We have found that pupil dynamics differ significantly among the three differently sized tortricid moths. Furthermore, temperature affects pupil dynamics differently in each species. Eye size in these small microlepidoptera, as estimated by eye radius r, is close to (230 and 290 µm in LB and CP) or below (190 μm in GM) the proposed theoretical minimal size for a functional optical superposition insect eye (r > 250 µm; Meyer-Rochow and Gál 2004). In spite of the calculated lower limit, the superposition pupil is well resolved even in the smallest species, GM (Fig. 1), indicating that further theoretical and experimental work is needed to fully understand optical superposition eyes. All three species have functional superposition optics, but because pupil aperture decreases with decreasing eye size, the species with the smallest superposition aperture should have the lowest light sensitivity, following the sensitivity formula of Warrant and Nilsson (1998), which is based on anatomical parameters, assuming that, in our case, photoreceptor diameter and photoreceptor length also get smaller with eye size. Accordingly, GM might have the lowest sensitivity of the three species, which is in agreement with its diurnal life style, while the intermediate and highest sensitivity could be expected in the crepuscular and nocturnal species LB and CP, respectively.

Sex differences in eye size, although statistically significant, were minimal (less than 0.5% of the total ANOVA model variance) and contributed only slightly to differences in pupil dynamics between sexes (females, having smaller eyes than males, reached half-time opening earlier than them). Females had larger dorso-ventral eye diameters than males in CP and GM, which is in accordance with the bigger size of females, but not in the crepuscular moth species LB. Pupil dynamics was controlled by the intrinsic circadian rhythms in the three tortricids. Circadian rhythmicity of eye pigments was already shown in CP (Nordstrom and Warrant 2000), but not in the other two species. Goldsmith and Bernard (1974) reviewed daily rhythms of pigment position and showed that many insects preserve it under constant illumination conditions, but in some species such as the Pyralid moth Ephestia kuehniella (Zeller) there is no circadian rhythm (Day 1941). The beetle Tenebrio molitor L. (Ro and Nilsson 1993) and the bedbug Triatoma infestans Klug (Reisenman et al. 2002) also show circadian rhythms of pigment migration under constant dark or light conditions. External temperature does not affect screening pigment movement in T. infestans (Lazzari et al. 2011), whereas in CP exceeding a threshold temperature induces a fast migration to the light-adapted position regardless of external light conditions (Nordström and Warrant 2000). This suggests that, in some species, temperature could be an important factor for triggering pupil dynamics, due to temperature differences between the day and the night, which may be used by an organism to indirectly estimate the light conditions, usually associated with a given temperature range.

Circadian rhythm observations showed that in the tortricids species opening and closing have a similar duration under constant illumination conditions and, in fact, our curve for CP is very similar to the one reported by Nordström and Warrant (2000) for this species. Ro and Nilsson (1993) also reported similar durations for pupil opening and closing of a beetle species during constant light and constant dark conditions, but there are not many references comparing this parameter under constant illumination conditions. We observed that induced pupil closing was noticeably shorter than induced pupil opening, whereas the duration of these two periods was similar in insects maintained under constant darkness. Dreisig (1981) demonstrated that the position and speed of screening pigments depend solely on the absolute illumination level, producing similar movement curves at both dawn and dusk. However, other studies—such as Bernard et al. (1984) on pyralids; Stavenga et al. (1977) on butterflies; and Nordtug (1990) and Nordtug and Melo (1992) on noctuids—reported significantly faster closing curves compared to opening ones. These discrepancies likely arise from the highly variable experimental conditions, as each study employed stimulus lights with substantially different wavelengths and intensities—the latter often differing by several orders of magnitude.

Therefore, it is plausible that an intense light stimulus accelerates the closing process to prevent photoreceptor saturation and protect the insect from temporary blindness. Alternatively, the observed asymmetry may result from the abrupt and artificially intense change in illumination during the experiment, or due to the use of monochromatic UV light for stimulation, which could have caused microvillar membrane degradation (Langer et al. 1986), or the accumulation of long-lived photo-products, such as the metarhodopsin isoform, which can potentially trigger the pupil closure until enzymatic degradation (Stavenga and Hardie 2011). We note that we deliberately used the integrated ring illuminator LEDs for both monitoring and stimulation, to demonstrate the applicability of a commercial-grade USB microscope, and to keep the set-up small, so that it could be placed within the climatized incubator. The chosen stimulation wavelength sufficiently overlaps with the sensitivity of all photoreceptor classes, which is expected in Tortricidae (Martin-Gabarella et al. 2023). Thus, we demonstrate that a simple, off-the shelf device could be used in future studies for monitoring the effects of environmental pollution in insects with superposition eyes. However, to properly measure the pupil action spectrum, an epi-illumination optical pathway with a beam splitter would be required.

To investigate interspecific differences in pupil closure dynamics, we compared the temporal response profiles of the three studied species by inducing pupil closure with UV light (Fig. 4a and c). All curves have a brief initial delay of 2–4 s before the onset of the closing process (not shown in the figure). In LB and CP, this is followed by a rapid decline in pupil brightness that decelerates progressively. In contrast, GM displays a more gradual and uniform reduction throughout the closing period. LB and CP exhibit a bimodal pattern—an inflection or plateau—that becomes more pronounced at lower temperatures. A similar curve shape was described by White et al. (1983) in the moth Manduca sexta (L.) (Lepidoptera: Sphingidae), though without considering temperature or offering a mechanistic explanation. Bernard et al. (1984) described a triphasic reflectance response in the moth Amyelois transitella (Walker) (Lepidoptera: Pyralidae), consisting of an initial slight decrease due to pigment migration from photoreceptors, a transient increase caused by pigment congregation at the distal tip of the rhabdom, and a final sustained decline as pigments accumulate from both photoreceptors and accessory pigment cells. Nordtug (1990) later reported that noctuids exhibit only the second and third phases. In our data, the closure curves appear to begin directly at the third phase, with no detectable evidence of the initial two. Thus, the bimodal features observed in LB and CP do not align with the triphasic pattern previously described. We assume that in small superposition eyes, the different phases of light adaptation overlap in time and can’t be separated without further experimental manipulation. In our study, LB had the fastest closing speed (lowest t₅₀). GM, however, had the steepest slope (b), reflecting its steady, linear-like closure. The lower b values in LB and CP likely result from the mid-curve inflection, which reduces overall steepness.

The rapid dynamics observed in LB are consistent with its ecological context, as it is active during periods characterized by abrupt changes in ambient light. Such conditions likely favour rapid pupil adjustments to maintain visual performance (Warrant and Somanathan 2022). Conversely, GM—despite having the smallest eyes and presumably the shortest pigment migration paths—displays the slowest yet most consistent closure kinetics, preserving the shape of the closing curve to a greater extent than other species. Temperature was found to prolong closure time in all three species; however, GM exhibited the strongest temperature dependence, with closure duration increasing more markedly as temperature declined. This raises the possibility that thermal sensitivity of pupil dynamics may depend by morphological traits such as eye size. After the “bump”, statistical analysis reveals no significant differences among species—unlike the closing phase, where interspecific variation was evident. This suggests that the opening response in darkness is less species-specific and less finely tuned than the closing mechanism, and is also uniformly slowed by lower temperatures. The observed effects of sex remain unexplained.

The main interspecific differences lie in the closing process. LB, a crepuscular species, closes its pupil fastest, with dynamics similar to the nocturnal CP, which is bigger and slower. In contrast, the diurnal and smallest species, GM, shows distinct closure dynamics and greater temperature sensitivity. This may relate to its smaller size, which can increase heat loss (Kozlowski et al. 2004), or to diurnal activity, which provides less natural light variation but higher photon availability. Consequently, GM may rely on alternative protective strategies, such as lower visual sensitivity or adaptations typical of diurnal species with apposition eyes (Warrant and McIntyre 1996). Our hypothesis that pupil speed is inversely proportional to eye size has been therefore disproved. The similarity in the sigmoid portion of the opening curves across species may result from all experiencing a drop below their respective light thresholds for pupil opening (Dresig 1981). However, the abrupt light-off in our experiment likely masked any interspecific threshold differences.

In summary, our results highlight clear interspecific differences in pupil dynamics, with LB exhibiting the fastest and most responsive closure, likely reflecting adaptation to environments with rapid light fluctuations. In contrast, GM displays the slowest and most thermally sensitive response, with a markedly different, more uniform closure profile. These distinct dynamics may be linked to differences in ecological niches, visual ecology, or eye morphology, suggesting divergent evolutionary pressures shaping the pupil mechanisms across species.