Journal of Insect Physiology 47 (2001) 1021–1027www.elsevier.com/locate/jinsphys

Enhanced drought tolerance of a soil-dwelling springtail by pre-acclimation to a mild drought stress

H. Sjursen*, M. Bayley 1, M. HolmstrupDepartment of Terrestrial Ecology, National Environmental Research Institute, Vejlsøvej 25, PO Box 314, DK-8600 Silkeborg, Denmark

Received 25 October 2000; accepted 1 February 2001

Abstract

The springtailFolsomia candida has a highly permeable cuticle, but is able to survive several weeks at 98.2%RH. This correspondsto a water potential deficit of about 17 bars between the environment and the normal osmotic pressure of the body fluids of thisanimal. Recent studies have shown a water vapour absorption mechanism by accumulation of sugars and polyols (SP) inF. candida,which explains how this species can survive dehydrating conditions. In the present study, adultF. candida were pre-acclimated at98.2%RH to induce the accumulation of SP, and were subsequently exposed for additional desiccating conditions from 98 to94%RH. Activity level, water content, osmotic pressure of body fluids and SP composition were investigated. After the desiccationperiod, the animals were rehydrated at 100%RH and survival was assessed. The results showed thatF. candida survived a moresevere drought stress when it had been pre-acclimated to 98.2%RH before exposure to lower humidity. This species was able tomaintain hyperosmosity to the surroundings at 95.5%RH, suggesting that it can absorb water vapour down to this limit. Below thislimit, trehalose levels increased while myo-inositol levels decreased. We propose that this is a change of survival strategy whereF. candida at mild desiccation levels seek to retain water by colligative means (remain hyperosmotic), but at severe desiccationlevels switches to an anhydrobiotic strategy. 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Springtail; Desiccation; Acclimation; Water content; Sugar; Polyol

1. Introduction

Springtails (Collembola) are among the most wide-spread terrestrial arthropods, with a global distribution.They are found in a wide range of habitats, includingpolar and alpine areas, sea shores and deserts (Yosii,1966; Leinaas and Ambrose, 1992; Suhardjono andGreenslade, 1994). Vannier (1983) has classified spring-tails into three groups: when exposed to desiccating con-ditions, the Type I or hygrophilic species show no con-trol over water loss. The Type II (mesophilic) and TypeIII (xerophilic) species show increasing control overwater loss. Hygrophilic springtails are generally vulner-able to desiccation because of their small size and per-meable cuticle. Availability of soil water is hence

* Corresponding author. Tel.:+45-8920-1400; fax:+45-8920-1413.E-mail address: hes@dmu.dk (H. Sjursen).

1 Present address. Department of Zoology, Institute of BiologicalSciences, University of Aarhus, Universitetsparken, Building 135, DK-8000, Aarhus C, Denmark.

0022-1910/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved.PII: S0022-1910 (01)00078-6

expected to play an important role in determining speciesdistribution, spatial patterns, community compositionand productivity. Soil invertebrate communities arehighly variable within their habitat and often reflect themoisture status of their microhabitat and the more gen-eral soil water characteristics (Hodkinson et al., 1999).However, it is well-known that it is not the soil wetness,but rather the soil water potential that controls theresponse of soil invertebrates to changing moisture con-ditions (Collis-George, 1959). Few studies haveaddressed the water availability in terms of soil waterpotential of the microhabitat of soil animals. When asoil-dwelling springtail is exposed to drought in its natu-ral environment, it will most likely be subjected to agradually increasing drought stress due to the bufferingcapacity of the soil to water loss. Experiments address-ing drought tolerance would therefore possibly havemore relevance if such conditions were considered,rather than exposing the springtails directly to a targetdrought level.

Folsomia candida (Willem) is a common soil-dwell-

1022 H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

ing springtail with a highly permeable cuticle, but is ableto survive relative humidities (RH) down to 97% formore than a week, even when directly exposed to theselevels (Holmstrup, 1997). In a recent study, Bayley andHolmstrup (1999) described a water vapour absorptionmechanism in F. candida which explains how this spec-ies can survive dehydration and be active below the per-manent wilting point of plants (i.e. 98.9%RH). F. can-dida actively increases its osmotic pressure and becomeshyperosmotic to its surroundings, by synthesising myo-inositol and glucose. This enables the animals to pass-ively absorb water molecules from an unsaturated atmos-phere and avoid fatal dehydration. This mechanism ispotentially a survival strategy during summer droughtperiods.

In the present study, adult F. candida were pre-acclimated at 98.2%RH to increase the sugar and polyol(SP) concentration, and were subsequently exposed todesiccating conditions from 98 to 94%RH (the lethallimit of directly exposed F. candida specimens isapproximately 96%RH). Activity level, water content,osmolality and SP composition in the desiccated animalswere investigated. This enabled us to test the hypothesisthat animals which have been allowed to raise their SPconcentration during a pre-acclimation period will sur-vive desiccating conditions better than animals whichhave not been pre-acclimated.

2. Materials and methods

2.1. Test animals

Specimens of F. candida for these experiments wereobtained from a parthenogenetic culture held in a climateroom at 20±1°C on moist plaster of Paris/charcoal andfed dry baker’s yeast, as described by Krogh (1995). Thetaxonomic key of Fjellberg (1980) was followed foridentification. All test animals were female adults ofapproximately the same age (�40 days). F. candida inthis culture reach sexual maturity at the age 16–19 days(Krogh, 1995). The average length and fresh weight ofthe animals were approximately 2 mm and 150 µg. Theculture was fed and watered regularly to ensure animalsof good health and with minimal previous drought stress.Prior to the experiments the animals were starved for24 h on moist plaster of Paris/charcoal for clearing ofthe gut, but were given access to free water (drinking orabsorption via ventral tube from the film of water coatingthe plaster of Paris) to ensure that they were fullyhydrated (average water content 1.6 g/g DW) before theoutset of the experiments.

2.2. Pre-acclimation and treatment

The springtails were pre-acclimated prior to the actualtreatment. Pre-acclimation took place at 20±0.1°C for six

days. The animals were placed in open-top plastic sam-ple vials (4 cm high, 2 cm diameter) in the centre of a200 ml plastic cup containing 25 ml NaCl solution(31.6 g/l), sealed with a tightly fitting plastic lid. The airin this small closed system rapidly equilibrates with thesalt solution (following Raoult’s law), providing an RHof 98.2% (Lang, 1967;Weast, 1989). Control series (nopre-acclimation) were placed in similar chambers con-taining distilled H2O, providing 100%RH. Following thepre-acclimation/no pre-acclimation period, all specimenswere exposed to controlled drought treatment for 6 days,by placing them in similar plastic chambers containingNaCl solutions of known concentrations. Table 1 showsthe concentrations used, and the resulting RH and waterpotential (bar) in the test chambers.

2.3. Activity and survival

Four replicates containing 10 pre-acclimated and 4containing 10 control specimens were placed at 98.2,97.5, 97.0, 96.5, 96.0, 95.5, 95.0, 94.5, 94.0, 93.5 and93.0%RH. After 6 days, activity was assessed by classi-fying the individual animals in each replicate as either0 (inactive), 1 (movement of leg or antennae) or 2 (ableto walk) after tapping the container gently to stimulatemovement, and a total activity ranking was calculatedfor each sample. The animals were subsequently rehy-drated for 48 h in the same type of containers at100%RH, whereafter survival was assessed. The animalswere considered alive if they showed co-ordinated walk-ing behaviour after rehydration. The difference betweenactivity and survival in both pre-acclimated and controlgroups, and the variance in survival between the twogroups were analysed with logistic regression.

2.4. Osmotic pressure of body fluids

Four replicates of 40 pre-acclimated and 4 of 40 con-trol specimens were placed at 98.2, 97.5, 97.0, 96.5,

Table 1NaCl concentrations with the corresponding relative humidity (RH)and water potential (bar)

g NaCl/l dH2O RH (%) Water potential(bar)

0 100 031.60 98.2 �24.753.80 97.0 �34.862.50 96.5 �42.471.20 96.0 �50.181.20 95.5 �57.990.34 95.0 �65.899.40 94.5 �73.7

108.46 94.0 �89.7117.52 93.5 �99.1126.57 93.0 �107.3

1023H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

96.0, 95.5, 95.0 and 94.5%RH for 6 days (controls downto 96.5%RH only, due to difficulties in determiningwhether the springtails were alive or not below this RH).Individual specimens displaying activity level 1 or 2were considered to be alive after treatment, and wereused for osmometry measurements. A sample of 10–15specimens were taken from each replicate and osmoticpressure of body fluids was measured using a WescorDewpoint Microvoltmeter (Wescor, Logan, Utah) asdescribed by Bayley and Holmstrup (1999). Differencesin osmotic pressure between pre-acclimated and controlgroups were analysed with one-way ANOVA.

2.5. Water content and SP

Four replicates of 40 pre-acclimated and 4 of 40 con-trol specimens were placed at 98.2, 97.5, 97.0, 96.5,96.0, 95.5, 95.0 and 94.5%RH for 6 days (controls downto 96.5%RH only). Individual specimens displayingactivity level 1 or 2 were considered to be alive aftertreatment, and were used for water content and SPmeasurements. A sample of 10 specimens were takenfrom each replicate and weighed on a Cahn 4700 auto-matic electrobalance (accuracy ±1 µg), and dry weightwas determined by drying the samples completely at60°C for 24 h and weighing them again. Mean watercontent at the end of the treatment was expressed as gwater/g dry weight (g/g DW). The dried samples werestored at �80°C and later extracted for analysis of SP.Extraction and high-pressure liquid chromatography(HPLC) was performed to quantify the SPs in thesamples as described by Bayley and Holmstrup (1999).Sorbitol (Supelco — standard R422500) was used asinternal standard, and glucose (d(+)glucose, Supelco —standard 4-7249), myo-inositol (Fluka Biochemica) andtrehalose (Merck 1.08353) were used for identificationand quantitative calibration. Variance in water contentand SP concentrations between pre-acclimated and con-trol groups was analysed with one-way ANOVA.

3. Results

3.1. Survival

There was a highly significant difference in survival(p�0.0001) between pre-acclimated and control F. can-dida (Fig. 1(a) and (b), Table 2), with the pre-acclimatedspringtails showing a higher survival on each treatmentlevel. Pre-acclimated animals survived RH down to94.5%, whereas only a few control specimens surviveddown to 95.5%RH. This corresponded to an increase inthe tolerable water potential from about �57 to �73 bar(Table 1) when springtails had been pre-acclimated.

Fig. 1. Percent of F. candida in each activity group before rehy-dration (solid, walking; hatched, movement of leg or antenna; open,no movement), and mean (±SD) survival after rehydration (circles),in (a) pre-acclimated and (b) control springtails exposed to differenttreatments (%RH).

3.2. Activity

There was a good agreement between survival afterrehydration and the sum of activity levels 1 and 2 beforerehydration (Fig. 1(a) and (b), Table 2), and there wasno significant difference between the two in either pre-acclimated or control springtails. This was an importantobservation that validated the assumption that survivalof dehydrated springtails could be assessed without rehy-dration. It was important to have such a criterion whenmeasuring parameters on dehydrated, living springtails.

3.3. Water content, SP and osmotic pressure

The water contents of pre-acclimated F. candida weregenerally higher than that of controls at the humiditylevels below 100%RH (Fig. 2), suggesting that pre-acclimated springtails were able to reduce water lossand/or increase the water vapour absorption compared

1024 H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

Table 2Statistical analyses of differences in survival and SP concentrationbetween groups (pre-acclimated and control), and between survival andthe sum of activity levels 1 and 2 for both groups of F. candidaexposed to different treatments (RH). Dev=deviance, df=degrees offreedom, c=chi square, SS=sum of squares

Parameter Test Test statistics p-value

Survival Logistic Dev=84.24, �0.0001regression df=57,between groups c=60.8498

Survival vs. Logisticactivity level regression1+2

pre-acclimated Dev=2.90, df=2, 0.9990c=0.0001

not pre- c=0.0025 0.9602acclimated

Glucose One-way F=7.23, df=9, 0.3734ANOVA R2=0.62,between groups SS=53.41

Trehalose One-way F=4.75, df=9, 0.2298ANOVA R2=0.51,between groups SS=10.18

Myo-inositol One-way F=16.21, df=9, 0.9326ANOVA R2=0.78,between groups SS=0.58

Fig. 2. Mean (±SD) water content (g/g dry weight) in pre-acclimated(filled circles) and control (open circles) F. candida exposed to differ-ent treatments (%RH).

to the control group. However, this difference was onlystatistically significant at 96 and 96.5%RH (paired T-test, p�0.01). The missing data in the control group atthe lower drought levels was due to a difficulty inobtaining living specimens from these drought regimes.When the survival rate of each treatment was plottedagainst the total water content it was evident that pre-acclimated animals were able to tolerate lower bodywater contents than control animals (Fig. 3). Forinstance, pre-acclimated springtails with a final watercontent of 0.8 g/g DW had a survival rate of 70%,whereas only 25% of control animals survived the same

Fig. 3. Correlation of mean water content (g/g dry weight) and mean(±SD) survival in pre-acclimated (filled circles) and control (opencircles) F. candida exposed to different treatments (%RH).

water content. This was equivalent to only 50% of thewater content of fully hydrated specimens. Droughtexposed F. candida accumulated myo-inositol and glu-cose as the only SP compounds at humidities above97%RH (Fig. 4). Below 97%RH trehalose was alsoaccumulated, and the concentration was correlated withincreasing drought stress. The accumulation of trehalosewas apparently occurring at the expense of myo-inositoland, to a lesser degree, glucose. The total concentrationof each SP in the pre-acclimated animals was signifi-cantly higher than in controls at treatments between 98.2and 97.5%RH (Fig. 4, paired T-test, p�0.02). For bothgroups there was a slight increase in the total SP concen-tration between 98.2 and 97.5%RH. At RHs below thislevel the total SP concentration remained constant in thecontrol group down to the lowest treatment, but in thepre-acclimated animals there was an increase below96%RH, when trehalose accumulation increased. Acomparison between the osmotic pressure of the spring-tails and the surrounding water potential for the different

Fig. 4. Mean (±SD) sugar and polyol concentrations in pre-acclimated (filled symbols) and control (open symbols) F. candidaexposed to different treatments (%RH). Circles, trehalose; triangles,glucose; squares, myo-inositol.

1025H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

treatments (Fig. 5) showed that at all treatments in bothgroups the osmotic pressure of the animals was higherthan the surrounding water potential, i.e. all springtailsremained hyperosmotic to the surroundings. The osmoticpressure differences between pre-acclimated and controlgroups were statistically significant at 98.2 and97.5%RH (paired T-test, p�0.01). The osmotic pressureof the initial solutes and induced SPs could in bothgroups explain a significant part of the measured osmoticpressure (Fig. 5).

4. Discussion

Drought tolerance has been studied in many differenttaxa of arthropods, and several strategies for survivingthe exposure to dry environments have been discussed(Hadley, 1994). In the terminology of Hadley (1994)hygrophilic springtails (including F. candida) mainly usetwo strategies, behavioural avoidance of drought(Verhoef and Nagelkerke, 1977; Verhoef and Van Selm,1983) and water sorption dependence (Bayley andHolmstrup, 1999). In a previous study we have investi-gated the tolerance limits of acute drought stress(Holmstrup, 1997). However, the soil environment is ahighly buffered habitat with respect to desiccation. It isthus likely that acclimation to a mild desiccation cantake place during summer droughts. The present studyshows that this type of buffering, with a pre-acclimationperiod where the springtails are exposed to a milddrought stress prior to a more severe drought stress, willsignificantly increase survival. The pre-acclimatedspringtails had a significantly higher survival at each

Fig. 5. Mean (±SD) measured osmotic pressure of body fluids(triangles) and mean (±SD) osmotic pressure of original solutes andtotal induced SPs (circles) of pre-acclimated (filled symbols) and con-trol (open symbols) F. candida, exposed to different treatments(%RH). The osmotic pressure of combined SPs was calculated on thebasis of osmotically active water using the conversion from total watercontent to osmotically active water given by Worland et al. (1998).Original solutes were assumed to increase in concentration pro-portional to the loss of osmotically active water. The dotted line indi-cates the water potential (in bar) of the surrounding air in the differenttreatments (%RH).

level of drought, and also survived exposure to a lowerrelative humidity than the springtails with no pre-acclim-ation. There was also a higher activity at lower humidityin the pre-acclimated springtails, which may facilitatemigration away from dry areas under field conditions.The pre-acclimated springtails also had slightly higherconcentrations of total SP than the control springtailsafter six days of drought exposure. An obvious expla-nation for the beneficial effects of pre-acclimation is thatthese animals would have a more favourable vapourpressure gradient between their body fluids and the sur-rounding air, thus reducing the acute water loss. Ourresults also show that pre-acclimated animals were ableto maintain a higher total water content at a given RH,and in general were able to tolerate lower internal watercontents. Pre-acclimated animals may have been able totake up water (i.e. reaching hyperosmosity) early in theexperiment while control animals needed more time toreach the state where water uptake was possible.

Several studies of invertebrates have shown that pre-acclimation to a particular form of stress provides toler-ance to a later and more severe stress. This has beendemonstrated in flies for heat stress (Krebs andLoeschcke, 1994), cold stress (Chen et al., 1987; Czajkaand Lee, 1990; Kelty and Lee, 1999), and desiccationstress (Hoffmann, 1990). Of particular interest for thepresent study are the studies of Crowe and Madin(1975), and Higa and Womersley (1993) on desiccationtolerance of the nematode Aphelenchus avenae. Theseauthors showed that gradual exposure to increasing des-iccation regimes greatly enhanced the survival rate com-pared to a direct transfer to the particular desiccationlevel. Moreover they showed that the enhanced survivalwas correlated with the accumulation of trehalose duringpre-conditioning, a situation similar to the results of thepresent study. A. avenae is known to be anhydrobiotic,and able to tolerate the loss of all osmotically activewater (Crowe and Madin, 1975). The main explanationof this ability is given by the “water replacement hypoth-esis” (Crowe et al. 1984, 1992) according to whichsugars and polyols (in particular trehalose) is able toreplace vicinal (“structural” ) water around proteins andmembranes, and thus aid the retention of such cellcomponents. There are also examples of anhydrobioticspringtails, e.g. the species Folsomides angularis(Belgnaoui and Barra, 1988). These authors showed thatcellular glycogen granules disappeared during des-iccation, presumably for the accumulation low molecularweight sugars or polyols. In that connection it is interest-ing that both pre-acclimated and control animals in ourstudy produced trehalose at the lower RH levels, andthat this seemed to happen at the expense of myo-inosi-tol (and possibly glucose). This could indicate a shift instrategy, from water absorption to anhydrobiosis. How-ever, F. candida does not survive the loss of water belowa level where membranes are likely still to be hydrated,

1026 H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

thus the water replacement hypothesis may not apply inthis case. It is possible that the shift to trehalose syn-thesis is an adaptation related to relative solubilities oftrehalose and myo-inositol. Trehalose is much more sol-uble than myo-inositol, and at the reduced water contentsdisplayed in this study the solubility of the stabilizingsolute could be of significant importance. Trehalose is acommon disaccharide in desiccation resistant arthropods(Ring and Danks, 1998), and is commonly found in anumber of anhydrobiotic invertebrates, e.g. nematodes(Madin and Crowe, 1975; Womersley and Smith, 1981;Barrett, 1991) and tardigrades (Westh and Ramløv,1991). It is an effective glass former because its tran-sition temperature is relatively high (Ring and Danks,1998), and may also behave as rubber at temperaturesabove glass formation (Danks, 2000 and referencestherein). The body fluids of F. candida had a high vis-cosity when the springtails had been dehydrated at97.5%RH or lower RH, and were under those circum-stances possibly behaving as a rubber, which woulddecrease the evaporative water loss because of a restric-ted movement of water molecules.

Obviously, the ecological significance of the des-iccation tolerance of F. candida may be that the speciesis able to remain active even during periods of droughtand therefore maintain growth and reproduction. How-ever, this ability may also be important for rapid disper-sal mechanisms. Hygrophilic springtails have limitedabilities to actively disperse due to their small size, highcuticular permeability, and adaptations to a life in thesoil (e.g. reduced or absent furca) (Bengtsson et al.,1994a,b). Nevertheless, they are abundant on every con-tinent, including the Arctic and Antarctic (Hopkin,1997). Dispersal of springtails by wind has previouslybeen ruled out as an explanation due to the high cuticularpermeability in many species. However, our presentresults suggest that individuals of this species will toler-ate severe desiccation if acclimated at a mild droughtstress before a final severe drought stress is applied, pro-posing that springtails, in a partly dehydrated state, maybe transported by wind over long distances.

Acknowledgements

We thank M.P. Hansen for valuable comments on anearly version, and two anonymous reviewers for valuablecomments to the improvement of the manuscript. Thisstudy was made possible by a grant to H.S. from theNordic Arctic Research Programme, and a grant to M.B.from the Danish Natural Science Research Council(grant No. 9700177).

References

Barrett, J., 1991. Anhydrobiotic nematodes. Agricultural ZoologyReviews 4, 161–176.

Bayley, M., Holmstrup, M., 1999. Water vapor absorption in arthro-pods by accumulation of myoinositol and glucose. Science 285,1909–1911.

Belgnaoui, S., Barra, J.-A., 1988. Anhydrobiosis in the collembola Fol-somides angularis. Biological Structures and Morphogenesis 1,147–153.

Bengtsson, G., Hedlund, K., Rundgren, S., 1994a. Food-dependent anddensity-dependent dispersal — evidence from a soil collembolan.Journal of Animal Ecology 63, 513–520.

Bengtsson, G., Rundgren, S., Sjøgren, M., 1994b. Modelling dispersaldistances in a soil gradient — the influence of metal resistance,competition, and experience. Oikos 71, 13–23.

Chen, C., Denlinger, D.L., Lee, R.E., 1987. Cold-shock injury andrapid cold hardening in the flesh fly Sarcophaga crassipalpis.Physiological Zoology 60, 297–304.

Collis-George, N., 1959. The physical environment of soil animals.Ecology 40, 550–557.

Crowe, J.H., Madin, K., 1975. Anhydrobiosis in nematodes: evaporat-ive water loss and survival. Journal of Experimental Zoology 193,323–334.

Crowe, J.H., Crowe, L.M., Chapman, D., 1984. Preservation of mem-branes in anhydrobiotic organisms: the role of trehalose. Science223, 701–703.

Crowe, J.H., Hoekstra, F.A., Crowe, L.M., 1992. Anhydrobiosis.Annual Review of Physiology 54, 579–599.

Czajka, M.C., Lee, R.E., 1990. A rapid cold-hardening response pro-tecting against cold shock injury in Drosophila melanogaster. Jour-nal of Experimental Biology 147, 245–254.

Danks, H.V., 2000. Dehydration in dormant insects. Journal of InsectPhysiology 46, 837–852.

Fjellberg, A., 1980. Identification keys to Norwegian Collembola.Norsk Entomologisk forening, Oslo, Norway.

Hadley, N., 1994. Water Relations of Terrestrial Arthropods, 1st ed.Academic Press, San Diego, CA.

Higa, L.M., Womersley, C.Z., 1993. New insights into the anhydrobi-otic phenomenon — the effects of trehalose content and differentialrates of evaporative water loss on the survival of Aphelenchusavenae. Journal of Experimental Zoology 267, 120–129.

Hodkinson, I.D., Webb, N.R., Bale, J.S., Block, W., 1999. Hydrology,water availability and tundra ecosystem function in a changing cli-mate: the need for a closer integration of ideas? Global ChangeBiology 5, 359–369.

Hoffmann, A., 1990. Acclimation for desiccation resistance in Droso-phila melanogaster and the association between acclimationresponses and genetic variation. Journal of Insect Physiology 36,885–891.

Holmstrup, M., 1997. Drought tolerance in Folsomia candida Willem(Collembola) after exposure to sublethal concentrations of threesoil-polluting chemicals. Pedobiologia 41, 361–368.

Hopkin, S.P., 1997. Biology of the springtails (Insecta: Collembola).Oxford University Press, New York, 330pp.

Kelty, J.D., Lee, R.E., 1999. Induction of rapid cold hardening bycooling at ecologically relevant rates in Drosophila melanogaster.Journal of Insect Physiology 45, 719–762.

Krebs, R.A., Loeschcke, V., 1994. Costs and benefits of activation ofthe heat-shock protein response in Drosophila melanogaster. Func-tional Ecology 8, 730–737.

Krogh, P.H., 1995. Does a heterogeneous distribution of food or pestic-ide affect the outcome of toxicity tests with Collembola. Ecotoxic-ology and Environmental Safety 30, 158–163.

Lang, A.R.G., 1967. Osmotic coefficients and water potentials ofsodium chloride solutions from 0 to 40°C. Australian Journal ofChemistry 20, 2017–2023.

1027H. Sjursen et al. / Journal of Insect Physiology 47 (2001) 1021–1027

Leinaas, H.P., Ambrose, W.G., 1992. Utilization of different foraginghabitats by the purple sandpiper Calidris maritima on a Spitsbergenbeach. Fauna Norwegica 15, 85–91.

Madin, K.A.C., Crowe, J.H., 1975. Anhydrobiosis in nematodes:carbohydrate and lipid metabolism during dehydration. Journal ofExperimental Zoology 193, 335–342.

Ring, R.A., Danks, H.V., 1998. The role of trehalose in cold-hardinessand desiccation. Cryo-Letters 19, 275–282.

Suhardjono, Y.R., Greenslade, P., 1994. Folsomides arnoldi n.sp.(Isotomidae): a new Collembolan abundant in arid Australia, witha redescription of Folsomides denisi (Womersley). Proceedings ofthe Linnean Society of New South Wales 114, 21–27.

Vannier, G., 1983. The importance of ecophysiology for both bioticand abiotic studies of soil. In: Lebrun, P., Andre, H.M., deMets,A., Gregoire-Wibo, C., Wauthy, G. (Eds.), New trends in soilbiology. Dieu-Brichart, Ottignies-Louvain-La-Neuve, Belgium, pp.289–314.

Verhoef, H.A., Nagelkerke, C.J., 1977. Formation and ecological sig-nificance of aggregations in collembola. Oecologia 31, 215–226.

Verhoef, H.A., Van Selm, A.J., 1983. Distribution and populationdynamics of Collembola in relation to soil moisture. HolarcticEcology 6, 387–394.

Weast, R.C., 1989. CRC Handbook of Chemistry and Physics, 70thed. CRC Press, Boca Raton, FL, USA.

Westh, P., Ramløv, H., 1991. Trehalose accumulation in the tardigradeAdorybiotus coronifer during anhydrobiosis. Journal of Experi-mental Zoology 258, 303–311.

Womersley, C., Smith, L., 1981. Anhydrobiosis in nematodes — I.The role of glycerol, myo-inositol and trehalose during desiccation.Comparative Biochemistry and Physiology B 70, 579–586.

Worland, M., Grubor-Lajsic, G., Montiel, P., 1998. Partial desiccationinduced by sub-zero temperatures as a component of the survivalstrategy of the Arctic collembolan Onychiurus Arcticus (Tullberg).Journal of Insect Physiology 44, 211–219.

Yosii, R., 1966. Snow Collembola of the Siachen glacier in Karako-ram. Results of the Kyoto University Scientific Expedition to theKarakoram and Hindukush, vol. 8, pp. 407–410.