Water Vapor Absorption in Arthropods by Accumulation of Myoinositol and Glucose

See allHide authors and affiliations

Science  17 Sep 1999:
Vol. 285, Issue 5435, pp. 1909-1911
DOI: 10.1126/science.285.5435.1909

This article has a correction. Please see:


Hydrophilic soil arthropods have been thought to respond to soil desiccation exclusively by migrating to deeper soil layers. Numerous studies have shown that their survival below 90 percent relative humidity dry weight, is limited to hours. However, little attention has been paid to physiological adaptations to more realistic desiccation regimes, such as at the permanent wilting point of plants (98.9 percent relative humidity). A water vapor absorption mechanism is described that allows a common soil collembolan, Folsomia candida, to remain active down to below the permanent wilting point. A reevaluation of the water physiology of this widespread and diverse animal group is required.

Soil-dwelling arthropods have several characteristics that distinguish them from surface-living forms, in particular with respect to water balance. These characteristics include small size, epidermal respiration, and high integumental permeability to water (1). In fact, soil-dwelling Collembola are reported as having no physiological or metabolic means of regulating water loss and generally have high integumental permeability compared with other terrestrial arthropods (2, 3). These characters seem to match the soil environment, where the pore humidity is normally very close to 100% relative humidity (RH). However, during dry periods, evapotranspirative removal of water throughout the root zone may reduce pore humidity. At about 98.9% RH, at which the water potential is at −15 bars, plants normally wilt (4). The hemolymph osmolality of soil arthropods, such as Collembola, is usually about 300 mosm/kg (5), roughly corresponding to an osmotic pressure of −8 bars and 99.4% RH. Thus, at soil pore relative humidities below 99.4%, which occur frequently, Collembola and other soil arthropods must tolerate, or in some way avoid, a net efflux of water.

Folsomia candida (Willem) is able to survive more than a week at 98.2% RH (6). This RH corresponds to a water potential deficit of about 17 bars between the environment and the normal body fluid osmotic pressure of these animals. We investigated this phenomenon by monitoring the water content, the hemolymph osmotic pressure, and the sugar and polyol (SP) contents ofF. candida over a 7-day period of dehydration in an atmosphere maintained at a constant 98.2% RH. The response of these animals was compared with that of animals kept in an atmosphere at 99.6% RH without free water and with animals kept at 100% RH with access to free water.

As was expected, animals held in an atmosphere at 98.2% RH dehydrated rapidly during the initial 24 hours. They lost a third of their initial total water content, equivalent to a loss of half of their osmotically active body water (OAW) (Fig. 1), and became wrinkled in appearance and inactive, responding only to strong tactile stimulation. Controls in 99.6 and 100% RH remained unchanged in the same period. This pattern was repeated in the osmotic pressure of the body fluids, where the dehydrated animals approached water potential equilibrium with their environment at −24 bars (Fig. 2). Net water efflux was therefore close to zero at this time. The initial osmotic pressure of the animals was −7.5 bars. A loss of half of their OAW should result in an osmotic pressure of about −15 bars, so these animals should still be suffering from a water potential deficit with their environment. However, the animals actively increased their osmotic pressure to halt water efflux. Part of the explanation for this active increase in osmolality can be found in the synthesis of myoinositol (Fig. 3). This polyol, undetectable in control animals, constituted 1.4% of the dry weight of animals after the first 24 hours. Under the assumption that all myoinositol was dissolved in OAW, the contribution to the osmotic pressure would be about 5 bars, leaving less than 4 bars to be explained by other osmolytes.

Figure 1

(left). Delayed passive water vapor absorption in F. candida exposed to severe drought stress (18). Osmotically active water (OAW; mean ± SE; n = 5) in animals exposed to 98.2% RH (circles), 99.6% RH (squares), or 100% RH with access to free water (triangles). The vertical dotted line indicates the time at which animals were returned to Petri dishes with access to free water.

Figure 2

(right). Drought-stressed animals reestablished hyperosmoticity to their environment within the first 48 hours and were thus able to absorb water vapor. Body fluid osmotic pressure (mean ± SE; n = 3) in F. candida exposed to 98.2% RH (circles), 99.6% RH (squares), or 100% RH with access to freewater (triangles), measuredas described (9). The horizontal lines indicate the corresponding water potential at 98.2% RH (short-dashed line) and 99.6% RH (long-dashed line). The vertical dotted line indicates the time at which animals were returned to Petri dishes with access to free water.

Figure 3

(left). Sugar and polyol production is rapidly induced in drought-stressed animals (19). Contents (mean ± SE; n = 5) of myoinositol (circles) and glucose (triangles) in F. candidaexposed to 98.2% RH (solid line), 99.6% RH (short-dashed line), or 100% RH with access to free water (long-dashed line). The vertical dotted line indicates the time at which animals were returned to Petri dishes with access to free water.

The animals did not become iso-osmotic but within 48 hours of desiccation reestablished their hyperosmoticity to their surroundings at about the same level as under normal conditions (Fig. 2). This resulted in a net influx of water by passive absorption from the atmosphere. Accordingly, between the point of maximum dehydration at 24 hours and the end of the dehydration period at 169 hours, the animals increased their OAW content from 0.51 to 0.91 g of H2O g−1 of dry weight (Fig. 1). At the outset of the experiment, the animals contained 0.99 g of OAW g−1of dry weight, meaning that they had, in the following 6 days, recovered almost all of the water lost during the first 24 hours. This rehydration was accompanied by a clear return to the animals' normal appearance, which again was fully active as soon as their containers were opened instead of requiring strong tactile stimulation to induce movement. The increase in osmotic pressure was associated with sharp increases in myoinositol and glucose (Fig. 3), which had stabilized after 5 days at a combined concentration of 8% of the animals' dry weight. Under the assumption that all SP molecules made a colligative donation, these chemicals contributed 15 bars to the increased osmotic pressure after 48 hours (Fig. 4). At this time, the combined effects of water loss and SP production left only 11% of the increase in body fluid osmotic pressure unexplained. SP synthesis continued and reached a plateau at 120 hours; however, at this time, animals had rehydrated more than could be explained by SP production alone. A comparison of the animals' water content, SP content, and body fluid osmotic pressure leaves about 9 bars to be explained by unidentified osmolytes. At day 7, the dehydrated animals were returned to an environment at 100% RH with free liquid water. This abrupt shift to a rehydrating environment forces the animals to remove osmolytes from their body fluids to avoid a lethal influx of water. Indeed, about 50% mortality was observed during the following 48 hours. However, surviving animals absorbed water rapidly (Fig. 1), and the SPs disappeared completely in that time (Fig. 3).

Figure 4

(right). Sugars and polyols make a major contribution to the osmotic pressure in drought-stressed animals. Combined osmotic contribution (mean ± SE; n = 5) of myoinositol and glucose in body fluids ofF. candida exposed to 98.2% RH (circles). For comparison, the total body fluid osmotic pressure is also shown (triangles). The vertical dotted line indicates the time at which animals were returned to Petri dishes with access to free water.

Synthesis of SPs with colligative properties causing a lowering of the body fluid melting point, synonymous with depression of vapor pressure, has been reported in cold-hardy arthropods (7). It has previously been suggested that SP accumulation may therefore also reduce water loss under desiccating conditions (8,9). Accumulation of SPs is also associated with anhydrobiosis, well known in soil invertebrates including tardigrades (10), nematodes (11), and Collembola (12). However, in anhydrobiosis, the adaptive value of SP accumulation has been primarily ascribed to noncolligative properties involved in the protection of membranes and proteins (13). The role of SP accumulation in F. candida is apparently not associated with anhydrobiotic properties because the lethal limit for drought tolerance is above 95% RH (6). Instead, it appears that the SP accumulation in the present study is linked to water vapor absorption at low soil humidities.

A number of terrestrial arthropods have the capability to balance their water budget by actively absorbing water from highly unsaturated atmospheres. These truly terrestrial arthropods achieve this balance by a combination of a cuticle with very low permeability and locally creating extremely low water activity in specialized tissues (3). In common with most soil arthropods, F. candida has a highly permeable integument, making localized active water absorption inappropriate. This animal is therefore forced to maintain all its body fluids hyperosmotic to its surroundings to allow net water uptake from the atmosphere by passive diffusion along the gradient in water potential. We have shown that glucose and myoinositol account for a large portion of the measured increase in osmotic pressure. This type of water vapor absorption confers the capability of meeting the water requirements of a terrestrial arthropod under prolonged drought stress (14). The adaptive importance of this mechanism is obvious, allowing F. candidato remain active in the same range of drought intensities that plants are capable of surviving and that must therefore occur throughout the root zone (15). Hitherto, experimental designs in studies of hygrophilic soil arthropods have masked the discovery of such water-regulating abilities. The physiological adaptations of these animals to desiccation require reevaluation.

  • * To whom correspondence should be addressed. E-mail: mho{at}


View Abstract

Stay Connected to Science

Navigate This Article