Tuatara: Volume 23, Issue 1, July 1977
Overwintering Strategies in New Zealand Insects
Overwintering Strategies in New Zealand Insects
Insects are the most numerous of all living animal species. The reasons for their success are many and varied but can be directly related to adaptations found not only in the adult, but also in the various life cycle stages. In particular, those adaptations which enable survival during a period of severe environmental adversity are of considerable importance as the survival of the species depends to a large extent upon the success of these particular adaptations. Foremost among these are the overwintering adaptations which enable survival during the rigours of winter. For a large majority of species these include not only sub-optimal temperatures, frosts, and snow, but often the absence of the food source or host, so that normal growth, development and reproductive activity are inhibited. Many species of animals and plants faced with such environmental problems during winter have evolved a variety of survival mechanisms loosely referred to as dormancies, all of which involve either the slowing down or even the complete cessation of normal metabolic and behavioural activity. For such insects, the only alternative to evolving an overwintering survival stage (or dormancy mechanism) is migration out of the region for the duration of the winter, and in fact many insect species rely on this strategy instead of a period of dormancy.
Specific overwintering mechanisms such as diapause in insects, or its equivalent in plants — deciduousness — are presumed to have evolved during the Pliocene-Pleistocene Ice Ages in the Northern Hemisphere, when the selection pressures of severe climatic adversity brought about either the extinction of many species, or the development of overwintering survival mechanisms such as diapause. In the Southern Hemisphere, however, and particularly in New Zealand, the severity of this period is thought to have been sufficiently modified by our maritime climate so that relatively few insect or plant species were eliminated. As a result, we not only have a very high percentage of endemic plant and animal species, but there is also considered to page 2 be a very low incidence of diapause (and deciduousness) among our native fauna and flora (Dumbleton 1967). Instead, the occurrence and improvement of adaptations other than diapause, which conferred increasing cold hardiness and which were sufficient to enable survival during our relatively mild winters no doubt became widely established among surviving insect and plant species.
However, diapause is not solely an adaptive outcome of severe climatic selection pressures; it may also evolve in response to a number of other limiting factors in the life cycle such as periodic absences of food, or water (or even an excess of water in certain tropical latitudes). In some cases, in which the life cycle of the individual is closely related to that of a host species (animal or plant) a diapause stage may result which synchronises the insect with the seasonal cycle of the host species. According to the ‘Dumbleton hypothesis’, examples of climatically induced diapause in New Zealand insects should be rare or even absent; and any example of a diapause stage can therefore be presumed to have two probable sources: either it evolved in response to environmental pressures other than climate (such as have already been referred to); or the mechanism was evolved elsewhere and imported into this country along with the insect in question. Problems associated with the latter can be easily solved in those cases where the recent arrival of the insect has been observed and recorded. Difficulties can occur, however, in those cases in which the insects' antiquity in this country cannot be accurately determined.
The major objectives involved in overwintering studies are thus three-fold. The first involves detailed field studies, noting in particular how many, and which particular stage(s) are present during the winter months and their state of activity. This may range from normal daily activity to various degrees of inactivity or dormancy. Observations on adult activity during winter also involve examination of the state of the reproductive organs, as imaginal reproductive diapause is known in many insects. Identification of any possible environmental limiting factors operating in the field during the winter will assist the second objective which is a laboratory study of the factors that induce and terminate the state of dormancy. Finally interpretations of the findings from both field and laboratory observations have to be based upon a clear understanding of the meaning and definition of the various types of dormancy states commonly referred to in the literature. Unfortunately several of these terms are often used synonomously, i.e. hibernation, diapause and dormancy may all be used with reference to the same overwintering mechanism. Generally, the term dormancy can be used to encompass all types and degrees of growth arrest, which range from briefly and directly induced states of cold torpor to progressively intensifying states of inactivity including quiescence, through to the advanced and physiologically complex adaptation called diapause.page 3
Many attempts have been made to define these terms and in particular to identify the environmental factors which are involved in their induction (or onset). Recent literature on the subject (e.g. Way 1962; Lees 1962; Danilevskii 1964; Norris 1970; Mansingh 1971) reveal a consensus of opinion that two basic types of dormancy can be distinguished: quiescence and diapause. The former can be defined as cessation of activity for indeterminate lengths of time, in direct response to adverse environmental conditions (commonly temperature) with a correspondingly direct return to normal activity upon termination of the adversity. Quiescence can be found in any or in all stages of the life cycle which are subjected to the adverse conditions. By contrast, diapause is known to occur in only one species specific stage of the life cycle, in almost all known diapausing insects (Beck 1968). This stage, therefore, is the only one which is present and able to survive the period of seasonal adversity. Secondly, it can be distinguished from quiescence by being induced in advance of (i.e. in anticipation of) the seasonal adversity, and then continuing to proceed for a definite period regardless of changes in the external environment. Only when this period of ‘diapause development’ is completed will the diapause be terminated. Observations in the field on the anticipatory nature of diapause and of its ‘timing’ in order to synchronise a particular life cycle stage with the seasonal adversity, have led to the conclusion that an innate ‘clock’ mechanism is involved in this process. While temperature has long been known to influence both the incidence and intensity of diapause, its actual timing has been shown to be based on seasonal information received from the daily photoperiod. This provides an extremely stable and accurate environmental ‘time cue’ as to the time of year, and thus enables the insect to anticipate the onset of the unfavourable season and hence enter diapause before its arrival.
An interesting aspect of this discovery is that this reliance on photoperiod by diapausing insects has led to the observation that this photoperiodic process is latitudinally related, and in fact many examples of geographical ‘clines’ in a population of the same diapausing species are known to occur. This is particularly true in facultatively diapausing species in the Northern Hemisphere which can show a gradation in the percentage of diapausing individuals in any one generation ranging from virtually nil at low latitudes, through an increasing percentage with increasing latitude until at very high latitudes an apparently obligate diapause of 100% in every generation is manifested.
A survey of the literature concerning overwintering studies in New Zealand insects highlights most of the aspects already mentioned, i.e. problems concerning the introduced or endemic status of some insects; problems in the lack of parallel field and laboratory investigations into the nature of the overwintering stage(s); and problems in the interpretation of the type of dormancy stage. A good variety page 4 of studies is involved including representatives from different insect orders, and from geographical locations ranging from 36° S to 48° S. Despite their relatively small number they provide a fairly comprehensive test of the ‘Dumbleton hypothesis’, and enable some conclusions to be drawn.
According to Wilkinson (1964) the tiger-moth, Metacrias strategica, in Otago (47° S) appears to overwinter in the larval and pupal stages, the adults having only a very brief life span (approximately 21 days) confined to the summer months. Early instar larval feeding during warm periods then gives way in the cold winter months of June to August to an inactive ‘quiescent’ last larval instar, which overwinters until the approach of warmer spring weather. Because of individual variations in the duration of the larval instars, pupae too may overwinter, and Wilkinson noted that the duration of the pupal instars increased with decreasing winter monthly temperatures. Although no experimental investigations were carried out to determine the nature of the stimuli(s) involved in the onset of this ‘quiescence’, the absence of a total arrest in the life cycle prior to the onset of winter, and of a single specific overwintering stage, along with the observations concerning the relationship of the length of the pupal stage to temperature, are all strongly suggestive of a temperature induced quiescence rather than a diapause. (Unfortunately Dumbleton (1967) cited this as one of the two known examples of a winter diapause in New Zealand.) Spitzer (1970) looked specifically for evidence of either winter quiescence or diapause in the Manawatu (40° S) as indicated by the presence or absence, respectively, of all or only one stage of the life cycle during the winter months. Of at least 23 species of Noctuidae examined, in most the adults, although not numerous, could be found during the winter and in all cases they were reproductively mature. The larval and pupal stages during winter remained active but growth was slowed, and pupal quiescence was ‘probable’ in at least one case. Spitzer concludes that ‘at low altitudes in the North Island diapause does not occur’ (either in the adults, as reproductive diapause) or ‘even in the immature stages although … development may be somewhat retarded’ — this state of growth retardation he terms ‘quiescence’. Because of the low altitude and more northerly latitude of this study, Horak-Kaenel's (1969/70) investigation into the overwintering of Proteodes carnifex is of particular interest as it was carried out on an alpine species at high altitudes and latitudes (41° S at the St. Arnaud Range in Nelson, and 45° S at Te Anau). In the Northern Hemisphere at such latitudes and in a known univoltine species, diapause could be expected to occur. In fact, however, although P. carnifex constructed a protective winter hibernaculum, it only remained inactive within this cocoon for short spells during heavy snowfalls when feeding was impossible. At every favourable opportunity it emerged to feed on Nothofagus cliffortiodes. It appears that despite the high latitude page 5 and altitude, the winter temperatures are not sufficiently severe, at least for prolonged periods, to completely inhibit slow growth and development throughout the winter.
The 10 species of Lepidopteran Crambini studied by Gaskin (1975) included both uni- and bi-voltine species, and even geographical races within the one species, which vary according to latitude as to the number of generations per year. The life cycle of Orocrambus simplex, a uni-voltine species inhabiting sub-alpine tussock grasslands, is illustrative of the general overwintering trends in this group. The adults and eggs are both present during the summer months while the first four larval feeding stages can be found during autumn. The sixth larval instar appears to be the overwintering stage, remaining inactive (and non-feeding) from April to October. No adults of any of the 10 species were recorded during August and September. In O. flexuosellus (on which the greatest number of observations of each stage were made) there is a definite trend within each particular stage of the life cycle for an increase in its duration to occur with the approach of winter: e.g. January egg duration (9-10 days); July egg duration (24-29 days); larval stages 1-4 in March (14-30 days); in May (25-50 days). Without conclusive experimental evidence it is not possible to state categorically that the last larval overwintering stage represents a quiescent rather than a diapausing dormancy stage. However, the evidence of progressive growth retardation in all stages with the onset of lower winter temperatures, along with the overlap in several species of both late larval and pupal stages during winter months are again suggestive of a quiescent rather than a winter diapause mechanism.
It is important at this point to emphasise that all insects so far cited are not only endemic but are phytophagous, on evergreen host plants. The existence and the importance of such a year-round food supply for so many of our endemic insects is realised only from comparison with studies on insects lacking such a continuous food supply; in which case this can possible become the limiting factor promoting the evolution of a diapause overwintering mechanism. In the remaining studies, this and other potential diapause-producing factors are considered.
The Pompilidae (‘spider-wasps’) studied by Harris (1974) are a case in point. Unfortunately although a large number of these species were studied in considerable detail, only one life cycle is described. Priocnemis nitida is a uni-voltine insect present in the adult stage only during the summer months when, because of its strongly positive heliotropism, it is active only in full sunlight. The eggs are laid in summer (e.g. February) and along with a single spider to which each egg is attached, they are placed individually into a protective nest. Within two weeks the larva passes through its feeding stages, completely consuming the spider food supply before spinning the cocoon of the last larval (fifth) instar. Inside this, the page 6 prepupae spends nearly eight months in a completely inactive state before pupating (14-18 days duration) in late spring and emerging as an adult in early summer (e.g. November). Although Harris reports that the prepupal ‘diapause’ could be terminated by chilling at 0° C for three weeks, no other experimental investigations into stimuli affecting the timing and induction of this overwintering stage were carried out. However, there is a strong likelihood that in this insect the overwintering prepupae is in a true diapause, and that this has evolved as an adaptation to the absence of food during the autumn and winter. This absence is probably due not so much to the lack of spiders, but rather to the inability of the adult wasp to collect spiders during the autumn and winter when the heating effect from insolation is too low. Another example of a possible diapause mechanism having been evolved in a New Zealand insect is that of the weevil Praeolepra unifomis. This insect displays a close relationship with the seasonal phenology of its host plant Coprosma lucida. The eggs are laid in March in the axil of the green fruit and stem, and the developing larvae burrows into the fleshy mesocarp where the first instar feeds externally, while the second instar tunnels in through the hard endocarp and continues to feed inside the seed case on the carpels, and inner walls of the seed coat. The prepupa then overwinters inside the maturing fruit until emergence 16 to 18 months later (i.e. in July and September respectively of the following winter). (B. M. May, pers. comm.)
Although details of the degree of host specificity and the adult foodplant of this weevil are not certain, it is possible that the winter emergence of the adults is related to the winter flowering of the Coprosma plant. The interesting feature of this life cycle is that it illustrates another environmental factor which is known to contribute to the evolution of a winter diapause mechanism, and that is where the life cycle of the insect is closely adapted to the phenology of the host plant for food and/or shelter. Danilevskii (1964) cites as a ‘typical example’ of this situation the apple blossom weevil whose entire development occurs in the buds of the apple blossom. The adult beetle after emergence enters a diapause state for 10 to 11 months until the following spring flowering of the apple tree. It is quite possible that Praeolepra uniformis may have evolved a prepupal diapause for similar reasons; as Danilevskii (1964) notes, ‘such cycles are characteristic of many carpophagous insects.’ Alternatively, the lengthy period inside the Coprosma fruit may be spent in a combined larval feeding and prepupal quiescence.
The life cycle of Pericoptus truncatus, an endemic scarab beetle (Dale, 1963) provides another example which poses interesting ecological problems concerning the existence of a diapause or quiescent stage in the life cycle. This beetle spends its entire life, including mating and reproduction (except for an adult dispersal flight shortly before death), beneath the ground on sandy beaches. page 7 It depends for its food supply on the nutritionally rich but environmentally unstable driftwood zone, as well as on the more stable but nutritionally poorer marram grass zone. Although it thus has a year-round food supply, in both habitats dessication is a problem. Both an egg diapause and a prepupal diapause are suggested in Date's study but the occurrence of two distinct diapausing stages in a life cycle is rare (Beck, 1968). It is possible, however, that in such a precarious and dessication-prone habitat, moisture content may act as a (micro) climatic limiting factor on egg and/or on prepupal development, necessitating a diapause stage in at least some individuals. The resistance of diapause stages to dessication is well known (Danilevskii, 1964) and would thus ensure the survival of some prepupae over the summer dry period, until they were able to pupate the following year. Such a prepupal facultative diapause would result in both the two and three year cycles recorded by Dale. Yet until the environmental factors are known which enable the actual timing and induction of the diapause, the alternative possibility of a state of quiescence, at least in the egg, cannot be discounted.
The remaining overwintering studies worthy of note concern those insects which (unlike the endemic species previously mentioned) are of more doubtful, or of recently introduced origin. In such insects it is not surprising to encounter instances of diapausing mechanisms; the problem lies in determining whether they evolved in this country or were introduced with the arrival of the insect.
Hardwick (1965) reports a winter pupal diapause in Helicoverpa armigera conferta (of the corn earworm complex), varying from 19% at Rotorua (39° S) to 82% in the Nelson (41° S) population. The published status of this insect as a South Pacific sub-species is indicative of its lengthy presence in the area, but the origin of its diapause mechanism was most probably imported along with its adventitious spread into the Pacific. (J. S. Dugdale, pers. comm.)
Similarly, the codling moth, Cydia pomonella, is of known recent introduction, and it also exhibits a diapause in which the percentage variations in incidence per generation (and hence the occurrence of uni or bi-voltinism) are correlated with different latitudinal locations. (H. C. Waring, pers. comm.)
The cricket species in New Zealand represent the most interesting and problematical of overwintering studies, and indeed one species, Teleogryllus commodus, is the second of the two examples of winter ‘diapause’ cited by Dumbleton (1967). However, two major problems present themselves when trying to determine the exact nature of the overwintering mechanism in the Gryllidae; one is the uncertainty of origin (particularly of Teleogryllus commodus) and the other is (once again) the lack of experimental evidence concerning the environmental stimuli responsible for the timing and induction of the ‘diapause’. It is commonly observable that the uni-voltine life cycle of T. commodus in the field, even in Auckland (37° S) page 8 involves an egg diapause of approximately 98% (R. L. Hill, pers. comm.) and that adults are completely absent during winter. In contrast to this high diapause incidence are studies on native Nemobiinae crickets; two studies in particular illustrate a geographical (and hence latitudinal) gradient in diapause occurrence. McIntyre (1969) has looked at differences between a Christchurch (43° S) and a Kaikoura (42.° 5 S) population of Pteronemobius sp. and concluded that they represent distinct ecological races, with the Christchurch egg diapause being ‘obligate’ and of ‘high intensity’ while in the Kaikoura population it was of ‘lower intensity’ and ‘facultative’. Comments on diapause formed only a minor part of McIntyre's investigation and were based on insufficient evidence to verify this conclusion. However, her work, together with evidence from cultures of Pteronemobius bigelowi and P. nigrovus at Victoria University (G. W. Gibbs, pers. comm.) indicate that an egg diapause mechanism is present in these crickets but that its control and significance are poorly understood at present. It must be facultative in both populations studied by McIntyre, the differences in incidence and intensity merely representing variations in the broad spectrum possible in facultative species.
Parkes (1972) also examined two populations of Pteronemobius sp. at Hamilton, and the more coastal Kawhia (38° S). Here field studies showed that adults were present throughout the winter months in the Hamilton population, while at Kawhia ‘both adults and nymphs could be found during most of the year’. Parkes concludes that in the Hamilton population a partial second generation may occur in favourable years while at Kawhia, ‘continuous year-round breeding probably occurs’ — which rules out any diapause mechanism at least in the latter population. By contrast T. commodus populations at both locations had only a single generation/year, the adults dying out in autumn and only the eggs overwintering in diapause.
Although Bigelow (1964) points out that certain morphological differences exist between the New Zealand and Australian populations of T. commodus, suggesting its long standing in this country, Parkes (1972) comments that both Gryllinae species (T. commodus and Modicogryllus tepidus) are of introduced origin.
In conclusion it appears that to date the investigations into overwintering mechanisms of endemic New Zealand insects do not indicate the existence of a diapause state as distinct from other forms of dormancy, such as quiescence. However, possible exceptions may be the zoophagous ‘spider wasps’, the carpophagous weevil Praeolepra uniformis, and the sand-inhabiting scarab Pericoptus truncatus. It is interesting to note that each of these insects introduces ecological factors other than that of severe macro-climatic adversities which more commonly promote the evolution of a diapause mechanism.
It thus appears that Dumbleton's (1967) thesis relating the lack of a sufficiently severe Pliocene-Pleistocene with the presence of a high page 9 percentage of non-diapausing (or non-deciduous) endemic fauna and flora, can largely be confirmed by our present knowledge of overwintering studies in New Zealand insects. The emphasis of this suggestion is, however, on phytophagous insects with evergreen host plants. If further investigations indicate the existence of a winter diapause in an endemic insect, then factors other than climatic adversities, such as the type of food source or the phenology of the host plant or animal, or even more subtle micro-climatic factors may provide the clue to the selection pressures necessitating the adoption of this overwintering mechanism.
Beck, S. D., 1968: Insect photoperiodism. Academic Press; N.Y., Lond. Bigelow, R. S., 1964: Note on the black field cricket in New Zealand. N.Z. Ent. 3 (3): 9-10.
Dale, P. S., 1963: Ecology, Life History and Redescription of Pericoptus truncatus (Fabricius). Trans. Roy. Soc. N.Z. (Zool.) 3 (3): 17-32.
Danilevskii, A. S., 1964: Photoperiodism and seasonal development of Insects. Oliver and Boyd: Edin. and Lond.
Dumbleton L. J., 1967: Winter dormancy in New Zealand biota and its Paleoclimatic implications. N.Z. Jl. Bot. 5: 211-222.
Gaskin, D. E., 1975: Information on the life cycles of some New Zealand Crambini (Lepidoptera: Pyralidae: Crambinae). N.Z. Jour. Zool. 2 (3): 365-376.
Hardwick, D. F., 1965: The corn earworm complex. Mem. Ent. Soc. Can. 40.
Harris, A. C., 1974: A revision of the New Zealand Pompilidae. Unpublished M.Sc. thesis, Victoria University of Wellington.
Horak-Kaenel, M., 1970: Systematic and ecological study of Lepidoptera on Nothofagus cliffortioides in New Zealand. Unpublished thesis produced for E.T.H. Entomological Institute, Switzerland.
McIntyre, M. E., 1969: A comparison of two geographical populations of small black field crickets. Part 1: An assessment of variations based on diapause characters. Unpublished M.Sc. project, University of Canterbury.
Mansingh, A., 1971: Physiological classification of dormancies in insects. Can. Ent. 103, 983-1009.
Norris, K. R., 1970: In: The Insects of Australia, pp. 107-109. C.S.I.R.O., Melbourne University Press.
Parkes, H. D., 1972: Aspects of the biology of Pteronemobius sp. (Gryllidae; Nemobiinae). Unpublished M.Phil. thesis, Waikato University.
Spitzer, K., 1970: Observations on adult activity of Noctuidae (Lepidoptera) during winter in the Manawatu, New Zealand. N.Z. Journ. Sci. 13 (2): 185-190.
Way, M. J., 1962: Definition of diapause. Ann. Appl. Biol. 50: 595-596.
Wilkinson, L. L., 1964: Notes on Metacrias strategica (Meyrick). Trans. Roy. Soc. N.Z. 4 (14): 192-200.