EVOLUTION AND COADAPTATION OF THERMOREGULATORY BEHAVIOR AND WING PIGMENTATION PATTERN IN PIERID BUTTERFLIES

This paper addresses the question of how the relationship between morphological structure and functional performance differs in related groups of organisms. I describe the relationship between a suite of phenotypic characters (behavioral posture and the pattern of wing pigmentation) and one function of these characters (thermoregulatory performance) for two groups of butterflies in the family Pieridae, focusing on how behavior and wing pattern interact to affect specific aspects of thermoregulation. Using both natural and experimentally created variation in wing-melanization patterns, I develop and test a series of predictions about the relations among thermoregulatory posture, melanization pattern, body temperature, and flight activity. Results show that increased melanization in different wing regions has positive, negative, or neutral effects in increasing body temperature of Pieris butterflies. The angle of the wings used during basking alters the relative importance of different modes of heat transfer and thereby determines the contribution of different dorsal wing regions to thermoregulation. Experimentally increased dorsal melanization can either increase or decrease the onset of flight activity and can directly alter thermoregulatory posture. For Pieris, dorsal melanization affects basking and flight, while ventral melanization primarily affects overheating. These results are used to generate a functional map relating melanization pattern to thermoregulatory performance in Pieris. Reflectance-basking posture, white background color, and melanization pattern represent coadapted characters in Pieris that interact to determine thermoregulatory performance. The differences in thermoregulatory posture and background color between pierid butterflies in the subfamilies Pierinae and Coliadinae have led to a reorganization and partial reversal of the thermoregulatory effects of melanization pattern. I suggest that this change in the physical mechanism of thermoregulatory adaption in pierids has qualitatively altered the nature of selection on wing-melanization pattern.     

Thermal ecology of Pieris butterflies (Lepidoptera: Pieridae): a new mechanism of behavioral thermoregulation

Summary I document a new mechanism for behavioral thermoregulation, not previously described in animals, called reflectance basking. This behavior, described here for Pieris butterflies, involves the use of the wings as solar reflectors that reflect solar radiation onto the body to increase body temperature. Results show that Pieris require thoracic (body) temperature. between 29° and 40° C in order to take off and fly, and achieve these elevated temperatures by basking. Diurnal patterns of population flight activity are closely correlated with patterns of body temperature during basking. Behavioral studies indicate that 1) Pieris orient to solar radiation, 2) they use thermoregulatory postures consistent with reflectance basking, and 3) they do not use the basking postures found in other Pierid butterflies (i.e., the Coliadinae). There are consistent differences in wing angles used in reflectance basking between Pieris in different subgenera. Results are discussed with respect to thermoregulation and wing color in other Pierid butterflies, and suggest that a re-evaluation of the functional significance of melanization in Pieris is needed.     

A novel method of behavioural thermoregulation in butterflies

Abstract The requirement for efficient thermoregulation has directed the coevolution of specialized morphological and behavioural traits in ectotherms. Adult butterflies exhibit three thermoregulatory mechanisms, termed dorsal, lateral and reflectance basking. In this study, we investigate a potential fourth mechanism whereby individuals perch with their wings fully spread and angled downwards such that the margins are appressed to the substrate. We find that mate‐locating male Hypolimnas bolina (L.) (Nymphalidae) adopt this posture when operational thoracic temperatures are lowest (less than approximately 34 °C). As thoracic temperature increases, males perch with wings increasingly closed and ultimately select shaded microhabitats. Using thermocouple‐implanted dead models, we show that appressed posture individuals warm faster than those adopting the conventional dorsal‐basking (horizontal wing) posture. This thermal advantage is not mitigated by shading of the outer 60–70% of the wing area, which suggests that – as with the conventional dorsal posture – only the basal wing surfaces contribute to heat gain via the absorption of solar irradiation. These investigations suggest that appression represents a novel extension of conventional dorsal basking behaviour in butterflies.     

Seasonal phenotypic plasticity of wing melanisation in the cabbage white butterfly, Pieris rapae L. (Lepidoptera: Pieridae)

Abstract.  1. Effective thermoregulation is crucial for the fitness of small flying insects. Phenotypic plasticity of the ventral hindwing of pierid butterflies is widely recognised as adaptive for effective thermoregulation. Butterflies eclosing in cooler environments have more heavily melanised wings that absorb solar radiation, thus allowing flight under these cool conditions.
2. Many pierids also exhibit phenotypic plasticity of dorsal forewing melanisation but in this case, cooler environments reduce melanisation. It has been hypothesised that this plasticity is also adaptive because it increases solar reflection from the wing surfaces onto the body in certain basking postures.
3. The degree of seasonal variation in ventral hindwing and dorsal forewing melanisation of wild-caught Pieris rapae was quantified to determine if it shows patterns of plasticity similar to that documented for other Pieris species.
4. Male wing melanisation on both wing surfaces shows the characteristic seasonal, adaptive plasticity. However, only some dorsal forewing pattern elements of females conformed to the predictions of the hypothesis of adaptive dorsal forewing melanisation. Sexual dimorphism of wing pattern plasticity may result from, and/or affect, sexual dimorphism of behaviour and physiology of these butterflies.     

DISSECTING CORRELATED CHARACTERS: ADAPTIVE ASPECTS OF PHENOTYPIC COVARIATION IN MELANIZATION PATTERN OF PIERIS BUTTERFLIES

In this study we address the question of how much of the covariation among phenotypic characters observed in natural populations is adaptive. We examine covariation among a set of phenotypic characters that describe the wing-melanization pattern of Pieris butterflies. Previous functional analyses of thermoregulatory performance allow us to predict a priori whether and how different wing melanic characters should be correlated. We quantify and analyze the variation in the wing-melanization pattern within species for a series of Pieris populations from relatively cool environments in North America and compare these results with the predictions based on our adaptive hypothesis. We consider adaptive covariation both for biogeographic variation among populations and for seasonal polyphenism (phenotypic plasticity) within populations. Our hypothesis correctly predicts many of the qualitative features of covariation in melanization among major regions of the wings, at the level of biogeographic variation among populations, for both males and females of Pieris occidentalis. When within-population variation is considered, agreement with the adaptive predictions varies considerably in different populations for both P. occidentalis and P. napi males and females. Agreement for P. napi, particularly the females, is generally poorer than for P. occidentalis. In both species, there is a consistent difference in melanization pattern between alpine and arctic sites; this difference is discussed in relation to the differences in the radiative environment between these two types of “cold” habitats. Our results suggest that some important aspects of phenotypic correlation among wing melanic characters in Pieris are adaptive. We emphasize the important distinction between covariation and co-occurrence of characters, and we discuss these results in relation to the extensive biogeographic variation and phenotypic plasticity (seasonal polyphenism) in Pieris wing-melanization patterns.     

The adaptive significance of alpine melanism in the butterfly Parnassius phoebus F. (Lepidoptera: Papilionidae)

Summary The adaptive significance of alpine melanism, the tendancy for insects to become darker with increased elevation and latitude, was investigated using the butterfly Parnassius phoebus. The effects on temperature dependent activity of five components of overall wing melanism, as well as size, were examined. The components of wing melanism examined were the transparency of the basal hindwing and distal fore-wing areas, the width of the black patch in the basal hind-wing area and the proportion of black to white scales in that area, and the proportion of the distal fore-wing covered by predominantly black scaling.The body temperature of dead specimens was correlated with air temperature, solar radiation, the width of the black patch at the base of the wings, and the proportion of black to white scales at the base of the wings. The minimum air temperatures and solar radiation levels required for initiation of flight did not vary with wing melanism of P. phoebus, in contrast to the results found for Colias butterflies by Roland (1982). However, under environmental conditions suitable for flight initiation, males with a higher proportion of black to white scales in the basal area of the hind-wing, and wider basal black patches, spent a greater proportion of time in flight at low air temperatures and low insolation. Increased basal wing melanism was also associated with increased movement of males within a population. In contrast, melanism in the distal area of the wings has no effect on activities which are dependant on body temperature. The amount of time spent feeding did not vary with differences in wing melanism. I suggest that in dorsal basking, slow-flying butterflies (Parnassius) basal wing color affects body temperature primarily during flight (rather than while basking), such that butterflies with darker wing bases cool down less rapidly because they absorb more solar radiation during flight.     

Why Small Is Beautiful: Wing Colour Is Free from Thermoregulatory Constraint in the Small Lycaenid Butterfly,Polyommatus icarus

We examined the roles of wing melanisation, weight, and basking posture in thermoregulation in Polyommatus Icarus, a phenotypically variable and protandrous member of the diverse Polyommatinae (Lycaenidae). Under controlled experimental conditions, approximating to marginal environmental conditions for activity in the field (= infrequent flight, long duration basking periods), warming rates are maximised with fully open wings and maximum body temperatures are dependent on weight. Variation in wing melanisation within and between sexes has no effect on warming rates; males and females which differ in melanisation had similar warming rates. Posture also affected cooling rates, consistent with cooling being dependent on convective heat loss. We hypothesise that for this small sized butterfly, melanisation has little or no effect on thermoregulation. This may be a factor contributing to the diversity of wing colours in the Polyommatinae. Because of the importance of size for thermoregulation in this small butterfly, requirements for attaining a suitable size to confer thermal stability in adults may also be a factor influencing larval feeding rates, development time and patterns of voltinism. Our findings indicate that commonly accepted views of the importance of melanisation, posture and size to thermoregulation, developed using medium and large sized butterflies, are not necessarily applicable to small sized butterflies.     

Behavioural thermoregulation and the relative roles of convection and radiation in a basking butterfly

Poikilothermic animals are often reliant on behavioural thermoregulation to elevate core-body temperature above the temperature of their surroundings. Butterflies are able to do this by altering body posture and location while basking, however the specific mechanisms that achieve such regulation vary among species. The role of the wings has been particularly difficult to describe, with uncertainty surrounding whether they are positioned to reduce convective heat loss or to maximise heat gained through radiation. Characterisation of the extent to which these processes affect core-body temperature will provide insights into the way in which a species׳ thermal sensitivity and morphological traits have evolved. We conducted field and laboratory measurements to assess how basking posture affects the core-body temperature of an Australian butterfly, the common brown (Heteronympha merope). We show that, with wings held open, heat lost through convection is reduced while heat gained through radiation is simultaneously maximised. These responses have been incorporated into a biophysical model that accurately predicts the core-body temperature of basking specimens in the field, providing a powerful tool to explore how climate constrains the distribution and abundance of basking butterflies.     

Females show greater changes in wing colour with latitude than males in the green‐veined white butterfly,Pieris napi (Lepidoptera: Pieridae)

In butterflies, wing colour may simultaneously be under sexual selection in the context of mating selection and natural selection in the context of thermoregulation. In the present study, we collected mated females of the green‐veined white butterfly (Pieris napi) from locations spanning 960 km of latitude across Fennoscandia, and investigated sex‐specific latitudinal wing colour variation in their offspring raised under identical conditions. We measured wing colour characteristics, including reflectance at wavelengths 300–700 nm and the degree of wing melanization. At all latitudes, females reflected more light in the short wavelengths (< 400 nm) and less in the long wavelengths (> 450 nm), and they were more melanized than males. However, female wing colour varied more with latitude than that of males. Among females, long wavelength reflectance decreased, whereas short wavelength reflectance and melanization increased, towards the north. By contrast, among males, latitudinal variation was found only in the ventral hindwing melanization. These results are consistent with the idea that the balance between natural and sexual selection acting on wing colour changes with latitude differently in males than females. The dark wing colour of females in the north may be a thermoregulatory adaptation, although males may be constrained from evolving the dark dorsal wing colour favoured by natural selection because of constant sexual selection across latitudes. © 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, ?? , ??–??.     

Thermal characterization of butterfly wings—1. Absorption in relation to different color, surface structure and basking type

1. 1. Spectral integral reflectance, transmittance and the resulting absorption of intact and descaled butterfly wings of the black-winged Pachliopta aristolochiae (Papilionidae), the white-winged Pieris brassicae (Pieridae), and the yellow-winged Gonepteryx rhamni (Pieridae) were determined between 350 and 800 nm.

2. 2. Whereas in the black forewing of the dorsal basking Pachliopta almost all incident light is absorbed nearly independent of the wavelength and thus converted into heat, the white forewing of the body basking Pieris absorbs less than 20% in the visible range of the spectrum.

3. 3. The yellow hindwing of the lateral basking Gonepteryx absorbs to a higher degree than the Pierid wing, but—due to the sparsely arranged scales—transmittance is clearly increased (40–50% between 525 and 800 nm).

4. 4. The varying thermal characteristics of the different wings with reference to the color and arrangement of the scales and the different basking strategies of the butterflies are discussed.

Predation,thermoregulation, and wing color in pierid butterflies

Summary This paper explores two hypotheses about the relationships among predation, thermoregulation, and wing color in butterflies: First, that butterflies are susceptible to predation during thermally marginal periods (e.g., cool weather) when effective thermoregulation and flight are not possible; second, that Pieris butterflies are relatively unpalatable to visual predators, supporting the idea that the white wing pigment of Pieris represents aposematic coloration. Field experiments with Pieris and Colias in 1984 and 1985 demonstrate that substantial predation may occur during the morning period before butterflies are able to actively fly. Circumstantial evidence is presented to suggest that at least some of the predation is by small, cursorial mammals. Feeding experiments in the field using Grey Jays as predators indicate that Pieris napi and P. occidentalis are less palatable than other sympatric butterflies, including confamial Colias alexandra. These and previous results suggest that Pieris are edible but less preferred as prey by birds, and that the degree of palatibility may vary among Pieris species. The relatively low palatability of these Pieris is consistent with the hypothesis that their white pigmentation represents aposematic coloration; however, the cues by which potential bird predators might discriminate against Pieris have not been established.     

Body temperatures and behavioural thermoregulation strategies of threePieris butterflies in relation to solar radiation

Behavioural thermoregulation of 3Pieris butterfly species,P. rapae, P. melete andP. napi, was examined in relation to the intensity of solar radiation. To evaluate solar radiation intensity, the temperature (T_wr) was measured with a mercury thermometer whose bulb was covered with white cloth and exposed to direct sunlight. On clear days, the diurnal air temperature was between 16 and 28°C. The T_wt varied between 18 and 45°C, while the temperature in the shade was under 25°C. When the T_wt was under 28°C, the body temperatures (T_h) of butterflies closely coincided with it. Butterflies with T_b's under 26°C were resting, while those with T_b's between 26 and 28°C were basking. When T_wr was between 28 and 40°C, the butterflies were active and their T_b's were always lower than T_wr, never exceeding 36°C, though body temperatures could be artificially elevated easily up to the level of T_wr. When T_wr exceeded 40°C, butterflies showed species-specific heat-avoiding behaviour.P. rapae, whose habitat resources exist in the sun, intercepted solar radiation by closing the wings over the body.P. melete andP. napi, however, whose main habitat resources exist in the shade, moved into the shade. Strictly speaking, it is concluded that both butterflies, in many cases, leave shaded habitats for sunny habitats to elevate their T_b rather than enter the shaded habitats for heat-avoiding.     

AERODYNAMICS,THERMOREGULATION, AND THE EVOLUTION OF INSECT WINGS: DIFFERENTIAL SCALING AND EVOLUTIONARY CHANGE

We examine several aerodynamic and thermoregulatory hypotheses about possible adaptive factors in the evolution of wings from small winglets in insects. Using physical models of Paleozoic insects in a wind tunnel, we explore the potential effects of wings for increasing gliding distance, increasing dispersal distance during parachuting, improving attitude control or stability, and elevating body temperatures during thermoregulation. The effects of body size and shape, wing length, number, and venation, and meteorological conditions are considered. Hypotheses consistent with both fixed and moveable wing articulations are examined. Short wings have no significant effects on any of the aerodynamic characteristics, relative to wingless models, while large wings do have significant effects. In contrast, short wings have large thermoregulatory effects relative to wingless models, but further increases in wing length do not significantly affect thermoregulatory performance. At any body size, there is a wing length below which there are significant thermoregulatory effects of increasing wing length, and above which there are significant aerodynamic effects of increasing wing length. The relative wing length at which this transition occurs decreases with increasing body size. These results suggest that there could be no effective selection for increasing wing length in wingless or short-winged insects in relation to increased aerodynamic capacity. Our results are consistent with the hypothesis that insect wings initially served a thermoregulatory function and were used for aerodynamic functions only at larger wing lengths and/or body sizes. Thus, we propose that thermoregulation was the primary adaptive factor in the early evolution of wings that preadapted them for the subsequent evolution of flight. Our results illustrate an evolutionary mechanism in which a purely isometric change in body size may produce a qualitative change in the function of a given structure. We propose a hypothesis in which the transition from thermoregulatory to aerodynamic function for wings involved only isometric changes in body size and argue that changes in body form were not a prerequisite for this major evolutionary change in function.     

Seasonal,environmental and anthropogenic influences on nocturnal basking in turtles and crocodiles from North-Eastern Australia

Many ectotherms bask in the sun as a behavioural mechanism to increase body temperature and facilitate metabolism, digestion or gamete production, among other functions. Such behaviours are common during the day, but some nocturnal species are also known to thermoregulate at night, in the absence of solar radiation, through shifts in body posture or microhabitat selection. Additionally, recent work has documented nocturnal basking in freshwater turtles in tropical Australia, though the purpose of the behaviour remains unknown. Here, we have built upon that work to test: 1. seasonal differences, 2. the influence of environmental factors and 3. the influence of anthropogenic development (e.g. river-front houses) on nocturnal basking behaviour. We visually surveyed transects repeatedly at night on the Ross River, Townsville, QLD, Australia from March to November 2020 and documented nocturnal basking in both freshwater turtles (Emydura macquarii krefftii) and freshwater crocodiles (Crocodylus johnstoni). For both taxa, we found significantly more nocturnal basking activity during the hotter months. Likewise, water surface temperature significantly influenced nocturnal basking in both taxa, especially when water temperatures were both high and warmer than air temperatures. We propose that nocturnal basking provides a mechanism for thermoregulatory cooling when water temperatures are high (e.g. 30°C) and above-preferred temperatures. After accounting for availability in basking habitat, both turtles and crocodiles basked more frequently on the undeveloped side of the river, suggesting avoidance of human activity or disturbance. This study is the first to document nocturnal basking activity temporally throughout the year as well as the first to identify the influences of environmental factors. Nocturnal thermoregulation has been documented in many reptiles, however, thermoregulatory cooling in tropical systems is less well-known.     

The cost of melanization: butterfly wing coloration under environmental stress

Evolutionary studies typically focus on adaptations to particular environmental conditions, thereby often ignoring the role of possible constraints. Here we focus on the case of variation in dorsal wing melanization in a satyrine butterfly Pararge aegeria. Because melanin is a complex polymer, its synthesis may be constrained if ambient conditions limit the resource budget. This hypothesis was tested by comparing melanization among butterflies that fed as larvae on host grasses experiencing different drought-stress treatments. Treatment differences were validated both at the level of the host plant (nitrogen, carbonate, and water content) and of the butterfly (life-history traits: survival, development time, and size at maturity). Melanization rate was measured as average gray value of the basal dorsal wing area. This area, close to the thorax, is known to be functionally significant for basking in order to thermoregulate. Individuals reared on drought-stressed host plants developed paler wings, and development of darker individuals was slower and less stable as estimated by their level of fluctuating asymmetry. These results provide evidence that melanin is indeed costly to synthesize, and that differences in environmental quality can induce phenotypic variation in wing melanization. Therefore, studies dealing with spatial and/or temporal patterns of variation in wing melanization should not focus on adaptive explanations alone, but rather on a cost-benefit balance under particular sets of environmental conditions.     

Butterfly wing morphology variation in the British Isles: the influence of climate, behavioural posture and the hostplant-habitat

Gradients (isophenes) in modifications of butterfly wing morphology (colour, pattern, size) to the north and west of Britain are shown to correlate closely with contemporary environmental gradients, whereas their alleged formation as infra-specific units in Devensian refugia off western Britain is unsubstantiated. A model is described which explains the transformation in phenotypes in relationship to climate, especially ambient temperatures and radiation levels. In cooler, less predictable summer conditions to the north and west, selection has favoured modifications in adult phenotypes that maintain efficiency in thermoregulation, mate advertisement and predator escape. The form that wing modifications take depends mainly on basking posture (lateral, dorsal-absorption and reflectance), which determines the allocation and interaction of functions on different wing surfaces. It is also dependent on hostplant-habitat structure, which influences thermal stability and the milieu of predators and conspecifics, and other behavioural norms (mate-locating behaviour) and biological attributes (size, robustness, speed and mode of flight, chemical defences) which affect their relationships with predators and conspecifics. The significance of Quaternary palaeoenvironments to phenetic transformations is discussed as is the relevance of the model to the development of phenotypes in arctic endemic butterflies. Differences in phenotypes of butterflies which occupy arctic and temperate montane environments are also predicted by the model.     

The rôle of butterfly wings in regulation of body temperature

Free and unnarcotized butterflies in a vertical basking position were exposed to radiation from a halogen lamp. Warming rate and equilibrium excess temperatures were recorded by means of microthermistors on the cuticle. Living, dead, and dried specimens were irradiated partly and totally. If the wings are shaded, the excess body temperature is reduced by about 30 per cent. The major portion of the heat transferred from the wing to the body originates from 15 per cent of the wing surface nearest to the body. There is no significant difference in excess thoracic temperatures of living and freshly dead specimens. After drying, the body temperature level rises about 1·4 to 2·2°C, remaining almost constant between 15°C (not radiated) and 37°C (radiated). The influence of air convection was tested with dried specimens under varying spatial orientation, keeping incident radiation constant. In an approximately horizontal position the heat arising from the wing increases to about 40 per cent by accumulation of warm air under the wing base. The ecological implications of heat supply by the wings and adaptive significance of wing pattern are discussed with respect to different modes of heat transport     

Wing pigmentation in Calopteryx damselflies: a role in thermoregulation?

Body melanization may show adaptive variation related to thermoregulation ability, and it is to be expected that the degree of melanization will change among populations or closely related species across environmental gradients of solar radiation and/or environmental temperature. Some melanized secondary sexual traits may also play a role in sexual selection, leading to interpopulation variation, which would not be predicted by thermoregulation pressures alone. We studied the relationships between the interpopulation variation in wing pigmentation level (i.e. melanized secondary sexual trait) of two closely related species of Calopteryx damselfly, and both solar radiation and maximum environmental temperature estimates. Wing pigmentation differs between these species, is gender specific and is used in species' discrimination. Only Calopteryx virgo meridionalis males showed a significant negative partial correlation between wing pigmentation degree and temperature. However, C. virgo meridionalis females showed a positive significant partial correlation between wing pigmentation degree and solar radiation. Wing pigmentation in Calopteryx xanthostoma males was not related to solar radiation or temperature. Thus, thermoregulation pressures poorly explained the observed variations in wing pigmentation between populations, although they might have an adaptive significance at the species' level. As wing pigmentation showed important latitudinal variation, several other selection pressures which might act on melanized traits are briefly discussed. © 2011 The Linnean Society of London, Biological Journal of the Linnean Society, 2011, 103 , 36–44.     

Testing the role of ecological selection on colour pattern variation in the butterfly Parnassius clodius

Colour pattern has served as an important phenotype in understanding the process of natural selection, particularly in brightly coloured and variable species like butterflies. However, different selective forces operate on aspects of colour pattern, for example by favouring warning colours in eyespots or alternatively favoring investment in thermoregulatory properties of melanin. Additionally, genetic drift influences colour phenotypes, especially in populations undergoing population size change. Here, we investigated the relative roles of genetic drift and ecological selection in generating the phenotypic diversity of the butterfly Parnassius clodius. Genome‐wide patterns of single nucleotide polymorphism data show that P. clodius forms three population clusters, which experienced a period of population expansion following the last glacial maximum and have since remained relatively stable in size. After correcting for relatedness, morphological variation is best explained by climatic predictor variables, suggesting ecological selection generates trait variability. Solar radiation and precipitation are both negatively correlated with increasing total melanin in both sexes, supporting a thermoregulatory function of melanin. Similarly, wing size traits are significantly larger in warmer habitats for both sexes, supporting a Converse Bergmann Rule pattern. Bright red coloration is negatively correlated with temperature seasonality and solar radiation in males, and weakly associated with insectivorous avian predators in univariate models, providing mixed evidence that selection is linked to warning coloration and predator avoidance. Together, these results suggest that elements of butterfly wing phenotypes respond independently to different sources of selection and that thermoregulation is an important driver of phenotypic differentiation in Parnassian butterflies.     

Thermoregulation and flight activity in territorial male graylings,Hipparchia semele (Satyridae), and large skippers,Ochlodes venata (Hesperiidae)

Some male butterflies defend specific mating sites, e.g. sandy patches (Hipparchia semele) or plants (Ochlodes venata). When perching within its territory, a male orients the body axis and tilts its wings and body in order to control the body area exposed to the sun, and thereby keeps its body temperature (T _b) as close to a preferred level as possible. In accordance with a model presented here, these behaviours can be separated into three successive phases. At low temperatures, the males maximized the heat load by exposing the maximum body area (sun-basking). This raised T _b above the temperature of a non-regulating animal by c. 3° C. At an intermediate range of temperatures, T _b was kept constant at the preferred level by means of a gradual change of body orientation and posture (graded phase). At high temperatures, the heat load was minimized by exposing the minimum body area. This lowered T _b below that of a non-regulating animal by c. 2.5° C. H. semele went through all three phases, but O. venata only reached the basking phase due to a more moderate microclimate. Three types of thermoregulation in ectothermic animals and their functions are discussed. Thermoregulation in territorial male butterflies serves to prepare the animal for efficient flight performance if another male should try to take over the territory, or a predator attacks. The males also made frequent short flights, spontaneously or elicited by other insects. Their duration was independent of temperature, and they may function as a sexual signal.