Role of water and moisture in diapause development ( A review )

While observational studies led to the assumption that water or moisture (W/M) is a prerequisite for diapause development, the experimental research indicates rather the opposite: usually W/M is needed as late as for the post-diapause resump­ tion o f morphogenesis. Recent examples for this type o f regulation o f dormancy are given: Eggs o f the tettigoniid Stictophaula armata, eggs o f the grasshopper Oedaleus senegalensis, adults o f the bruchid Bruchidius atrolineatus, adults o f the endomychid Stenotarsus subtilis (= S. rotundus). In the late diapause of the noctuid Busseola fusca and in eggs o f the chrysomelid Homichloda barkeri, moisture is assumed to be the diapause terminating cue. Fall in temperature is assumed decisive for termination o f pupal dia­ pause in the saturniid Schausiella santarosensis, although the effect o f intense rain after a long dry period has not yet been excluded. Effects o f intense changes in environmental conditions and o f gradual decrease in diapause intensity with time have often been neglected.


INTRODUCTION
Most studies of seasonal ecology of insects focus on photoperiod and temperature as principal regulating cues (Tauber et al., 1986;Danks, 1987).Although particularly photoperiod has an essential role due to the high reli ability of its precise repetition year after year, moisture should not be underestimated -as was rightly indicated by Tauber et al. (1998).Detailed experimental protocols for the study of moisture were suggested in the mentioned forum paper, but not many such experiments have been performed since then.A serious survey can nevertheless be based on several detailed and precise experimental studies and a quite long series of observations.
There is an obvious causal relation between the regrowth of vegetation, produced by rain after a period of drought, and the resumption of the active phase of insect life.The principal function of diapause is to assure syn chronization of these two phenomena.These facts repre sent the basis for a widely accepted assumption that moisture or water (M/W) is the factor regulating the development of such diapauses that are in some way related to alternation of periods of drought and high humidity.It appears that the impact of moisture on dia pause development is not so simple.It is, in fact, so com plex that often the interpretation of findings remains ambiguous.Andrewartha's (1952) remark, that it is difficult to unravel the true causal relationships between water and diapause, is certainly true.According to his opinion "the absorption of water is rarely if ever the primary stimulus which initiates diapause development".

NEGATIVE VS. POSITIVE INFLUENCE OF W ATER/RAIN
Increasing relative humidity and precipitation was recorded as a seasonal signal for annual vertical migration of arthropods to trees in flooded plains of central Ama zonia (Adis & Junk, 2002).The ascent to tree trunks occurs 3-4 months before forest inundation.The repro duction of arthropods (that are usually univoltine), such as tiger beetles (Pentacomia egregia), is postponed to the terrestrial phase after inundation (Table 1).
Not only flooding, but also very high humidity can endanger the development of descendants and delay oviposition to the post-rain season (Table 1, below).The females of a Mexican bruchid beetle, Acanthoscelides obtectus, spend the wet period between April and  November in reproductive diapause in rolled dead leaves hanging from plants, and postpone the oviposition (mid-November to late December), i.e. to the period when pre cipitation is below 50 mm per month (Hodek et al., 1981).
The larvae developing inside bean seeds may thus be saved from the highly probable dying in decaying beans.
A contrasting example of synchronization of reproduction with high humidity can be mentioned.The reproduction in the tropical endomychid beetle, Stenotarsus subtilis (= S. rotundus; Roubik & Skelley, 2001) occurs in May, i.e. the first month of the wet season (Tanaka et al., 1987a, b;Tanaka, 2000; Table 1, above; see also the last chapter).It feeds most probably on fungi like the other beetles from this family, but a direct evidence is still missing.The dif ference in food requirements of larvae could explain the contrasting timing of reproduction in relation to rainy sea sons in the above two species of beetles.

EFFECT OF PHOTOPERIOD AND TEM PERATURE IN SUM M ER DIAPAUSE
Before proceeding to cases where a more or less active role of M/W in diapause termination was recorded (or at least supposed) it should be stressed that in many sum mer diapauses no decisive effect was found.Diapause ter mination has often been reported to be governed rather by decrease in photophase and temperature (Masaki, 1980;Tauber et al., 1986;Danks, 1987;Hodek, 1996Hodek, , 2002)).
A recent example of summer diapause completed by low temperature is Delia antiqua (Ishikawa et al., 2000;Fig. 1).Sometimes, however, diapause in tropical insects is terminated by several days exposure to high tempera ture (25°C), as was reported in parasites of Drosophila (Carton & Claret, 1982) or in sarcophagids (Denlinger, 1986).These relations were already discussed in the classic reviews of Masaki (1980) and Denlinger (1986).In Table 2 some recent examples are given.It should, however, be indicated that in most cases the effect of M/W was not excluded.
In a detailed study on Stenotarsus subtilis (=S.rotun dus) (Tanaka et al., l987 a, b;Tanaka, 2000; see the last chapter), where the resumption of activity and dispersal from dormancy sites were recorded at the beginning of rainy period, the role of photoperiod was not only observed in the field, but also proved by experiments as decisive.An experimental proof that the photoperiod and not moisture is the decisive factor for diapause develop ment was given for two chrysomelids, Chrysolina hyperici and C. quadrigemina (Schops et al., 1996;Fig. 2).It also was proved by experiments in prepupae of the Mediterranean arctiid moth Cymbalophora pudica (Kostal & Hodek, 1997) that photoperiod is the decisive factor for diapause development.In this moth the action of M/W and temperature was excluded.

ROLE OF M OIST AND DRY PERIODS IN DIAPAUSE DEVELOPM ENT
Tropical borers have been the most popular example of diapause terminated by rains.
In spite of several studies, the role of water in termina tion of larval diapause in several species of African borers remained only partially understood until the series of clever experiments on the noctuid Busseola fusca per formed in Kenya by Okuda (1988Okuda ( , 1990Okuda ( , 1991)).Larvae of this sorghum stem borer spend 6 to 8 months in diapause in dry stalks and pupate after the onset of rains in spring.Okuda (1991) found that water stimulates pupation only in late diapause and he proved by experiments that a pre vious exposure to drought is a prerequisite for the stimu latory effect of wetting (Figs 3 & 4).Diapause develop ment evidently proceeds at dry conditions, while wet con ditions are then needed for the resumption of development and pupation.Okuda (1990) was also able to find out that the stimulation is produced rather by a continuous contact with moist surroundings than through drinking.When diapausing larvae were provided with water by means of a pipet no pupation was recordedwhen they did not become wet (Okuda, 1990).The drinking larvae did not pupate although their constant weight indicates that they had higher water content than the wet larvae (Fig. 5).
Similar positive effect of a dry period was found in Thailand in diapausing eggs of a tettigoniid Stictophaula  (Okuda, 1988) armata by Ingrisch (1996).While the apparently post diapause "dry" eggs (after 6 months = 180 day storage) hatched early after wetting (the peak of hatching was after about 4 wk), the constantly moist eggs had a bimodal hatching, with only about 30% of (probably non diapause) eggs hatched within 5 weeks.Hatching of the other 30% was delayed until about 200-300 days.The pre-hatch period of the second mode of constantly moist eggs was thus longer than in the case of eggs that spent  (Okuda, 1991) about 180 days in low RH (Fig. 6).Moreover, in another experiment, Ingrisch (1996) achieved the same effect after only 2-3 months storage in 60% RH, as after 6 months.
The interpretation of these similar results by the two authors is not the same.While Okuda (1991) suggested that a tachytelic moist completion of diapause followed after a period of horotelic processes of diapause, Ingrisch (1996) assumed that a post-diapause hygric quiescence was terminated by water.He also suggested that "drought as a proximate factor for the elimination of diapause has been rarely discussed" and "could be more commonly  (Ingrisch, 1996) detected in areas with a pronounced dry season, if it was looked for".
The importance of dry periods -although in a different sense -was clearly shown also in egg diapause of the African chrysomelid Homichloda barkeri introduced to Australia (Nahrung & Merritt, 1999).Significantly higher hatch rate was recorded under the regime of alternated wetting and drying than under constant moisture.A similar proportion of larvae hatched from eggs wetted daily and from those wetted every 23rd day.In the latter group hatching lasted more than six times longer (Fig. 7).It was evident that each subsequent wetting stimulated development of a further proportion of embryos.The con stant period of 12-14 days between wetting and eclosion indicates that at each wetting, embryos within the devel oping cohort begun development at the same stage of embryogenesis.In Homichloda, diapause development was thus completed within only 2 to 4 weeks through daily wetting and the delay in the cohort with less fre quent wetting events was due only to longer intervals between wetting events.For the same total hatching rate, however, much less wetting events were needed (9) com pared to the daily wetting regime (27).Two factors had a complementary role: Wetting events and the passage of time.It may be suggested that horotelic processes of dia pause development alternated with tachytelic activation events.The authors consider wetting events per se responsible for "breaking" diapause.They assume thus that diapause of Homichloda is in this aspect different from that of Diabrotica virgifera (Krysan, 1978) where moisture triggers development after a previous sponta neous diapause termination.
There are many cases similar to Diabrotica, in which the completed diapause development is followed by a quiescence due to low moisture.The post-diapause mor phogenesis is not resumed until water content is replen ished to/above a critical level.Thus, e.g., the larvae of the wheat blossom midges, Contarinia tritici and Sitodiplosis mosellana need high soil moisture for pupation after the temperature regulated diapause development (Basedow, 1977).
Another example of such a sequence of low moisture quiescence after diapause completion was reported as one of the possible developmental pathways in eggs of the Australian plague locust, Chortoicetes terminifera (Acrididae).The development of the so called "second quiescent stage", that is morphologically indistinguish able from the previous diapause stage, cannot start without replenishing about 10% of moisture lost during diapause (Wardhaugh, 1988).Diapause development lasts 30-60 days at 10-15°C and hatching occurs in warmth in spring.Also Gehrken & Doumbia (1996) concluded from detailed observations in Mali that the eggs of the grass hopper Oedaleus senegalensis complete diapause devel opment after 7-8 months and remain thereafter dormant in "a state of postdiapause quiescence until the return of moist conditions".
In addition to other environmental cues (mainly short days) also high humidity (70-80 %) was needed for ter mination of reproductive diapause in the Sahelian bruchid Bruchidius atrolineatus (Lenga et al., 1993;Glitho et al., 1996).Vitellogenesis was considered here as a criterion of diapause termination.Most probably the dormancy scenario could be similar in the bruchid adults and in the cases mentioned earlier, eggs of orthopterans or larvae of the stem borer: Also here diapause development was completed during the long dry season and the cues of the rainy season (high humidity and pollen of host plants) switched on the post-diapause development.
Action of several environmental factors is apparently combined in the already classical model of tropical dia pause, the adult dormancy of the endomychid beetle Stenotarsus subtilis (= S. rotundus) in Panama (Tanaka, 2000).This beetle spends a long, up to 10 months (6 months of the wet season and 4 months of the dry season) dormancy within large aggregations (Fig. 8).Below -photoperiod and humidity conditions at the Panaman site, with the assumed sensitivity to photoperiod.(Tanaka et al.,1987a, b;Tanaka, 2000) muscles.Similar to many species, it is difficult in S. subtilis (=S.rotundus) to pin-point when diapause develop ment is completed and the post-diapause quiescence begins.Mating and dispersal coincide with the onset of rains.Usually a short early period of rain during the rainy season is sufficient for triggering dispersal, but in some years dispersal may be bimodal (Tanaka, 2000).Probably the beetles with still less developed flight muscles "wait" for the next rain.Effects of exposures of 2 months to con trasting relative humidities of 100 vs. 75% confirmed the existence of stimulation by the higher humidity on early development of several organs: Ovaries, male accessory glands and flight muscles (Denlinger, 1994;Tanaka, 2000).It seems that moisture stimulates post-diapause development in only those beetles that had already com pleted diapause development before the start of experi mental humidity exposures in mid-February.Another sample was collected more than 2 months earlier, at the end of wet season, in early December (Tanaka et al., 1987a, b) and exposed for the same period of 2 months to the identical high humidity.A comparison of the develop ment of organs between these two samples indicates that an important progress of diapause development occurred already during 10 weeks of dry season.
The humidity exposures did not affect the size of testes in S. subtilis (=S.rotundus).This parameter is not impor tant for monitoring its diapause, because reproductive diapause in coleopteran males involves the accessory glands rather than testes, as was shown in coccinellids (Hodek & Ceryngier, 2000).
In a detailed study on seasonality of Lepidoptera in a Costa Rican tropical dry forest, Janzen (1987) is suspi cious to the widely accepted assumption that rains are the cue for pupal eclosion.He suggests on the contrary that the cue is the "temperature and/or its changes, rather than the actual moisture brought by rainfall".Thus, e.g., an experimental exposure to moisture does not re-initiate development in the pupae of the saturniid Schausiella santarosensis.The about 2-4 months old pupae remain dormant also through the "little wet season" (before the subsequent 6 month dry season) but the pupae eclose at the beginning of the second wet season, when they are about 8 months old.A dramatic drop in temperature at the beginning of the second rainy season occurs about a week before "really soaking" rains arrive (Fig. 9).Also several other saturniids were observed to eclose after a series of exceptionally cool days with nocturnal lows of 17-18°C (Janzen, 1987).Janzen's assumption sounds quite plausi ble.However, two other possibilities remain to be excluded.(1) In the 2-4 month old pupae the horotelic processes of diapause development are, most probably, not yet progressed enough, i.e. diapause development is not completed.(2) The effect of contrast in physical con ditions (heavy rains after the long dry period) can play an important role, similar to the effect of contrasts in photo period or temperature, suggested in temperate diapauses (Hodek, 1978(Hodek, , 1983(Hodek, , pp. 19-24, 2002;;Zaslavsky, 1978).

CONCLUSION
It can be concluded from detailed experimental studies that the assumption of positive direct effects of W/M on the processes of diapause development is mostly erro neous -when based merely on observations.On the con trary, diapause development is often completed spontaneously during a dry period and is followed by a hygric quiescence.It is during this second part of dor mancy, in early post-diapause, when the loss of water is replenished and the resumption of morphogenic develop ment is enabled (Tables 3 & 4).In discussions on the role of various factors in comple tion of diapause, three factors are mostly invoked, photo period, temperature and water or moisture (W/M), while two further factors are almost always neglected, a neces sary passage of time and the effect of contrast.In many insect species the horotelic processes of diapause devel opment cannot be substituted by activation too early; diapause development simply needs to be completed  (Janzen, 1987).It seems that a previous period of drought predisposes the insects to the effect of the increase in moisture.
The difficulty to discriminate easily between these two periods of dormancy leads some authors to avoid the problem by using the ambiguous term "termination" of diapause (to include both diapause development and post diapause development) against which Tauber et al. (1986Tauber et al. ( , 1998) ) and Danks (1987) rightly warned.When we misuse this term, we elude the challenge to understand how the two certainly different phases of dormancy -diapause vs. post-diapause quiescence are regulated.We should strive to elucidate this problem by well focussed experimental studies.The methodical viewpoint paper by Tauber et al. (1998) could help in planning such trials.

Fig. 1 .
Fig. 1.Effect o f temperature on completion o f the summer diapause o f Delia antiqua.(Before experiments, diapause was induced at 25°C, 16L : 8D.) a -percentage o f diapause comple tion obtained when individuals were exposed to five tempera tures, for time intervals indicated on the right margin.Values near the symbols indicate the numbers o f puparia used; beffect o f five temperatures, indicated on the right margin, on completion o f diapause in respect to the duration o f treatment.(Ishikawa et al., 2000)

Fig. 2 .
Fig. 2. Termination o f summer diapause in females o f two Chrysolina spp.under contrasting photoperiod and humidity conditions, recorded by production o f faeces (a, b) and cumulative proportion o f ovipositing females (c).(Schops et al., 1996)

Fig. 3 .
Fig. 3. Pupation o f larvae o f Busseola fusca, diapausing from July and transferred in December and January to laboratory dry and wet conditions.(Okuda, 1988)

Fig. 4 .
Fig. 4. Pupation o f larvae o f Busseola fusca, diapausing from July and transferred in August and September to laboratory wet conditions.Before the experiment the insects were always stored for 2 months; at either wet conditions (a), or at dry condi tions (b, c).(Larvae in the September sample lived one month longer in the field drought than the August sample.)(Okuda, 1991)

Fig. 5
Fig. 5. a -pupation o f larvae o f Busseola fusca (diapausing from July, collected in mid January, n = 37), when drinking water without contact to moist environment (O) vs. in contact with moist environment (wetted, • ); b -fresh weight o f larvae under the two experimental procedures as indicated in (a).(Okuda, 1990) Fig. 7. Hatch o f diapause eggs o f Homichloda barkeri after three wetting procedures; inlet in (a) wet only once.a -wet every 23rd day; b -wet every day.(Nahrung & Merritt, 1999) While the breeding sites and food requirements of the beetle probably still remain unknown, experiments with dormant beetles reveal the role of environmental cues, namely photoperiod, in diapause development, and moisture in the resumption of post-diapause development.Although at the low latitude of Panama (9°N) the difference between the longest and shortest day is only one hour, the increasing photophase appears to stimulate from March the early weak growth of CA, primary oocytes and flight pora allata (d) in diapausing adults o f Stenotarsus rotundus.

Table 3 . Species in which moisture was reported as an environmental cue for diapause termination; mostly for post-diapause resumption o f activity/development, but also diapause development.
*more lepidopteran species discussed in this paper.