Abstract COLD HARDINESS OF FIRST INSTAR LARVAE OF THE FOREST TENT CATERPILLAR, MALACOSOMA DlSSTRIA (LEPIDOPTERA: LASIOCAMPIDAE) A. G. RASKE l Northern Forest Research Centre, Canadian Forestry Service, Edmonton, Alberta Can. Ent. 107: 75-80(1975) Unfed forest tent caterpillar (Malacosoma disstria Hbn.) larvae were subjected to temperature treatments of -180, -120, _70, _10,40, and 22 C, each at durations of 2 days, 7 days, and 14 days. Each temperature-duration treatment was initiated 3 days before expected hatch, and 3 days after hatch. Fed larvae were subjected to identical treatments 2 days after hatch. Temperature treatments of the before hatch group resulted in appreciable mortality only at -18 C for 7 days and at -18 C and -12 C for 14 days. Temperature treatments of the after hatch group resulted in nearly 100% mortality at -18 C for all durations, at -12 C for 14 days, and about 50% mortality at -7 C for 14 days. combinations caused little mortality. fed larvae did not differ from unfed larvae of the after hatch group. Resume Other temperature-duration The mortality resulting from temperature treatments of Des larves de la Livree des forets (Malacosoma disstria Hbn.) privees de nourriture furent soumises a des temperatures de -18, -12, -7, -I 0, _10, 40 et 22 C durant, 2, 7 et 14 jours. Chaque traitement debuta 3 jours avant I' eciosion estimee ou 3 jours apres I' eclosion. l'eciosion. Des larves nourries furent soumises a des traitements identiques deux jours apres A la suite du traitement, chez Ie groupe pre-eclosion, une mortalite appreciable se produisit seulement a -18 C et 7 jours, et a _180 C et -12 C et 14 jours. Chez Ie groupe deja eclos, la mortalite fut de presque 100% a-18 C et n'importe quelle duree, et a -120 C et 14 jours; il y eut environ 50% de mortalite a -7 C et 14 jours. Les autres traitements causerent peu de mortalite. La mortalite resultant de traitements de larves nourries ne differa pas de celie qui se produisit chez les larves privees de nourriture appartenant au groupe eclos. Introduction Forest tent caterpillar population levels are characterized by large fluctuations and sudden collapses of populations. Of particular interest is the simultaneous collapse of outbreak populations over large areas. When this occurred subnormal temperature was suspected to be the cause of these. Lowe (1899), Tothill (1920, 1923), and Sweetman (1940) reported a correlation between early instar mortality of the forest tent caterpillar and cool spring weather. Hodson (1941), however, concluded that multiple factors cause the collapse of outbreaks, but noted that spring frosts after the hatching date were destructive. Blais et al. (1955) were the first to document the collapse of a forest tent caterpillar outbreak simultaneous with a spring snow storm. Prentice (1954) and Gautreau (1964) reported correlations of mortality within the eggs with unfavourable weather. Davidson and Prentice (1968) considered weather to be the only natural factor capable of bringing an outbreak to an end within the first 3 years. Witter and Kulman (1972) suggested that warm spring temperatures followed by freezing temperatures caused about 40% mortality by killing pharate larvae (fully-developed first-instar larvae within eggs). Ives (1973) concluded that a subnormal cumulative-total of heat units during the early larval feeding period was associated with population declines. However, the effect of controlled temperatures on the survival of forest tent caterpillar larvae seldom had been studied. Prentice and Heron (1956) studied the effect of unusual warm fall temperatures on egg yolk depletion and concluded that prolonged warm temperatures in autumn do not cause mortality by egg yolk de ple- 'Present address: Newfoundland Forest Research Centre, Canadian Forestry Service, P.O. Box 6028, Sl. John's, Nfld. 75
76 THE CANADIAN ENTOMOLOGIST January 1975 tion. Hanec (1966) studied the relationship between glycerol content and ability to supercool eggs and first instar larvae, and concluded that larvae survived at lower temperatures than those temperatures reported by Blais et al. (1955) to have caused the collapse of the outbreak. Wetzel et al. (1973) subjected pharate larvae to cold temperatures of -1 ooe to - 300e at 5 intervals. They found that the critical spring temperature of pharate larvae under laboratory conditions is about -200e, since warmer temperatures did not cause mortality levels very different from controls. This paper reports the relationship between low temperature and mortality of first instar larvae. Pharate and Unfed Larvae Methods Eggbands were collected on 11 January 1967, east of Drayton Valley, Alberta, where the outbreak was beginning, stored at 1O. 8 e, and transferred to room temperature at an R.H. of 70-75% on 4 March 1968. Previous experiments established that at this time of year, and under these temperatures, eggbands would hatch in 5-6 days, making it possible to initiate treatments 3 days before general hatch. A factorial experiment in a completely randomized design was chosen, and since the data were expected to be highly variable, 20 replications for each treatment were thought necessary. Eggbands were placed one to a vial, and the vials were sealed with cotton plugs. The plugs were kept moist throughout the experiment to prevent desiccation of the eggs. Eggbands were chosen at random, placed in lots of 20, and each lot was randomly assigned to a temperature, duration, and timing. Temperature treatments were -18, -12, -7, - 10, 4 and 22 e each at a duration of 2, 7, and 14 days. Treatments were started on 6 March for the pharate larvae and on 12 March for the first instar larvae. Two days after cold treatment the number of living and dead larvae in each vial was recorded. Fed Larvae To test the cold-hardiness of fed larvae, newly-hatched larvae from several eggbands were fed for 2 days beginning on 14 March and then separated into groups of 100 larvae which were randomly assigned a treatment of temperature and duration. Temperature and duration treatments were identical with those used for unfed larvae. Two days after treatment the number of living and dead larvae was recorded. Data Analysis Direct mortality of hatched larvae could be examined only for the first instar larvae but the number of survivors could be determined for both groups of larvae and these data were used for statistical analyses. An ANOVA was used to test for differences among treatments. The expected numbers of living caterpillars (1') were calculated from multiple regression analyses. Tukey's Multiple Range Test was used to test the difference between means of the expected numbers. A contingency table analysis was used to test for significant differences among mean numbers of survivors of fed larvae. Pharate and Unfed Larvae Results The mortality of newly-hatched larvae, directly attributable to the cold treatment in the laboratory, was severe only at temperatures below those of normal spring cold periods. At the 2-day duration, mortality was 99. 7% at -18 e, and less than 1 % at all other temperatures; at the 7-day duration, mortality was 100% at -18 e, 17.1% at -l2oe, and less than 1 % at warmer temperatures; at the I4-day duration, mortality was
Volume 107 THE CANADIAN ENTOMOLOGIST 77 more than 99% at -18cC and -12cC, 67.1 % at -7cC, and less than 1 % at warmer temperatures. The ANOV A showed highly significant differences in the number of survivors for temperature, duration, timing, and for all interactions. The expected number of living larvae (Y) per treatment was computed as follows: Pharate larvae y = 100.30974-0.20272T + 0.00000949DT3 + 0.57279DT - 0.28624D2-0.01459DT2 + 0.00198T3 where D = duration of cold treatment in days, and T = temperature in cf, R(multiple regression coefficient) = 0.849 (d.f. = 342 and 6), significant at the I % level. S.E.r = 9.21. Unfed first instar larvae y= 17.52001 + 13.68063T - 0.40982T2-0.16705D2 + 3.52858DT2 + 0.00324T3-0.00329DT3-0.7702lDT R = 0.940 (d.f. = 342 and 6), significant at the 1 % level S.E.r= 7.75. The observed (Y) and calculated (Y) numbers of living larvae survlvlng the treatment are given in Table I. Tukey's Multiple Range Test showed that a difference in means (Y) of 21 to 25 larvae was statistically significant. Examination of the means shows that significant differences were concentrated at the temperatures of -18 C and -12 C, but significant differences also occurred at other temperatures, and in the controls. Fed larvae Fed larvae survived cold treatments of 2 and 7 days, as low as -12cC (Table II) but none survived at -18 e. At treatment durations of 14 days, temperatures of -7 C and lower resulted in 100% mortality, and only six larvae survived at -l e. A Contingency Table Analysis of the data gave a X 2 10 value of 192.50; highly significant. When the 14-day duration was omitted from analysis, the 2 X 10 was 7.09; not significant. This 2 X shows that the mortality caused by 2-day duration of cold treatments did not differ from the 7 -day duration. Discussion First instar forest tent caterpillar larvae under laboratory conditions can survive temperatures of -12 C for at least 7 days. At temperatures that approach the lows of Alberta spring weather (-7 C to -ICC), there was little mortality. This agrees with Hanec (1966), who found the mean supercooling point for newly-hatched larvae to be -13.4cC and that maximum survival time of newly-hatched larvae at -12cC to be 8 days. About 3 days before hatching, pharate larvae withstood temperatures as low as -12cC with little mortality, while about 45% survived temperatures of -18cC for the longest duration, 14 days. These percentages are not directly comparable with those reported by Wetzel et al. (1973) because my statistics are reduction in survival while theirs are percentage mortality of pharate larvae within eggs. There is general agreement since they reported the critical spring temperature for forest tent caterpillar pharate larvae under laboratory conditions to be about at -20 e. This approximates my results of little change in survival at -18 C for the 2-day duration.
-.J 00 Table I. Expected and observed average number of living caterpillars hatched from forest tent caterpillar eggbands at various temperature-duration combinations. (Expected numbers were estimated from multiple regression analysis equations) Duration of cold treatment in days 2 7 14 2 7 14-18 C -Irc -7 C -1 C 4 C Controls 22 C E p. Obs. E p. Obs. E p. Obs. E p. Obs. E p. Obs. E p. Obs. Y Y Y Y Y Y Y Y Y Y Y Y Pharate 102. 4 104.2 108.1 102.2 112. 3 IIS.4 116.3 114.9 121.4 120.7 160.3 160.1 76.S 69.0 121.2 144.4 133.6 11S. 4 124. 7 121.7 108.2 11S.2 ISS.7 ISS.l 4S.1 SO. S 99.0 7S.8 128.4 103.8 139. 2 IIS.1 137. S 143.8 116.6 116.6 First instar 13.3 0. 4 9S.9 134.3 136.3 117.1 139.7 119.4 122.3 141. S 97.6 96. 2 -S. 6 0.0 8S.6 97.8 132.7 146. 2 147.S 111.7 142.8 162.6 109. 3 IOS.2-7.9 0.1 17. 0 0.2 S8.4 S2.1 IOS. 6 140.9 147. 6 126.1 121. S 122.8.., :r: '" (') z " ;;;: z '" z.., 0 ;:: 0 r- 8 u;..,
Volume 107 THE CANADIAN ENTOMOLOGIST 79 Table II. Number of fed first instar forest tent caterpillar larvae surviving at various temperature-duration combinations Duration of Temperature ec) cold treatment Controls in days -18-12 -7 -I 4 22 C 2 0 82 77 99 97 100 7 0 85 82 89 99 98 14 0 0 0 6 95 97 The larvae were less tolerant to cold temperatures after they hatched. Presumably, physiological changes within the caterpillar rapidly reduced their ability to supercool, since there was only 6 days difference between the pharate and hatched groups. The residual mean square in the ANOV A was quite large indicating a very high inherent variability of the data. Therefore, any test used to determine significant differences between means is very conservative. Significant differences between means of the control treatments (22 C) and means of the temperature treatments (Table I) are common. The means of the control colonies are decidedly lower four times and decidedly higher two times. The significant differences among the control means indicate that some unaccounted-for factor caused a certain variability. It did so in the field as well (Prentice 1954). It caused differences of up to 50 larvae per eggband. Adding the differences of 25 larvae (obtained by Tukey's method) needed for statistical significance gives a needed difference in means of 75 larvae. Therefore, it is necessary to have this difference before it can be assumed safely that it is caused by temperature and duration. When a difference of 75 larvae is used to judge significance, then significant differences among means occurred only at -18cC and -12 C and at 7 and 14 days. Feeding the larvae for 2 days after hatching did not seem to alter their cold tolerance. Large scale mortality occurred only below -12 C (Table 11). Of interest is the difference in behavior of larvae at the time of population collapse and of larvae subjected to cold temperatures. Smith and Raske (1968) reported that in the year of population collapse, larvae would lower en masse from the twigs when disturbed, travel up and down the tree trunk, not feed on the foliage, and would disappear abruptly. The behavior of larvae in these experiments was opposite of the above in the two characteristics that are comparable. Larvae would not lower from the egg masses and twigs when disturbed, but clung tightly to the substrate and readily ate foliage before and after the cold treatment. Larvae also survived low temperatures under field conditions. In 1967, eggbands were taken into the mountains, attached to aspen trees and allowed to hatch. After they hatched, temperatures of -9. 5 C on two successive nights caused no apparent mortality. Newly-hatched larvae survived at laboratory temperatures lower than those encountered during Alberta spring weather. They also survived low temperatures for longer durations than the usual cold periods of Alberta spring weather. They are therefore well adapted to an environment of highly variable spring weather, and it seems likely that temperature alone does not cause the near 100% reduction in population levels during the collapse of an outbreak. I would agree with Hanec (1966) that the high spring mortality of larvae in 1953 (Blais et al. 1955) was probably not due to direct killing by low temperatures (-6cC). He felt that the high mortality may have been caused by starvation of larvae in that all the aspen foliage had been killed. However, Smith and Raske (1968) concluded that starvation is likely to be an insignificant
80 THE CANADIAN ENTOMOLOGIST January 1975 mortality factor for whole populations because larvae can survive longer than it takes aspen stands to refoliate. Other indirect effects of low temperatures, such as the effect on larvae with sublethal dosages of virus diseases, are not known. It appears that the cause of the sudden collapse of whole populations in the first instar cannot be solely attributed to low temperature. Acknowledgments I thank A. W. Douglas of the Biometrics Research Services in Ottawa for his helpful advice on the data analyzed, and for the use of the computer in Ottawa. I thank B. M. Dahl and G. C. Bigalow for their assistance throughout the study, and several colleagues for reading the manuscript. References Blais, 1. R., R. M. Prentice, W. L. Sippell and D. R. Wallace. 1955. Effects of weather on the forest tent caterpillar Malacosoma disstria Hbn., in central Canada in the spring of 1953. Can. Ent. 87: 1-8. Davidson, A. G. and R. M. Prentice. 1968. Aspen insects and diseases. In 1. S. Maini and 1. H. Cayford (Eds.), Growth and utilization of poplars in Canada. Forestry Branch Publ. 1205. Gautreau, G. 1. 1964. Unhatched forest tent caterpillar eggbands in northern Alberta associated with late spring frost. Can. Dep. For.. Bi-mon. Prog. Rep. 20: 3. Hanec, W. 1966. Cold-hardiness in the forest tent caterpillar Malacosoma disstria Hbn. (Lasiocampidae, Lepidoptera). J.lnsect Physiol. 22: 1443-1449. Hodson, A. C. 1941. An ecological study of the forest tent caterpillar. Malacosoma disstria Hbn., in northern Minnesota. Tech. Bull. Minn. agric. Exp. Sm. No. 148. Ives, W. G. H. 1973. Heat units and outbreaks of the forest tent caterpillar, Malacosoma disstria (Lepidoptera: Lasiocampidae). Can. Ent. 105: 529-543. Lowe, V. H. 1899. The forest tent caterpillar. Bull. N.Y. agric. Exp. Sin. No. 159, pp. 35-60. Prentice, R. M. 1954. Decline of populations of the forest tent caterpillar in central Saskatchewan. Can. Dep. For.. Bi-mon. Prog. Rep. 10: 2. Prentice, R. M. and R. 1. Heron. 1956 (unpub.). Preliminary studies on the effect of temperature treatment on the development and hatching on forest tent caterpillar eggs. Div. For. Bioi. Interim Rep. 1955-1966. Winnipeg. (Cited with permission.) Can. Dep. Agric. Sci. Serv. Smith, G. 1. and A. G. Raske. 1968. Starvation experiments with first instar forest tent caterpillar larvae. Can. Dep. For.. Bi-mon. Res. Notes 24: 39. Sweetman, H. L. 1940. The value of hand control for the tent caterpillars, Malacosoma americana Fabr. and Malacosoma disstria Hbn. (Lasiocampidae, Lepidoptera). Can. Ent. 72: 245-250. Tothill, 1. D. 1920. Insect outbreaks and their causes. 50th A. Rep. ent. Soc. Ont.. 1919: 31-33. --- 1923. Notes on the outbreaks of spruce budworm, forest tent caterpillar, and larch sawfly in New Brunswick. Proc. Acad. ent. Soc.. 1922: 172-182. Wetzel, B. W., H. L. Kulman, and 1. A. Witter. 1973. Effects of cold temperatures on hatching of the forest tent caterpillar, Malacosoma disstria (Lepidoptera: Lasiocampidae). Can. Ent. 105: 1/45-1149. Witter, 1.. A. and H. M. Kulman. 1972. Mortality factors affecting eggs of the forest tent caterpillar, Malacosoma disstria (Lepidoptera: Lasiocampidae). Can. Ent. 104: 705-710; (Received 1 May 1974)