CRITICAL THERMAL LIMITS OF DESERT HONEY : POSSIBLE ECOLOGICAL IMPLICATIONS1

CAROL ANN ROLPH KAY2 AND WALTER G. WHITFORD Department of Biology, New Mexico State University, Las Cruces, New Mexico 88003 (Accepted 11/1/77)

Critical thermal maxima (CTMAX) and minima (CTMIN), approximate upper and lower lethal temperatures, and the temperature limits of activity were determined for workers of diurnal and nocturnal of . The CTMIN of diurnal species were between 11.2 and 12.0 C, and that of the nocturnal species was -0.4 C. All species had CTMAX between 43.3 and 48.4 C. Lower lethal tempera- tures of individuals of all species exposed for 2 h were below -5 C, and upper lethal temperatures were between 40 and 45 C. Diurnal species were active at 13-46 C air temperature (Ta) and 13-60 C soil surface temperature (Ts), and the nocturnal species was active at Ta and Ts of 2-30 C.

INTRODUCTION define ecological temperature-tolerance limits of their , that is, the Upper and lower lethal temperatures, which are known for many temperature limits for escape from a hostile environment (Fry 1967). Animals (Cloudsley-Thompson 1970), define phys- iological temperature-tolerance are limits, heated or cooled, usually at 1 C/min, until locomotion becomes impaired. Tech- that is, the temperature limits for sur- vival when exposed to high or lowniques tem- that approximate those used with peratures for a fixed period vertebratesof time have been employed in only (Fry 1967; Cloudsley-Thompson a few 1970). studies of invertebrates. Tempera- ture-tolerance limits equivalent to Unfortunately, data for different CTMAX and CTMIN have been deter- often cannot be compared because in- vestigators have not used a single mined expo- for several large flying insects and sure time (Cloudsley-Thompson referred 1970). to as threshold temperatures of Lethal temperature studies require cold largeor heat torpor (Heath et al. 1971; numbers of animals and considerable Heath, Wilkin, and Heath 1972; May 1976). Recently Schumacher and Whit- time. Investigators studying reptiles and ford (1974), Whitford and Ettershank amphibians commonly use parameters known as critical thermal maxima (1975), and Whitford, Kay, and Schu- (CTMAX) and minima (CTMIN) macher to (1975) have measured CTMAX and CTMIN of several species of desert 1 We wish to thank Roy R. Snelling for identifying ants. ants for the study. The work was part of a Ph.D. During our investigations, CTMAX dissertation completed at New Mexico State University by C. A. R. Kay. Portions of the study and CTMIN were measured for five were funded through NSF grant no. GB-15886 speciesto of honey ants (Myrmecocystus the US/IBP Desert Biome Program, Jornada Validation Site. romainei Cole, M. depilis Forel, M. mimicus Wheeler, M. mexicanus Wes- 2Present Address: Department of Biology, California State University, Los Angeles, California mael, and M. navajo Wheeler) all com- 90032. monly encountered in the deserts of southern New Mexico (Snelling 1976). @ 1978 by The University of Chicago. 0031-935X/ 78/5102-7760$00.81 Approximate upper and lower lethal 206

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms THERMAL LIMITS OF ANTS 207

LETHAL TEMPERATURES temperatures of some species were ob- tained for comparison to Lethal critical temperature thermal studies were con- limits. In the field, determinations ducted in October 1972 wereon workers of made of the temperature M. romainei, ranges M. depilis, over and M. mexi- which foraging occurred. canus. TenWe workers hoped of a tospecies were detect possible physiological placed in eachadaptations of 24 30-cc vials on brass that allow diurnal Myrmecocystus screening. Twelve vials contained(M. moist romainei, depilis, and mimicus) sand below the to screening forage to humidify at the high temperatures the air, characteristic and 12 contained anhydrous of their desert environments and to silica gel below the screening as a drying investigate the physiological correlates agent. Four vials of each species (two of nocturnality of some species containing(M. moist sand and two contain- mexicanus and navajo). ing silica gel) were placed in a controlled- temperature incubator at one of six MATERIAL AND METHODS temperatures: -5, 5, 15, 35, 40, or 45 C. All ants were collected on the New After 2 h the vials were removed from Mexico State University Experimental the incubators and the ants were given Ranch, 30-40 km NNE of Las Cruces, a 1-h period for recovery. The number Dofia Ana County, New Mexico. Worker of living and dead ants in each vial was ants for laboratory studies were aspi- then recorded. rated from the soil surface at their nest CRITICAL THERMAL LIMITS entrances. Males and alate queens were aspirated from galleries less than 20 cm The apparatus used during measure- beneath the soil surface. The insects ments of critical thermal limits consisted were transported and maintained ofin a 250-ml beaker containing an 8-mm layer of white gypsum sand. A thermistor plastic boxes on moist paper toweling. probe and telethermometer were used The laboratory remained at about 24 C to measure the temperature at about and 20% relative humidity throughout 1 mm above the sand surface. It was not the investigations. possible to measure the temperature of

FIELD OBSERVATIONS the individual ants because of their

Observations of workers entering small size and and high activity, but the Ta measured was the temperature to which leaving nests of M. romainei, M. depilis, M. mimicus, and M. mexicanus the ants were were exposed. During measure- ments of CTMIN the beaker was cooled made at intervals during a 24-h period. in a container of chipped ice and water With each observation of nest activity, (diurnal Myrmecocystus spp.) or chipped air temperature (Ta) at 0.5 drycm iceabove (M. mexicanus). An electric the substrate and soil surface heating tempera- mantle (flask warmer) with a ture (Ts) were measured using variable a rheostattele- was used to heat the thermometer with a thermistor beaker probe. during measurements of CTMAX. Soil surface temperature was measured At the beginning of each measurement with the probe placed on the soilperiod, surface five ants were placed in the and covered with a thin layer beaker of atsoil. room temperature, and the Studies were repeated at about beaker 2-wk was heated or cooled at about intervals from August 1972 1to C/min. August Each was observed until 1973. locomotion became impaired. The tem-

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms 208 CAROL ANN ROLPH KAY AND WALTER G. WHITFORD perature at which locomotion the of soil each only a few seconds, climbing ant ceased but before all movement every object encountered as they moved ceased was considered to be the CTMIN between the nest and the plants on or CTMAX of that individual. Pilot which they were foraging. Diurnal ants studies using M. depilis workers did had not forage at night but often engaged shown that the rate of temperature in activities related to nest maintenance. increase was an important consideration Workers of M. mexicanus were active at since the CTMAX of ants heated at Ta and Ts of 2-30 C. Myrmecocystus 2 C/min averaged 2 C lower than mexicanus that usually ceased activity at of ants heated at 1 C/min. During dawn and resumed activity at dark, but critical thermal minimum experiments, occasionally on rainy days a few workers a ring of cardboard 1 cm high was wereplaced above ground 1-2 h after dawn or against the sides of the beaker because 0.5-1 h before dark. Activity of all contact with the cold glass caused species the decreased as temperatures ap- ants to immediately become torpid. proached the upper and lower activity In April, June, and July 1973, limits. mea- All species except M. romainei surements of seasonal effects on CTMAX foraged throughout the year when tem- and CTMIN were made on small, peratures permitted. Myrmecocystus ro- medium, and large workers of five mainei ceased activity in mid-October species of Myrmecocystus. Experiments and resumed activity in mid-March. were performed within 1-2 h of collec- LETHAL TEMPERATURES tion. In April 1973, experiments were conducted to examine acclimation of the After exposure for 2 h to temperatures between 5 and 35 C, two M. romainei CTMAX of workers of M. depilis. Ants workers and one M. depilis worker were were held in the laboratory for about 12 h and then placed in controlled- dead. No pattern in the deaths could be detected related to either temperature temperature incubators at 0, 5, 15, 25, or humidity, so the deaths were probably or 35 C on moist sand in plastic con- the result of injuries. After 2 h at -5 C, tainers with dilute honey as food. Ants were removed from the incubators after about 15% of the ants of all species were dead in dry air, but about 25% of the 0, 12, 25, 30, or 35 h exposure, and CTMAX were measured. ants were dead in wet air, probably because water condensed and froze on Means and .95 confidence intervals were calculated for all data sets. The thet'- ants at high humidities. After 2 h statistic was used to examine differences at 40 C death rates were 0% for M. depilis in both wet and dry air, 0% for between means with unequal variances M. romainei in wet air and 25% in dry (Steel and Torrie 1960). The .05 proba- air, and 15% for M. mexicanus in wet bility level was used for tests of sig- nificance. air and 70% in dry air. Ants of all species were weakened by the 40 C RESULTS exposure in dry air (but not in wet air), FIELD OBSERVATIONS and many died later, probably as a Workers of Myrmecocystus resultromainei, of dehydration. A plastic box of M. depilis, and M. mimicus wetwere sand seen containing M. depilis workers above ground at Ta of 13-46 was C maintained and Ts at 40 C for 12 h without of 13-60 C throughout a day.casualties. At At Ts 45 C all ants died in 2 h. above about 50 C workers remained on Therefore, the lower lethal temperatures

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms THERMAL LIMITS OF ANTS 209

(defined as the temperature smallest species, resulting M. romainei, in had the 50% mortality) of all lowestspecies CTMAX. were Larger below workers of a -5 C when exposed in species either usually wet had orslightly dry higher air for 2 h. The upper CTMAX lethal than smaller tempera- workers (depilis, tures were between 40 and 45 C for all mean CTMAX in July = 1.8 C higher; species in wet air and for all species mimicus, 1.5 C higher; romainei, 0.3 C except M. mexicanus in dry air. higher; mexicanus and navajo, not mea- sured). Queens of M. romainei had a CRITICAL THERMAL LIMITS significantly higher CTMAX than work- In July the CTMIN of workers ers. of Males the of M. depilis had a lower nocturnal M. mexicanus (mean live CTMAX than workers, but not signifi- weight = 9.2 + 1.1 mg) was much lower cantly so. than that of diurnal M. mimicus (4.7 + The CTMAX of workers of M. ro- 1.5 mg), M. depilis (3.6 + 0.6 mg), or mainei and M. depilis (but not M. M. romainei (3.2 + 0.7 mg) (table 1). mimicus or M. mexicanus) were signifi- Workers of the largest diurnal species, cantly higher in July than in April. M. mimicus, had the lowest CTMIN, Workers of all diurnal species, except and workers of the smallest species, M. M. romainei in April, had higher romainei, had the highest CTMIN. CTMAX than workers of nocturnal Larger workers of a species usually had species. During the high temperature slightly lower CTMIN than smaller exposures, workers of diurnal Myrmeco- workers (depilis, mean CTMIN = 2.5 C cystus spp. elevated their bodies from lower; mimicus, 2.4 C lower; romainei, the substrate, moved rapidly, and tried not measured; mexicanus, 0.3 C higher). to climb every object encountered. Noc- Queens of M. romainei had a significantly turnal workers did not display this lower CTMIN than workers. behavior pattern. The CTMAX of workers of the two The CTMAX of temperature-accli- nocturnal species, M. mexicanus and M.mated workers of M. depilis are shown navajo, were not significantly different. in table 2. The mean of the measure- Of the diurnal species, workers of thements made at the beginning of the largest species, M. mimicus, had the acclimation periods (acclimation time = highest CTMAX, and workers of the 0) was considered as the control value.

TABLE 1

CRITICAL THERMAL MAXIMA AND MINIMA OF FIELD-ACCLIMATIZED ANTS OF THE "MYRMECOCYSTUS"

SPRING SUMMER FORAGING AND SPECIES CASTE CTMAX N CTMAX N CTMIN N

Diurnal: M. romainei ... Worker 43.3+ .8a 31 46.1+ .4b 26 12.0+.3e 20 M. romainei ... Queen 47.2+ .6c 10 8.4 +.6 10 M. depilis ..... Worker 46.2+ .8b 35 47.4+ .7c,d 26 11.6+.5e,f 20 M. depilis ..... Male 42.5+6.2a,b,c,d 5 M. mimicus..... Worker 48.4+ .9d 25 47.7+ .9c,d 20 11.2+.5f 20 Nocturnal: M. mexicanus... Worker 44.7+5.0a,b,c,d 3 44.2+ .9 20 -0.4 . 7 18 M. navajo ..... Worker 43.7 2.2a 5

NoTE.--CTMAX = mean critical thermal maximum + .95 confidence interval in +C; CTMIN = mean critical thermal mini- mum + .95 confidence interval in +C. Means followed by the same letter are not significantly different (P < .05).

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms 210 CAROL ANN ROLPH KAY AND WALTER G. WHITFORD

TABLE 2

CRITICAL THERMAL MAXIMA OF TEMPERATURE-ACCLIMATED WORKERS OF "MYRMECOCYSTUS DEPILIS"

ACCLIMATION TEMPERATURE (OC) 5 15 25 35

HoURsS CTMAXb N CTMAX N CTMAX N CTMAX N 0...... 47.2++1.7 10 47.2+1.7 10 47.2+1.7 10 47.2+1.7 10 12...... 48.0+1.5 5 30...... 25...... 47.8+ .546.4+ 5 47.8+ .7 5.8 47.0+1.3 10 548.0+ 51.4+1.5c 5.9 5 47.4+ .9 10 49.2+ .8o 10 35...... 48.3+ .7 5 47.2+1.2 10 49.2+1.7 5 51.9+1.5c 5

a Hours of acclimation. b Mean critical thermal maximum + .95 confidence interval in +C. o Significantly different from CTMAX at h = 0 (P <.05).

Only CTMAX of workers at 35 C for diurnal or crepuscular desert ants studied 25 h or longer were significantly different by Schumacher and Whitford (1974) or from the control. At 35 C, the CTMAX Whitford and Ettershank (1975) had increased with time in acclimation and CTMIN below 3.6 C. The CTMIN of all appeared to stabilize after about 30 h. of the ants that have been tested, but especially of M. mexicanus, are lower DISCUSSION than the cold torpor temperatures of Measurement of CTMIN and several large flying insects (Heath et al. CTMAX proved to be a simple 1971, 1972). and ecologically meaningful method The ofCTMIN ob- of diurnal Myrmeco- taining temperature-tolerance cystus limits spp. of were 1.5-8.4 C higher than Myrmecocystus spp. The CTMIN those of speciesof of Pogonomyrmex, Novo- workers of diurnal and nocturnal messor, species Formica, and Trachymyrmex of Myrmecocystus were within studied2 C of bythe Schumacher and Whitford minimum foraging temperatures, (1974) andsug- Whitford and Ettershank gesting that the ants do not(1975). cease Those investigators chose as the surface activity until the temperature CTMIN (or CTMAX) that temperature drops to near the point at which at which locomo- all movement ceased. Bertram tion would be impaired. However, (1935) determinedlight both a temperature intensity apparently partially of locomotor deter- impairment and a tempera- mined the type of activity ture in of which cessation of movement for 49 diurnal workers engaged (foraging different insectsvs. and found an average nest maintenance). Lower lethal difference tem- between the two temperatures peratures of Myrmecocystus ofspp. 4.1 C.were We felt that the temperature of much lower than foraging limits, locomotor sup- dysfunction was the tempera- porting Dely6's (1968) hypothesis ture at whichthat ants could no longer lower lethal temperatures are return of littleto their nests, and therefore ecological significance to desert the temperature ants. limit of ecological sig- Mellanby (1940) found that nificance. an arctic ant, Formica semirufa, had a chillThe upper coma temperature limits of activ- temperature of 1-3 C, comparable ity of diurnal to Myrmecocystus spp. were the CTMIN of -0.4 C found for close to their CTMAX but slightly workers of M. mexicanus. None of the higher than their upper lethal tempera-

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms THERMAL LIMITS OF ANTS 211

tures. However, the probablyupper have lethal thoracic tem- temperatures peratures represent very 2-h close exposures to the temperatures to measured, constant temperatures. but we Workers cannot rule outof the M. possibility mexicanus, on the other that the hand,largest ants ceased heated or cooled at surface activity at temperatures slightly lower rates. aboutThermal inertia 15 C below their CTMAX. Workers of may also have accounted for the higher the nocturnal species, M. navajo, CTMAX in and lower CTMIN of M. southern Utah were observed foraging romainei queens, which had weights only at temperatures lower than 25 about C, 30 times those of workers. The about 19 C below their CTMAX (Fau- slightly lower CTMAX of males of M. tin 1946). The upper temperature-toler- depilis (range, 37.1-47.7 C) could not ance limits probably determine the up-be attributed to size differences because per foraging limits of diurnal species males were about the same size as but not of nocturnal species. Light medium-sized workers. Males die quick- intensity may be the most important ly in the laboratory, and those with low factor controlling foraging of nocturnal CTMAX may have been weakened or species (Fautin 1946; Cazier and Sta-injured prior to experiments. The small tham 1962). Thus, nocturnal Myrmeco- effect of size on the temperature-toler- cystus spp. probably are never exposed ance limits of worker ants is probably of to dangerously high temperatures. This little ecological significance since smaller suggests that nocturnal species have antsnot predominate in the foraging popula- evolved the behavioral responses to hightions of most honey-ant species. temperatures as noted for diurnal Myr- Seasonal effects on the temperature- mecocystus spp. during our experiments tolerance limits have been demonstrated (elevation of the body and other attempts for only a few arthropods (Edney 1964; to escape from the hot environment). Cloudsley-Thompson 1969; May 1976). At high temperatures, the tempera- Seasonal changes in the CTMAX were tures of locomotor impairment of Myr- small for all species of Myrmecocystus mecocystus spp. were about 1 C below except M. romainei and probably of the temperatures of cessation of move- little ecological significance. The ment. Species of Myrmecocystus CTMAXhad of workers of M. romainei were CTMAX comparable with that of lower F. in April, immediately following perpilosa (45.2 C), an ant in the same winter dormancy, than in July. subfamily that also forages in the plant Exposures of 24 h or longer to sub- canopy (Schumacher and Whitford 1974), lethal high temperatures increase the and 3.3-10.5 C lower than those of upper lethal temperatures of many in- species of Pogonomyrmex and Novomes- sects, but acclimation to lower tempera- sor, ants that forage primarily ontures the usually has no effect (Cloudsley- soil surface (Whitford and Ettershank Thompson 1970). The CTMAX of M. 1975; Whitford et al. 1975). depilis workers responded to tempera- It was impossible to determine whether ture acclimation in a similar way. We small differences in CTMAX and also found that the CTMAX of 15 CTMIN among workers of different workers of M. romainei that had been sizes and among diurnal species held atof 0 C for 10 days was 46.5 + 1.2 C, Myrmecocystus were the result ofnot actual significantly different from the physiological differences or of differences CTMAX of field-acclimatized workers in thermal inertia. These small ants in July. The CTMAX of field-acclima-

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms 212 CAROL ANN ROLPH KAY AND WALTER G. WHITFORD tized workers was never elevated (tablethose of many insects that are not 1), even though air temperatures oftenadapted to desert life (Cloudsley-Thomp- remained above 35 C while the ants son 1970). Dely6 (1968) found that most were being transported to the laboratory. Sahara Desert ants do not have unusual- Foragers would not be exposed to tem-ly high upper lethal temperatures. A peratures of 35 C or above for more few than desert ants, notably species of 10 h in their natural environment, Cataglyphus sug- (Dely6 1968) and species of gesting that acclimation to sublethal Pogonomyrmex (Whitford and Etter- high temperatures seldom occurs shank in 1975), tolerate higher tempera- nature. tures. Investigators have hypothesized CONCLUSIONS that physiological differences among des- ert ants may be related to the "feeding can remain guild" (Whitford et al. 1975) or foraging active at much lower temperatures than behavior (Dely6 1968) of the ants. For most desert ants (Dely6 1968; Schu- example, ants such as Pogonomyrmex macher and Whitford 1974; Whitford and spp. that collect seeds or other relatively Ettershank 1975). The low temperature- dry food on the ground may have tolerance limit of this species probably evolved physiological tolerances to high- reflects an adaptation to nocturnal for- er temperatures and vapor-pressure defi- aging since temperatures below 15 C cits than ants such as Myrmecocystus occur frequently at night in warm spp. that spend much of their time in desert areas. The ability to forage at the plant canopy collecting nectar. De- temperatures below 10 C allows spite the M. fact that diurnal species of mexicanus to be active at night through- Myrmecocystus cannot tolerate unusually out much of the year in southern high temperatures,New workers forage at Mexico. In the Las Cruces area, the midday in the summer when few other diurnal Myrmecocystus spp. can also ants are active. Although high tempera- forage throughout the year because mid- tures preclude activity on the soil sur- day temperatures are well above their face, foragers can remain in the plant CTMIN during most of the winter. canopy during the hottest portion of the The upper temperature-tolerance lim- day and return to the nest when the its of Myrmecocystus spp. are similar to temperature drops.

LITERATURE CITED

BERTRAM, G. C. L. 1935. The low temperature DELYi, G. 1968. Recherches sur l'&cologie, la limit of activity of arctic insects. J. Anim. physiologie Ecol. et l'6thologie des Fourmis du 4:35-42. Sahara. Theses pr6senties B la Facult6 des CAZIER, M. A., and M. STATHAM. 1962. The Sciencesbe- de l'Universit6 d'Aix-Marseille. havior and habits of the myrmecophilous EDNEY, E. B. 1964. Acclimation to temperature in scarab Cremastocheilus stathamae Cazier with terrestrial isopods. Physiol. Zool. 37:364-377. notes on other species (Coleoptera: Scarabeidae). FAUTIN, R. W. 1946. Biotic communities of the J. New York Entomol. Soc. 70:125-149. northern desert shrub biome in western Utah. CLOUDSLEY-THOMPSON, J. L. 1969. Acclimation, Ecol. Monogr. 16:251-310. water and temperature relations of the FRY,wood F. E. J. 1967. Responses of vertebrate poikilo- lice Metoponorthus pruinosus and Periscyphis therms to temperature. Pages 375-409 in jannonei in the Sudan. J. Zool. (London) 158: A. H. ROSE, ed. Thermobiology. Academic 267-276. Press, New York. - -. 1970. Terrestrial invertebrates. Pages 15-77 HEATH, J. E., J. L. HANEGAN, P. J. WILKIN, and in G. C. WHITTOW, ed. Comparative physiology M. S. HEATH. 1971. Adaptations of the thermal of thermoregulation. Vol. 1. Academic Press, responses of insects. Amer. Zool. 11:147-158. New York. HEATH, J. E., P. J. WILKIN, and M. S. HEATH. 1972.

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms THERMAL LIMITS OF ANTS 213

Temperature responses SNELLING, of theR. R. 1976. cactus A revision dodger of the honey ants Cacama valvata (Homoptera, genus Myrmecocystus Cicadidae). (: Physiol. Formi- Zool. 45:238-246. cidae). Sci. Bull. Los Angeles County Nat. Hist. MAY, M. L. 1976. Thermoregulation and adaptation Mus. 24:1-163. to temperature in dragonflies (Odonata: STEEL, An- R. G. D., and J. H. TORRIE. 1960. Principles isoptera). Ecol. Monogr. 46:1-32. and procedures of statistics. McGraw-Hill, MELLANBY, K. 1940. The activity of certain New York. arctic insects at low temperature. J. Anim. WHITFORD, W. G., and G. ETTERSHANK. 1975. Ecol. 9:296-301. Factors effecting foraging activity in Chihuahuan SCHUMACHER, A., and W. G. WHITFORD. 1974. desertThe harvester ants. Environmental Entomol. foraging ecology of two species of Chihuahuan 4:689-696. desert ants: Formica perpilosa and Trachymyr- WHITFORD, W. G., C. A. KAY, and A. M. SCHU- mex smithi neomexicanus (Hymenoptera: Formi- MACHER. 1975. Water loss in Chihuahuan desert cidae). Insectes Soc. 21:317-334. ants. Physiol. Zool. 48:390-397.

This content downloaded from 128.123.44.24 on Fri, 23 Apr 2021 19:09:42 UTC All use subject to https://about.jstor.org/terms