 associated with the occurrence of elaborate courtship behaviors, the males dancing in a stereotypic 330 way in front of females and displaying their wings (Kopp and True, 2002;Revadi et al., 2015).

331
Although the empirical evidence of an effect of the spot on mating success is scarce (Fuyama, 1979; 332 Roy and Gleason, 2019), it is nevertheless conceivable that females choice might be influenced by the 333 male wing phenotype. Sexual selection may have favored the stability of the relative spot size and 334 reduced its plasticity, a common process (Fierst, 2013) although not universal (Price, 2006). This 335 would suggest that the focal trait might not be the spot size itself, but rather its extent on the wing, i.e.

336
the wing color pattern.

337
. CC-BY-ND 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/800417 doi: bioRxiv preprint Alternatively, it is possible that spot size plasticity is simply a structural consequence of wing size 338 plasticity to temperature, which has been suggested to be adaptive (Crill et al., 1996;David et al., 339 1997), and the fact that the boundaries of the expression of the genes involved in the spot formation 340 vary according to the positional information of the veins (Arnoult et al., 2013;Gompel et al., 2005).

341
Such developmental dependency of spot formation on wing development would thus impose a tight 342 correlation between spot size and wing size, leading to a structurally stable spot relative size. This 343 would thus be an example of an absolute constraint on the variation of the wing color pattern.

344
The analysis of the ratio variation within temperature provides an element in the discussion. It 345 was indeed found that this variation increases at high temperature in the three populations (Figure 3c).

346
This effect may indicate that hot temperatures might destabilize the processes involved in the spot 347 formation, which is in agreement with the idea that D. suzukii is adapted to cold temperatures (Jakobs

359
The hypothesis of adaptive plasticity in invasive populations (as opposed to native ones) has 360 frequently appeared in the literature (Davidson et al., 2011;Lande, 2015;Lee and Gelembiuk, 2008; 361 Richards et al., 2006;Yeh and Price, 2004). However, our results only show very small differences in 362 wing size plasticity between invasive and native populations.  The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/800417 doi: bioRxiv preprint