PHYSIOLOGY,BIOCHEMISTRY, AND TOXICOLOGY Diapause Traits of Melanoplus sanguinipes and Melanoplus borealis (Orthoptera: Acrididae)

DENNIS J. FIELDING1

USDAÐARS, P.O. Box 757200, University of Alaska, Fairbanks, AK 99775

Ann. Entomol. Soc. Am. 101(2): 439Ð448 (2008) ABSTRACT A thorough understanding of diapause development in insects is needed for prediction of population responses to climate change, for realistic simulation models, and for effective pest management. In Melanoplus sanguinipes F. and Melanoplus borealis (Fieber) (Orthoptera: Acrididae), diapause typically occurs in embryos at a late stage of development. Experiments were conducted to compare diapause traits in two populations of M. sanguinipes, one from Alaska and one from Idaho, and also in a population of M. borealis from Alaska. Respiration was measured at different stages of embryonic development. As the embryos entered diapause, respiration rates declined over a period of 4Ð8 d to levels Ϸ70% lower than peak rates of prediapause, late stage embryos, in all populations. In the Idaho population, subjecting prediapause embryos to a cold treatment (5ЊC) caused the embryo to avert diapause: respiration in these embryos continued to increase after chilling as the embryo developed directly to hatching. In M. borealis and M. sanguinipes from Alaska, chilling of prediapause embryos did not affect subsequent diapause development: respiration rates in these embryos increased until they entered diapause and then declined. About the same amount of time at 5ЊC was necessary to avert diapause in early stage embryos of the Idaho population as was required to terminate diapause in late stage embryos. Respiration rates of diapausing embryos increased with increasing time spent at 5ЊC. Respiration continued to increase after transfer to warm temperatures in those that had completed diapause, but in those that had not, respiration soon declined back to diapause levels. In general, there was less ßexibility in the diapause program among grasshoppers from Alaska.

KEY WORDS diapause, Melanoplus, chilling, respiration

Diapause in insects is characterized by cessation of predict spread of invasive species and to predict the morphological development, even though tempera- response of insects to climate change (Bradshaw and ture and other environmental conditions may be fa- Holzapfel 2001, Regniere and Nealis 2002, Olfert et al. vorable for continued development. Diapause enables 2004). insects to synchronize growth and reproduction to Acridids are no exception when it comes to the predictable, seasonal changes in environment importance of diapause in their life cycle. From trop- (Tauber et al. 1986). Traits related to diapause, e.g., ical to subarctic climates, an understanding of the the stage at which it occurs, whether obligate or fac- diapause process is vital to understanding and pre- ultative, its duration, are tailored by natural selection dicting population dynamics of grasshoppers and lo- to favor adaptation of the organism to its environment custs (Valek and Coppel 1972, Cherrill and Begon (Tauber et al. 1986, Bradshaw et al. 2000, Fielding 1991, Groeters 1994, Tanaka 1994, Gehrken and 2006). Diapause may inßuence population dynamics Doumbia 1996, Hunter et al. 2001, Fielding 2006). by controlling voltinism (Steinbauer et al. 2004) and Different species of acridids may diapause at different by affecting interactions between parasitoids and pathogens and their hosts (Ringel et al. 1998, Corley stages, but most common is the egg stage. Among those et al. 2004). Effective pest management, including species that diapause as embryos there is considerable management of resistance to insecticides, relies on an variation in the different aspects of diapause, such as understanding of diapause attributes (Carrie`re et al. the stage of embryonic development at which dia- 1995, Mitchell and Onstad 2005), and models of insect pause occurs, whether diapause is facultative or obli- populations often require realistic simulation of the gate, and the time required to complete diapause diapause process (Gage et al. 1976, Carter et al. 1998, (Moore 1948, Kreasky 1960, Gage et al. 1976, Tanaka Onstad and Gould 1998). Furthermore, a thorough 1994). There also may be considerable variation understanding of the diapause process is necessary to among populations within a species in characteristics related to diapause (Dingle et al. 1990, Dingle and 1 Corresponding author, e-mail: [email protected]. Mousseau 1994, Groeters 1994). 440 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 2

Melanoplus sanguinipes F. has a wide geographic different stages of development to determine the stage distribution, from central Florida and Mexico to the at which diapause occurs, whether diapause could be interior of Alaska (Capinera et al. 2002, Pfadt 2002), averted by chilling of prediapause embryos, and the and it exhibits a correspondingly diverse range of amount of time at cool temperatures required to ter- diapause characteristics. This species overwinters in minate or avert diapause. the egg stage only, but the embryos do not need to be in diapause to survive the winter. In southern portions Materials and Methods of their range, populations are bivoltine and embry- onic diapause may be under maternal control, in re- Laboratory colonies of M. sanguinipes from Idaho sponse to changing photoperiod (Dean 1982). In more were initiated with at least 200 individuals collected northern regions where populations are univoltine, near Lewiston, ID (46.38Њ N, 117.02Њ W, 450-m eleva- diapause prevents the eggs from hatching too late in tion) as fourth and Þfth instars in 2003 and 2004. the season to complete another generation. In the far Laboratory colonies of M. sanguinipes and M. borealis north, eggs of semivoltine populations overwinter from Alaska were initiated with similar numbers of twice, but not because of an extended diapause or insects collected near Delta Junction, AK (64.00Њ N, because they diapause twice. Reproduction in Alaska 145.73Њ W, 400-m elevation). Grasshoppers were main- begins in late summer. Soil temperature soon declines tained on romaine lettuce and wheat bran, at room below the threshold for development, and embryos temperature with incandescent lights (16-h photo- spend the Þrst winter in an early stage of development phase) suspended above the cages to allow the grass- (D.J.F., unpublished data). These eggs are not in dia- hoppers to self-regulate their internal temperatures. pause when overwintering for the Þrst time. The next Trays of moist sand were provided for oviposition. summer, embryos develop further until entering dia- Offspring from the Þeld-collected grasshoppers were pause. Diapause is terminated sometime before mor- reared under the same conditions described above, phological development resumes the next spring and and all experiments were conducted on F2 and F3 the eggs hatch. generation eggs to avoid maternal effects from the Most populations of M. sanguinipes diapause in a late Þeld-collected generation. Oviposition trays were stage of development, Ϸ80% of development from sifted daily to obtain egg pods of known age. Egg pods oviposition to hatch (Moore 1948, Dingle et al. 1990). were stored in moistened vermiculite in plastic cups Diapause at a late stage of development allows for a (50 ml) with perforated lids. relatively early hatch date because less development Hatching in Relation to Developmental Stage and remains to be completed before hatching in the spring. Chill Time. To determine the effect of stage of de- Dingle and Mousseau (1994) reported variation in the velopment and duration of chilling on diapause avoid- diapause stage within and among populations of M. ance or termination, eggs of different ages were sanguinipes along an altitudinal gradient, with some chilled for different periods of time and subsequent diapausing before katatrepsis. development at higher temperatures was observed. Previous investigations suggested that diapause was Katatrepsis (reorientation of the embryo within the obligate in an Alaskan population of M. sanguinipes egg, also referred to as blastokinesis Chapman 1982), (Fielding 2006). If chilled at early, prediapause stages was used as a benchmark to determine stages of em- of development (simulating overwintering for the Þrst bryological development. Katatrepsis in Melanoplus time) and then returned to warm temperatures, the spp. takes place shortly after the temporal midpoint of embryos continued development until diapause was embryological development (Moore 1948). Eggs were initiated and then would not hatch unless diapause held at 22 Ϯ 1ЊC until transfer to a refrigerator main- was terminated by chilling a second time. By contrast, tained at 5 Ϯ 1ЊC. Egg pods were transferred to the eggs of the same species from Idaho hatched after cool temperature at three different stages of devel- experiencing cool temperatures at any stage of devel- opment: before katatrepsis, after katatrepsis but be- opment. It seemed that embryos of the Idaho popu- fore diapause, and after diapause was initiated. At lation averted, or avoided, diapause, as opposed to 22ЊC. These stages of development correspond to ages entering diapause at an early stage of development. of 12Ð17 d, 22Ð27 d, and Ͼ36 d for both species of Church and Salt (1952) alluded to diapause avoidance Alaskan grasshoppers. Three to Þve eggs from each in Melanoplus bivittatus (Say), and Cherrill and Begon pod were examined to conÞrm the embryosÕ stage of (1991) describe a similar phenomenon in Chorthippus development at the time of transfer to 5ЊC. Eggs from brunneus Thunberg. the Idaho population showed somewhat more rapid The objective of the current study was to charac- development (Fielding 2006), and they were trans- terize the process of diapause development and avoid- ferred to 5ЊC at ages of 10Ð15, 20Ð24, and Ͼ36 d. ance in two populations of M. sanguinipes: one popu- Approximately one third of the eggs from each pod lation, from Alaska, with an apparently obligate was removed from the refrigerator after each of three diapause; and another population, from Idaho, able to periods of chilling: 28Ð32, 56Ð60, and 84Ð90 d. The avoid diapause. Because it is an important economic eggs were then incubated at 22ЊC and inspected daily pest in Alaska, diapause development in Melanoplus for hatching. After 26Ð32 d at 22ЊC, the eggs that had borealis (Fieber) also was studied. Respiration rates, not hatched were examined to see whether normal percentage of hatch, and days to hatch were measured development had taken place, the criterion being pig- for eggs chilled for different periods of time and at mentation of the compound eyes of late stage embryos March 2008 FIELDING:DIAPAUSE TRAITS OF Melanoplus SPECIES 441

visible through the chorion. These eggs were chilled Repeated measurements of CO2 production were a second time at 5ЊC for another 90Ð100 d, after which made on individual eggs. Between measurement pe- they were again returned to 22ЊC where hatching was riods, individual eggs remained in the syringes with recorded daily. The total number of eggs to hatch the stopcock open. These syringes were kept at 22ЊC (100 Ͻ N Ͻ 200) after the Þrst and second periods of in a plastic food storage box with a wet paper towel to chilling was used as the denominator in calculating the maintain humidity. proportion hatching after the Þrst period of chilling CO2 was measured with an infrared gas analyzer only. Examination of unhatched eggs after the second (LI-6252, LICOR Inc., Lincoln, NE). Two milliliters of cold period did not reveal any remaining viable eggs air from the syringes was injected into a CO2-free (Ͻ10% of all eggs in all treatments). airstream and passed through a magnesium perchlor- Differences in proportions of eggs hatching were ate column to remove water before passing through tested using logistic regression (PROC GENMOD, the gas analyzer. Flow rate was maintained at 100 ml SAS Institute 2003) with a binomial probability dis- minϪ1 STP with a mass ßow controller (C100 Smart- tribution and logit link function. Chill times (4, 8, or Trak, Sierra Instruments, Inc., Monterey, CA). Analog

12 wk) were treated as a continuous variable and signals from the CO2 analyzer were converted to dig- population and stage of development as categorical ital output (ppm) and recorded with a computer run- variables. Overdispersion was accounted for by using ning data acquisition software (Sable Systems, Inc., the SCALE ϭ DEVIANCE option in PROC GENMOD. Las Vegas, NV). Twenty readings were averaged and

Population and treatment effects were evaluated using recorded per 0.5 s. Continuous time plots of CO2 type III likelihood ratios. concentration in the airstream were converted to mil-

Incidence of Early Diapause. Incidence of early dia- liliters of CO2 per hour based on the ßow rate of 100 pause was assessed by incubating eggs at 22ЊC for 36 d. ml/min. By integrating the curve against time, the

Eggs were examined under a binocular microscope. If total milliliters of CO2 in the 2-ml air sample was eyespots were visible through the chorion, it was as- determined. CO2 produced by the egg was calculated sumed that embryos were in a late stage of development. by subtracting the initial volume of CO2 per 2 ml air Eggs with no visible development were segregated and from the Þnal volume, correcting for the total volume chilled for 90Ð100 d, after which they were maintained of the syringe while incubating (3 ml). Results were Њ at 22 C and hatching was recorded. Those eggs that had converted to microliters of CO2 per hour per egg. not exhibited development after 36 d, but hatched after Mass-speciÞc metabolic rates were not determined chilling, were assumed to have been in an early stage because it was not possible to determine the mass of diapause. Those eggs that did not hatch were assumed to the embryo without removing it from the egg; how- be nonviable. More than 500 eggs were examined in each ever, live weights were determined for 50 diapausing species and population. eggs of each species and population. Respirometry. In insects with an embryonic dia- Sample size ranged from 15 to 26 eggs at each mea- pause, respiration drops dramatically as the embryo surement. Respiration rates were normally distributed enters diapause (Gray et al. 1991). In the current and compared with analysis of variance (PROC GLM, study, I measured CO2 production in embryos as they SAS Institute 2003) with population, stage of devel- entered diapause and after diapause was terminated, opment, and duration of chilling (4, 8, or 12 wk) as in diapausing embryos after being chilled for different main effects. Differences between means were tested intervals, and in prediapause embryos after experi- using Bonferroni-corrected t-tests. encing a period of chilling. CO2 production from in- dividual eggs was measured using constant volume Results techniques. Measurements on prediapause, prechill- ing eggs were made at 12, 17, and 22 d of age and then Hatching in Relation to Developmental Stage and every 2 d until respiration dropped to diapause levels. Chill Time. Eggs from the Idaho population hatched Eggs were transferred to 5ЊC at different stages of after chilling at any stage of embryonic develop- development and maintained at 5ЊC for 4, 8, or 12 wk ment, whereas very few prediapause eggs from the as described above. Respiration was measured at 2, 4, Alaskan populations hatched after chilling (Fig. 1). 6, 9, and 15 d postchilling or until the egg hatched. There was a highly signiÞcant population by stage of Measurements were initiated on the second day after development interaction effect on proportion of transfer to 22ЊC to allow time for the embryos to eggs hatching (␹2 ϭ 149, df ϭ 8, P Ͻ 0.0001). When resume development. populations were analyzed separately, the propor- Eggs were carefully removed from pods with the tions of eggs hatching were strongly affected by chorion intact and placed in 5-ml syringes equipped stage of development and chill time in Alaskan pop- with stopcocks. Five milliliters of room air, which had ulations of M. sanguinipes and M. borealis (␹2 Ͼ 24, ϭ Ͻ passed through a soda-lime column to remove the CO2 df 2, P 0.0001 in all cases; Fig. 1A and B). In the and then through distilled water to prevent drying of Idaho population of M. sanguinipes, however, stage the eggs, was drawn into the syringes, which were then of development had no effect on proportion hatch- sealed. Syringes were incubated at 22 Ϯ 1ЊC for 18Ð24 ing (␹2 ϭ 2.1, df ϭ 2, P ϭ 0.34; Fig. 1C). Ϸ h. Because dissolved CO2 in the water was introduced Diapausing embryos of M. borealis required 8 into the airstream, CO2 levels were measured at the wk of chilling before hatching. There was no sig- beginning of the measurement period and at the end. niÞcant difference in hatching rates in diapausing M. 442 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 2

Fig. 1. Percentage of grasshopper eggs hatching after being chilled at 5ЊC for 4, 8, or 12 wk. Total number of eggs examined for each combination of treatments and populations are given in parentheses above the columns. Prekatastrepsis eggs were incubated at 22ЊC for 12Ð17 d before chilling, postkatastrepsis eggs were incubated for 20Ð27 d before chilling, and diapausing eggs were incubated for 36Ð42 d before chilling. (A) M. borealis from Alaska. (B) M. sanguinipes from Alaska. (C) M. sanguinipes from Idaho. Error bars indicate 95% CL. borealis after chilling for 8 versus 12 wk (␹2 ϭ 2.5, further 22Ð32 d of incubation at 22ЊC, indicating that df ϭ 1, P ϭ 0.11), with Ͼ90% hatch in either case. they were not in an early stage diapause. Diapausing embryos of M. sanguinipesÐAlaska The number of days for eggs to hatch after chilling was showed a more graded response to chill times, with greater for eggs chilled at earlier stages of development, steadily increasing percentage of hatch with in- as would be expected (Fig. 2). Postkatatrepsis eggs of M. creasing chilling (Fig. 1B). borealis that hatched did so within 11Ð15 d, and 90% of All stages of development in M. sanguinipesÐIdaho diapausing M. borealis hatched within 4Ð8 d. The nine of showed increasing percentage of hatch with increas- 148 prekatatrepsis M. sanguinipesÐAlaska eggs that ing chilling time (Fig. 1C). Of the prediapause stage hatched after 12 wk of chilling (Fig. 1B) required 21Ð24 eggs that did not hatch after the Þrst period of chilling, d (Fig. 2). Postkatatrepsis M. sanguinipesÐAlaska em- all had reached late stages of development after a bryos required 8Ð18 d to hatch, and diapausers hatched March 2008 FIELDING:DIAPAUSE TRAITS OF Melanoplus SPECIES 443

Fig. 2. Number of days to hatch after being chilled at 5ЊC for 4, 8, or 12 wk. Age of prekatastrepsis, postkatastrepsis, and diapausing embryos as in Fig. 2. Box indicates Ϯ 1 SD, bars indicate 5th and 95th percentiles of the population, and dots indicate extreme values. (A) M. borealis from Alaska. (B) M. sanguinipes from Alaska. (C) M. sanguinipes from Idaho. within 4Ð11 d. Of the M. sanguinipesÐIdaho prekatatrep- resumed development to hatching. The remaining 95% sis embryos that hatched, 90% did so within 11Ð24 d either were in an advanced stage of development after postchilling. Hatching occurred within 8Ð18 d and 5Ð14 36dat22ЊC or did not hatch at all. The stage at which d for postkatatrepsis and diapausing eggs, respectively, of early diapausers diapaused was not determined because M. sanguinipesÐIdaho. without bleaching or dissecting the early stages could not Early Stage Diapause. No eggs of the Alaskan species be determined, and if destroyed, could not be distin- of grasshoppers showed any evidence of early stage dia- guished from dead or moribund embryos. pause. A small portion of eggs, 23 of 504, from Idaho Diapause Development as Revealed by Respiration apparently did diapause at an early stage. After incuba- Rates. Respiration rates in prediapause embryos in- tion for 36 d at 22ЊC, the embryos were in an early stage creased as development progressed and then declined of development, and then after 12 wk of chilling, they as the embryos entered diapause (Fig. 3). Mean res- 444 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 2

Fig. 3. Respiration rates of eggs incubated at 22ЊC for 36 d, then held at 5ЊC for 12 wk, and then transferred back to 22ЊC until hatching. Error bars indicate Ϯ 1 SD. For each point, 14 Ͻ N Ͻ 27. (A) M. borealis from Alaska. (B) M. sanguinipes from Alaska. (C) M. sanguinipes from Idaho.

ϭ piration rates of diapausing embryos was about one differed among populations and species (F2, 83 12.2; third the rate at the prediapause peak. The Idaho P Ͻ 0.0001), with M. borealis (mean Ϯ SE, 0.53 Ϯ 0.026 ␮ Ϫ1 Ϫ1 Ϯ embryos tended to reach peak respiration rates and lCO2 egg h ) and M. sanguinipesÐAlaska (0.48 enter diapause in less time than did M. sanguinipes- 0.034) not differing from each other (P ϭ 0.58; Bon- Alaska or M. borealis (Fig. 3). After diapausing eggs ferroni corrected t-test), but both differing signiÞ- were chilled for 12 wk, the completion of diapause was cantly (P Ͻ 0.02) from M. sanguinipesÐIdaho (0.36 Ϯ indicated by high respiration rates after removal from 0.024). Some of these differences may be due to dif- 5ЊC (Fig. 3). Respiration rates 2 d after transfer to 22ЊC ferences in mass of the embryos: M. borealis diapaus- March 2008 FIELDING:DIAPAUSE TRAITS OF Melanoplus SPECIES 445

Fig. 4. Respiration rates of diapausing embryos at 22ЊC after being chilled for 4 wk (Þlled symbols) or 8 wk (open symbols). Error bars indicate Ϯ 1 SD. For each point, 14 Ͻ N Ͻ 27. (A) M. borealis from Alaska. (B) M. sanguinipes from Alaska. (C) M. sanguinipes from Idaho. ing eggs weighed 5.6 Ϯ 0.10 mg (mean Ϯ SE), M. Ϯ Fig. 5. Respiration rates of prediapausing embryos incu- sanguinipesÐAlaska weighed 5.7 0.11 mg); and M. bated at 22ЊC for 24 d, then held at 5ЊC for 12 wk, and then sanguinipesÐIdaho weighed (5.0 Ϯ 0.05 mg). Respira- transferred back 22ЊC until respiration dropped to diapause tion rates in these postdiapause embryos continued to levels or until hatched. Error bars indicate Ϯ 1 SD. For each increase until hatch in all populations (Fig. 3). point, 14 Ͻ N Ͻ 27. (A) M. borealis from Alaska. (B) M. Respiration rates of diapause-stage embryos were sanguinipes from Alaska. (C) M. sanguinipes from Idaho. indicative of diapause development after different pe- riods of chilling. Termination of diapause was evident Differences in respiration rates between eggs chilled in the higher respiration rate of eggs that hatched for four versus 8 wk were evident even in embryos that subsequent to chilling than in those that remained in did not terminate diapause. Two days after removal diapause. Two-days after transfer from 5 to 22ЊC, res- from 5ЊC, respiration was greater in eggs chilled 8 wk piration rates of eggs that subsequently hatched were than after 4 wk of chilling (0.33 Ϯ 0.015 versus 0.23 Ϯ Ϯ Ϯ Ϯ ␮ Ϫ1 [minu]1 ϭ Ͻ (mean SE) 0.52 0.020 compared with 0.29 0.011 0.012 lCO2 egg h )(F1, 98 25.5; P 0.0001, ␮ Ϫ1 Ϫ1 ϭ lCO2 egg h in eggs that did not hatch (F1, 142 averaged over all species). Too few eggs remained in 106.7; P Ͻ 0.0001, averaged over all populations and diapause after 12 wk of chilling for comparison. Pop- ϭ ϭ chill times). ulation effect was nonsigniÞcant (F2, 98 1.9; P 0.15), ϭ In diapause-stage embryos that did not terminate as was population by chill time interaction (F2, 98 0.7; diapause, respiration rates shortly after transfer from P ϭ 0.50). 5to22ЊC were elevated somewhat above prechilling Population differences were obvious in respira- rates, but then returned to diapause levels (Fig. 4). tion rates of prediapause embryos in which devel- 446 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 2

Fig. 6. Probabilities of the last freezing temperature in spring and Þrst freezing temperature in the fall occurring on or after a given date at Lewiston, ID (open circles) and Delta Junction, AK (black circles). Probabilities were determined from data recorded at National Oceanic and Atmospheric Administration weather stations from 1949 to 2005, in Lewiston, ID, and 1942Ð2003 in Big Delta, AK (WRCC 2004). opment was interrupted by chilling before entering Dingle and Mousseau (1994) reported early stage diapause (Fig. 5). All populations resumed normal diapause in some populations of M. sanguinipes. No respiration rates after 12 wk at 5ЊC, but in both Alaskan grasshoppers were found to diapause in an groups of Alaskan grasshoppers, respiration early stage, and only in a small proportion of the Idaho dropped as the embryos entered diapause (Fig. 5A population was early diapause indicated. Simulations and B), whereas in M. sanguinipesÐIdaho, respira- suggest that facultative averting of diapause is a better tion rates continued increasing as development pro- tactic than early stage diapause under almost all ceeded directly to hatching (Fig. 5C). lengths of season (Fielding 2006). Early diapause would interrupt normal development when time was still potentially available in most seasons for continued Discussion development. If the eggs were laid late in the season, In the Alaskan grasshoppers M. sanguinipes and M. or there was an unusually short growing season, the borealis, diapause was not averted in the majority of ability to avert diapause would provide the ßexibility embryos by chilling during early stages of develop- to hatch the next year without Þrst diapausing. ment, whereas in the Idaho population, diapause Detection of early stage diapause is problematic, could be averted by chilling of prediapause embryos. particularly when it is not the dominant stage of dia- About the same amount of time at 5ЊC was required to pause. The difÞculty lies in determining whether an avert diapause in the Idaho population as was required early stage embryo is in diapause or is developmentally to terminate diapause once diapause was induced. delayed by environmental stresses. Desiccation may This suggests similar mechanisms may be involved in delay development (Mukerji and Gage 1978), and it the two processes. also seems possible that saturated soil could limit ox- Differences in season length may explain the dif- ygen availability, resulting developmental delays. It ferences between the Idaho and Alaskan populations may be possible that these developmentally delayed of M. sanguinipes in their ability to avert diapause embryos revived during chilling such that they were (Fielding 2006). The average frost-free season at low- elevation sites in Idaho is Ϸ80 d longer than in interior able to subsequently resume development. Because Alaska (Fig. 6; WRCC 2004). M. sanguinipes and M. chilling would have averted diapause in the Idaho borealis in Alaska have a semivoltine life cycle because population, they would then readily hatch, giving the the eggs overwinter twice. If diapause were averted at appearance of early stage diapause. Respiration rates an early stage of development by overwintering the would not distinguish between diapausing and mori- Þrst time, hatching would occur later in the season, bund embryos, especially in early stages of develop- perhaps too late for the grasshoppers to mature and ment. By providing consistent conditions in the lab- reproduce. By contrast, the grasshoppers in Idaho oratory setting, these environmental stresses were have enough time to complete nymphal development minimized, and diapause may be the most parsimoni- even if they hatch late. ous explanation for development halting in the early March 2008 FIELDING:DIAPAUSE TRAITS OF Melanoplus SPECIES 447 stages, but I do not believe the evidence for early stage Cherrill, A., and M. Begon. 1991. Oviposition date and pat- diapause is unequivocal in this study. tern of embryogenesis in the grasshopper Chorthippus In many regions, consideration of prediapause de- brunneus (Orthoptera, Acrididae). Holarctic Ecol. 14: velopment and diapause avoidance may seldom be 225Ð233. important to understanding phenology. In localities Church, N. S., and R. W. Salt. 1952. Some effects of tem- where most adults die well before the end of the perature on development and diapause in eggs of Mela- season, most oviposition occurs early enough for em- noplus bivittatus (Say) (Orthoptera: Acrdididae). Can. J. bryos to reach diapause stage. But, overwintering Zool. 30: 173Ð184. Corley, J. C., A. F. Capurro, and C. Bernstein. 2004. Pro- in prediapause stage may be more common than longed diapause and the stability of host-parasitoid in- thought. Late season precipitation can cause a ßush of teractions. Theor. Popul. Biol. 65: 193Ð203. new vegetative growth, which then might fuel high Dean, J. M. 1982. Control of diapause induction by a change rates of egg production by surviving grasshoppers, or in photoperiod in Melanoplus sanguinipes. J. Insect dry soil conditions may slow or halt embryological Physiol. 28: 1035Ð1040. development temporarily (Mukerji and Gage 1978), Dingle, H., T. A. Mousseau, and S. M. Scott. 1990. Altitudinal leaving embryos to overwinter in an early stage of variation in life cycle syndromes of California populations development. Also, in areas with a growing season of the grasshopper, Melanoplus sanguinipes (F.). Oeco- long enough for a second generation, oviposition by logia (Berl.) 84: 199Ð206. the second generation may occur too late in the season Dingle, H., and T. A. Mousseau. 1994. Geographic variation to reach the diapause stage. in embryonic development time and stage of diapause in Consideration of diapause traits is important in a grasshopper. Oecologia (Berl.) 97: 179Ð185. modeling of pest populations. Phenological models to Fielding, D. J. 2006. Optimal diapause strategies in a grass- predict dates of grasshopper hatching are based on hopper, Melanoplus sanguinipes. J. Insect Sci. (http:// implicit or explicit assumptions regarding diapause www.insectscience.org/6.02). development. Gage et al. (1976) assumed diapause Gage, S. H., M. K. Mukerji, and R. L. Randell. 1976. A pre- dictive model for seasonal occurrence of three grasshop- avoidance, and thus they incorporated estimates of per species in Saskatchewan (Orthoptera: Acrididae). prediapause development that occurred the previous Can. Entomol. 108: 245Ð253. fall. Other phenological models that only include post- Gehrken, U., and Y. O. Doumbia. 1996. Diapause and qui- diapause development assume that embryos have en- escence in eggs of a tropical grasshopper Oedaleus senega- tered diapause before winter. If this assumption is not lensis (Krauss). J. Insect Physiol. 42: 483Ð491. valid, there may be an extended period of hatching, if Gray, D. R., J. A. Logan, F. W. Ravlin, and J. A. Carlson. 1991. diapause was averted by overwintering, or a portion of Toward a model of gypsy-moth egg phenologyÐusing the eggs may not hatch until the second year, if dia- respiration rates of individual eggs to determine temper- pause was not averted. ature time requirements of prediapause development. Environ. Entomol. 20: 1645Ð1652. Groeters, F. R. 1994. The adaptive role of facultative Acknowledgments embryonic diapause in the grasshopper Caledia captiva (Orthoptera: Acrididae) in southeastern Australia. Daniel LaBarre, Katherine Beatty, and Linda DeFoliart Ecography 17: 221Ð228. assisted with grasshopper colonies and respirometry. The Hunter, D. M., P. W. Walker, and R. J. Elder. 2001. Adap- manuscript was improved by the comments and suggestions tations of locusts and grasshoppers to the low and variable of A. Hagerty and A. Pantoja and anonymous reviewers. rainfall of Australia. J. Orthopt. Res. 10: 347Ð351. Kreasky, J. B. 1960. Extended diapause in eggs of high-alti- tude species of grasshoppers, and a note on food-plant References Cited preferences of Melanoplus bruneri. Ann. Entomol. Soc. Am. 53: 436Ð438. Bradshaw, W. E., S. Fujiyama, and C. M. Holzapfel. 2000. Mitchell, P. D., and D. W. Onstad. 2005. Effect of extended Adaptation to the thermal climate of North America by diapause on the evolution of resistance to transgenic the pitcher-plant mosquito, Wyeomyia smithii. Ecology Bacillus thuringiensis corn by northern corn rootworm 81: 1262Ð1272. (Coleoptera: Chrysomelidae). J. Econ. Entomol. 98: Bradshaw, W. E., and C. M. Holzapfel. 2001. Genetic shift in 2220Ð2234. photoperiodic response correlated with global warming. Moore, H. W. 1948. Variations in fall embryological devel- Proc. Natl. Acad. Sci. U.S.A. 98: 14509Ð14511. opment in three grasshopper species. Can. Entomol. 80: Capinera, J. L., C. W. Scherer, and J. M. Squitier. 2002. 83Ð88. Grasshoppers of Florida. University Press of Florida, Gainsville, FL. Mukerji, M. K., and S. H. Gage. 1978. A model for estimating Carriere, Y., D. A. Roff, and J. P. Deland. 1995. The joint hatch and mortality of grasshopper egg populations based evolution of diapause and insecticide resistanceÐa test of on soil moisture and heat. Ann. Entomol. Soc. Am. 71: an optimality model. Ecology 76: 1497Ð1505. 183Ð190. Carter, M. R., I. V. Macrae, J. A. Logan, and T. O. Holtzer. Olfert, O., R. M. Weiss, S. Woods, H. Philip, and L. Dosdall. 1998. Population model for Melanoplus sanguinipes (Or- 2004. Potential distribution and relative abundance of an thoptera: Acrididae) and an analysis of grasshopper pop- invasive cereal crop pest, Oulema melanopus (Coleoptera ulation ßuctuations in Colorado. Environ. Entomol. 27: : Chrysomelidae), in Canada. Can. Entomol. 136: 277Ð287. 892Ð901. Onstad, D. W., and F. Gould. 1998. Modeling the dynamics Chapman, R. F. 1982. The insects: structure and function, of adaptation to transgenic maize by European corn borer 3rd ed. Harvard University Press, Cambridge, MA. (Lepidoptera: Pyralidae). J. Econ. Entomol. 91: 585Ð593. 448 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 2

Pfadt, R. E. 2002. Field guide to western grasshoppers, 3rd Tanaka, H. 1994. Embryonic diapause and life cycle in the ed. Bulletin 912, Wyoming Agricultural Experiment Sta- Migratory locust, Locusta migratoria L. (Orthoptera: tion, Laramie, WY. Acrididae) in Kyoto. Appl. Entomol. Zool. 29: 179Ð191. Regniere, J., and V. Nealis. 2002. Modelling seasonality of Tauber, M. J., C. A. Tauber, and S. Masaki. 1986. Seasonal gypsy moth, Lymantria dispar (Lepidoptera: Lymantri- adaptations of insects. Oxford University Press, New idae), to evaluate probability of its persistence in novel York. environments. Can. Entomol. 134: 805Ð824. Ringel, M. S., M. Rees, and H.C.J. Godfray. 1998. The evo- Valek, D. A., and H. C. Coppel. 1972. Bionomics of an oak lution of diapause in a coupled host-parasitoid system. J. defoliating grasshopper, Dendrotettix quercus, in Wiscon- Theor. Biol. 194: 195Ð204. sin. Ann. Entomol. Soc. Am. 65: 310Ð319. SAS Institute. 2003. SAS for Windows, release 9.1. SAS In- [WRCC] Western Regional Climate Center. 2004. West- stitute, Cary, NC. ern Regional Climate Center, Reno, NV. (http://www. Steinbauer, M. J., D. J. Kriticos, Z. Lukacs, and A. R. Clarke. wrcc.dri.edu/index.html). 2004. Modelling a forest lepidopteran: phenological plas- ticity determines voltinism which inßuences population dynamics. For. Ecol. Manag. 198: 117Ð131. Received 13 June 2007; accepted 24 October 2007.