1. A method is described for establishing a stable relationship between body weight and hydration (i.e. 100% hydrated) in terrestrial slugs. This allows the use of ‘% initial body weight’ as an indicator of relative body hydration.

  2. During progressive dehydration in the slug, Umax maximus (L.), the haemolymph osmolality increases exponentially. This, together with measurements of wet weight/dry weight ratios of various tissues, indicates that the water loss during the initial stages of dehydration is from the haemolymph.

  3. Although long-term (days) cyclical changes in body hydration were not observed, a daily cycle is described with highs at 12.00 midnight and lows at 12.00 noon.

  4. It is shown that the huddling behaviour of slugs can reduce evaporative water loss from individual slugs.

  5. Dehydration of a slug to about 90 % of its initial body weight initiates a rhythmic cycle of pneumostome closures. Further dehydration results in an increase in the duration of the closures and reduction of the diameter of the pneumostome when open.

  6. The absolute and relative O2 uptake by the integument and lung are presented. In quiet slugs, the lung is responsible for about 20 % of the total O2 uptake, while in active slugs it is responsible for 50 % of the total. During activity there is a greater increase of O2 uptake across the lung than across the integument.

In most animals the maintenance of water balance involves modulation of water regulatory behaviour. Such behaviour includes drinking (Fitzsimons, 1972, 1979; Andersson, 1978), burrowing and cocoon formation by certain amphibians (e.g. Canziani & Canata, 1980), water orientation in insects (Edney, 1977) and withdrawal responses in shelled molluscs (e.g. McAlister & Fisher, 1968). These behavioural changes are initiated by drying conditions or actual dehydration of the animal and serve to minimize changes in total body hydration. In addition to initiating discret activity patterns, changes in the level of body hydration can modulate the general level of behavioural responsiveness (e.g. Wolf, 1938; Dainton, 1954a, b;,Roots, 1956).

Certain types of water regulatory behaviour result in a reduction of evaporative water loss. These responses include orientation to areas of high humidity and manoeuvres that result in reduced exposure of moist surfaces such as respiratory epithelia. Moist-skinned air-breathing animals, such as amphibians and terrestrial gastropods, are especially susceptible to dehydration because of the considerable loss of water across their integuments and lung surfaces (Machin, 1975; Carley, 1978). Reducing evaporative water loss is a general strategy that is particularly evident in these organisms. This strategy does, however, result in a serious dilemma. Reduced exposure of respiratory surfaces can minimize evaporative water loss, but in turn restricts respiratory gas exchange. Thus, in drying conditions, the behaviour of these animals must be a compromise between avoidance of dehydration and maintenance of sufficient respiratory function.

As part of an effort to understand the modulation of water regulatory behaviour, we have examined aspects of water balance and respiratory function in the terrestrial slug, Limax maximus. Our results indicate that in drying conditions body hydration in Limax is maintained primarily by behavioural means. This is accomplished, in part, by a set of discrete responses that can restrict evaporative water loss. These responses include progressive closure of the pneumostome (opening to mantle cavity) which reduces exposure of the lung surface, and ‘crowding behaviour’ among groups of slugs which reduces the area of exposed body surface.

In Limax, as in many other pulmonate gastropods, much of the mantle cavity is specialized as a vascularized lung surface (Runham & Hunter, 1970). The opening to the lung is the pneumostome, the diameter of which is controlled by circular and longitudinal bands of muscle (see Fig. 1). In the present study we have demonstrated that by controlling the exposure of the lung, the pneumostome can have an effect on both evaporative water loss and respiratory gas exchange.

Fig. 1.

(A) Umax maximus in its normal moving posture. The superior and inferior tentacles are extended and the pneumostome is fully open. (B) A slug with a plastic grommet sutured into the pneumostome to keep it open. (C) A cross section of the body wall with an implanted grommet.

Fig. 1.

(A) Umax maximus in its normal moving posture. The superior and inferior tentacles are extended and the pneumostome is fully open. (B) A slug with a plastic grommet sutured into the pneumostome to keep it open. (C) A cross section of the body wall with an implanted grommet.

When groups of slugs are exposed to drying conditions they gather into closely packed groups. Although this response was previously described as ‘crowding behaviour’ (Prior, 1981), we are now adopting the term ‘huddling behaviour’ which was used by Cook (1981) in a more extensive description of the same behaviour inL. pseudoflavus. This response to drying conditions is a result of the preference of slugs for any moist area, including other slugs. Due to the contact between the slugs there is a reduction of the total area of exposed body surface. In addition, due to evaporation from slugs in the group, a humid microenvironment is generated which could reduce further evaporative water loss. A preliminary report of a portion of this work has appeared in abstract form (Prior, 1979).

Specimens of Limax maximus were collected locally and kept in the laboratory in vented plastic refrigerator boxes lined with moist paper towels. Slugs were fed dry rat chow (Purina) and vegetables. Before use, individual slugs were isolated in small vented plastic sandwich boxes which were lined with moist paper towels. Following 7 days of fasting in this condition, the slugs were considered to be 100 % hydrated (see Results for a detailed description of the basis for this conclusion). In dehydration experiments, 100 % hydrated slugs were placed singly in large (155 mm diameter) dry plastic Petri dishes fitted with screen lids. The Petri dish, lid and slug were then weighed periodically during the test period (e.g. Fig. 4) to determine the extent of evaporative water loss.

The relationship between water loss and actual body hydration was determined by measuring the osmolality of haemolymph samples obtained from slugs that had been de-hydrated to varying extents. I n addition, the wet weight/dry weight ratios of whole slugs and specific tissues were determined. Haemolymph was obtained through a cut in a superior tentacle or in the posterior body wall. Following centrifugation, the osmolality of the haemolymph samples was determined by the method of freezing point depression (Advanced Instruments Osmette A) or with a vapour pressure osmometer (Wescor, 5120). Wet weight/dry weight ratios of tissues were determined by weighing (to 0·1 mg) tissues dissected from hydrated and dehydrated slugs, drying them at 50 °C to total dryness (5 days), and then weighing them again. The same was done for measurements on whole slugs. In all cases slugs were used that had been fasted for 5–7 days.

In the experiments designed to assess the effects of huddling behaviour on water loss, slugs were individually identified by their distinctive mantle markings, weighed and placed in vented plastic refrigerator boxes (32cm×25cm×10cm) that were either empty and dry or lined with wet paper towels.

To determine the effect of a group of slugs on the general humidity in a dry box, seven slugs were placed in a dry experimental box with the probe of a Lab-line Electro-hygrometer in the centre [at 18–20 °C, 31 % relative humidity (RH)]. The relative humidity in the box was measured once every 5 min for 1 h. The relative humidity remained between 31-34%, except when slugs moved over the probe shield. When two slugs were near the probe, the RH rose to 40–45 %, and when five or more slugs were on or next to the probe, the RH was as high as 64 %. As these slugs began to move away, the RH dropped immediately back to 34 %. Thus the presence of the slugs did not significantly raise the humidity in the entire box, but rather, as expected, just in the immediate vicinity of the slugs.

All measurements of pneumostome activity were made by direct visual observation using a hand-held ruler and stopwatches. Oxygen uptake measurements were made with a Gilson Differential Respirometer fitted with large-necked flasks (6 cm diameter; with potassium hydroxide wells) to allow easy transfer of slugs. Slugs were left in the respirometer flasks for 45–50 min before measurements were made and three sets of measurements (three-six, 5 min measurements/set) were made from each slug. Measurements of oxygen uptake by the integument alone were made by reversibly blocking the pneumostome. Hollow plastic grommets, which were sutured into the pneumostomes (see Fig. 1) of cold anaesthetized slugs, could be plugged with rubber stoppers. The slugs recovered from the anaesthesia within 10–20 min and were kept in a moist chamber for 2–3 h before the measurements were made. The oxygen uptake results were unaffected by the order of the measurement procedure (i.e. plugged or unplugged). A similar approach was used to assess the contribution of the lung to total evaporative water loss. Weight loss in normal slugs was compared with height loss in slugs with open grommets in their pneumostomes.

Body hydration cycles

It has been reported that the weight of a slug can fluctuate more than 25 % in 24 h and that the variation is due to cyclical changes in body hydration (Howes & Wells, 1934a, b; discussed in Runham & Hunter, 1970). Unpredictable variations of this sort could produce unreliable measurements of initial levels of body hydration and, therefore, of relative dehydration. It was therefore necessary to determine if cyclical variations in body hydration occur that could affect our results.

Individual slugs, maintained in high relative humidity, were weighed daily at approximately 9.00 a.m. for 14 days. During the first five days they had access to food. The daily variations in body weight during this initial period (days 1–5) were from 13–25 % (Fig. 2). To determine the effects of feeding on fluctuations in body weight, food was removed on day 5 of the experiment. After 5–7 days of fasting (days 10–12 in Fig. 2), the mean daily weight variation was only 1–2% and the range of the variations (i.e. % change in weight) was significantly reduced (see the S.D. of days 10–14, Fig. 2). The inset in Fig. 2 illustrates the typical daily variation in one slug (0·25 %). Note that the curves in Fig. 2 illustrate daily variation in body weight, not the actual weights of the slugs. Although these particular measurements did not reveal the type of cyclical variation reported by Howes & Wells (1934a, b), we examined the hourly variation in weight during the course of days to determine if there was a daily cycle.

Fig. 2.

The mean % daily variation ( ± S.D.) m body weight of slugs isolated in small boxes lined with wet paper towels for a period of 14 days. Until day 5 the slugs had access to food. From day 5 to day 14 the slugs were fasted. The inset illustrates the variation of a single slug.

Fig. 2.

The mean % daily variation ( ± S.D.) m body weight of slugs isolated in small boxes lined with wet paper towels for a period of 14 days. Until day 5 the slugs had access to food. From day 5 to day 14 the slugs were fasted. The inset illustrates the variation of a single slug.

Fig. 3 illustrates the variation of body weight [% of initial body weight (% IBW)] in fasted slugs over a 56 h period. Although the measurements extended for only two complete 24 h periods, it is clear that there is a cyclical variation in body weight. It appeared that there was minimal variation in the weights of individual slugs in the period from 8.00a.m. to 12.00 noon. The mean (± S.D.) of the three 8.00 a.m. measurements was calculated for each slug. When expressed as the percentage of the mean weights, it is evident that there was little variation in the standard deviations (x% variation of S.D. = 3·6 ± 2·2). This indicates the reliability of body weight measurements at 8.00a.m.

Fig. 3.

Variation in weight (% IBW) of 100 % hydrated slugs as a function of time. The % IBW was calculated by dividing each weight measurement by the initial 8.00 a.m. measurement for that slug. This was done for each weight measurement on each slug. Each point is the mean % IBW ( ± S.D.) of 15 slugs.

Fig. 3.

Variation in weight (% IBW) of 100 % hydrated slugs as a function of time. The % IBW was calculated by dividing each weight measurement by the initial 8.00 a.m. measurement for that slug. This was done for each weight measurement on each slug. Each point is the mean % IBW ( ± S.D.) of 15 slugs.

The effectiveness of our fasting procedure in establishing a reliable level of body weight and hydration (i.e. 100% hydrated slug) is further indicated by the regularity of whole body and tissue water content (Table 1) and haemolymph osmolality (see Fig. 5). Measurements of total body wet weight of 100% hydrated slugs showed considerable constancy ( ± S.D.% wet weight = 86·2 ± 1·2, N = 21; Table 1). Likewise the hydration levels of selected tissues (body wall, digestive gland, brain) from 100% hydrated slugs were very constant, in each case the standard deviation being less than 1 % of the mean (Table 1).

Table 1.

Percentage wet weight of whole slugs and isolated tissues

Percentage wet weight of whole slugs and isolated tissues
Percentage wet weight of whole slugs and isolated tissues

We conclude that by using our fasting procedure and beginning experiments approximately 8.00 a.m., we can accurately represent the relative level of hydration as ‘percent of initial body weight’ (% IBW). Slugs standardized in this manner were used in all subsequent experiments.

Evaporative water loss

The particular susceptibility of slugs to dehydration (Dainton, 1954a) is due to evaporation of water from the integument, the lung (via the pneumostome; Fig. 1), and the mucous trail deposited during locomotion. To measure the extent and time course of water loss, we allowed slugs to move about in large dry Petri dishes fitted with screen lids at room temperature (19–22 °C) and relative humidity (25–30%). Active slugs continued locomotor activity during the test period and lost about 33 % of their initial body weight (Fig. 4). In contrast, inactive slugs, which remained stationary throughout most of the test period, lost only 8% of their initial body weight. In addition to losses due to a mucous trail, the greater water loss during locomotion may be due to changes that increase the total area of exposed evaporating surfaces. For example, during locomotion, the body is elongated, which increases the surface area/volume ratio, and the pneumostome is kept fully open, which results in maximum exposure of the lung.

Fig. 4.

The decline in body weight of inactive ○ – – ○ and active ●——● slugs due to evaporative water loss. Each point is the mean ( ± S.D.) of measurements from 10 slugs.

Fig. 4.

The decline in body weight of inactive ○ – – ○ and active ●——● slugs due to evaporative water loss. Each point is the mean ( ± S.D.) of measurements from 10 slugs.

One of the major consequences of dehydration in Limax is a dramatic increased haemolymph osmolality (Fig. 5A). Evaporation of water from slugs is dependent upon the difference in vapour pressure between the haemolymph and the atmosphere. During dehydration both the partial pressure of water in the atmosphere and the total amount of solute in the slug remain constant. Therefore the progressive loss of water during dehydration should have resulted in an exponential increase in haemolymph osmolality. This is illustrated in Fig. 5B where the data from Fig. 5A were replotted as log haemolymph osmolality against % IBW.

Fig. 5.

Haemolymph osmolality plotted against the level of body hydration (% IBW). Haemolymph osmolality increases as a function of dehydration ( A). This exponential increase is particularly evident in (B) where the data from (A) are replotted semi-logarithmically and result in a straight line function. Each point represents one measurement from one slug. The regression line was calculated by the method of least squares.

Fig. 5.

Haemolymph osmolality plotted against the level of body hydration (% IBW). Haemolymph osmolality increases as a function of dehydration ( A). This exponential increase is particularly evident in (B) where the data from (A) are replotted semi-logarithmically and result in a straight line function. Each point represents one measurement from one slug. The regression line was calculated by the method of least squares.

To determine if tissues experienced similar reductions in water content, we measured the wet weight/dry weight ratios of body wall, digestive gland and brain from dehydrated slugs (Table 1). No significant changes in tissue water content were observed in slugs dehydrated to less than 80–70% IBW, suggesting that the loss of body weight at this level of dehydration was primarily due to water loss from the haemolymph. But at 65–50 % IBW, the percent wet weight of the tissues was lowered and at 48–37% IBW it was further reduced (Table 1). In the 48–37% IBW group there was an 8–10% reduction in percent wet weight of brain and digestive gland, while in the case of body wall the reduction was 16·9%, which is probably due to drying of the exterior surface.

Behavioural responses to dehydration

Progressive dehydration in Umax leads to behavioural responses that can reduce the rate of dehydration. Among these responses are (1) huddling by groups of slugs and (2) reduction in total exposure of the lung surface. Both of these illustrate the strategy of reducing the total area of exposed evaporating surface during drying conditions.

Huddling behaviour

We observed that in culture boxes that had been allowed to become relatively dry, the slugs were usually crowded closely together. This suggested that ‘huddling behaviour’ (Prior, 1981; Cook, 1981) could be initiated by drying conditions and possibly serve to reduce dehydration. To test this possibility, 100% hydrated slugs were individually identified, weighed and placed as groups (5–6 individuals), into large dry, vented plastic containers. These were left overnight (12–15 h) at room temperature and humidity (18–21 °C, 25–30% RH). In the morning the locations and weights of the individual slugs were recorded. With the exception of two slugs that were not with the group in one trial, all of the slugs in each trial were found huddled together in a corner of their respective containers. The mean overnight weight loss from grouped slugs in dry conditions was 22·4% (Table 2), which was less than the loss from a single active slug in only 2h (see Fig. 4). The experiment was repeated using containers lined with moist paper towels. In these conditions, only 2 of 21 slugs were found together and the weights of the overall group were not significantly altered (Table 2). When the experiment was repeated using one slug per dry container to prevent huddling, the mean weight loss was 54·8% which is more than twice that recorded in dry huddled slugs (Table 2). The relative humidity of a dry container was increased only in the immediate vicinity of the group of slugs (see Materials and Methods, for details of the measurements), thus the presence of the slugs did not alter the general humidity in a box. These results suggest that huddling behaviour of slugs, which is initiated by drying conditions, is effective in minimizing water loss.

Table 2.

Effects of huddling on water loss in Limax

Effects of huddling on water loss in Limax
Effects of huddling on water loss in Limax

Pneumostome responses

Ina fully hydrated slug, the pneumostome is kept open most of the time, especially during locomotion. Dehydration results in initiation of rhythmic closures of the pneumostome and progressive reduction of the open diameter (Fig. 6). In addition, there is an increase in the duration of the closures (Fig. 7). These responses, which reduce the exposure of the lung, occur in sequence as a function of the level of dehydration, the rhythm being initiated at about 90 % IBW, and the graded increase in closure duration and decrease in open diameter beginning at about 90–80 % IBW (Figs 6, 7). As dehydration progressed, the duration of the closures increased until the pneumostome was kept closed most of the time. There was considerable variation in this response, but as seen from the four cases illustrated in Fig. 7, moderate dehydration resulted in significant increases in the total time the pneumostome was closed. Further dehydration (70–80% IBW) resulted in cessation of the pneumostome rhythm and maintenance of a very small diameter opening (about 1-0mm). Regardless of the duration of closures, the frequency of the rhythm in a particular slug remained the same (1 or 2 closures/min).

Fig. 6.

The activity and diameter of the pneumostome at different levels of body hydration (% IBW). At approximately 90% IBW the closure rhythm of the pneumostome was initiated. Each point represents a single measurement; measurements from 21 slugs (initially 100% IBW) are included.

Fig. 6.

The activity and diameter of the pneumostome at different levels of body hydration (% IBW). At approximately 90% IBW the closure rhythm of the pneumostome was initiated. Each point represents a single measurement; measurements from 21 slugs (initially 100% IBW) are included.

Fig. 7.

The percentage of time the pneumostome is closed plotted as a function of % IBW. Measurements from four slugs during progressive dehydration are presented.

Fig. 7.

The percentage of time the pneumostome is closed plotted as a function of % IBW. Measurements from four slugs during progressive dehydration are presented.

Effect of pneumostome activity on evaporative water loss

To assess the effect of pneumostome closure on evaporative water loss from the lung we measured the rate of weight loss in slugs with normal pneumostome activity, and then in the same slugs when their pneumostomes were kept open with implanted grommets (see Fig. 1). Fully hydrated slugs were anaesthetized to allow implantation of the grommets (‘normal’ slugs were likewise anaesthetized). Following 2 h of recovery, they were placed in large Petri plates fitted with screen tops at room temperature and humidity (18–21 °C, 70–80 % RH) and were weighed every 30 min. The data are presented in Fig. 8, as the means ( ± S.D.) of measurements on the six slugs when: (1) their pneumostomes were kept open with grommets (N =6 measurements), and (2) their pneumostomes were allowed normal activity (N = 12 measurements). Water was lost at the same rate in the two conditions until dehydration reached 90 – 85 % IBW. At this level of dehydration the slugs lost less water when they had normal pneumostome activity than they did when their pneumostomes were kept open. It was at about this level of dehydration that the pneumostome rhythm and progressive closure were normally initiated (see Fig. 6). Regardless of the exact point of divergence of the curves in Fig. 8, when slugs had normal pneumostome activity, they lost less water . This is also seen in the inset in Fig. 8, in which the final weights of the individual slugs with and without grommets are compared. Although pneumostome closure alone could not prevent dessication, it does contribute to the overall effectiveness of the behavioural responses that serve to minimize dehydration.

Fig. 8.

Evaporative water loss (% IBW) against time. Three measurements were made from each slug (N = 6), two in which normal pneumostome activity was possible and one when the pneumostome was kept open with a grommet (see Fig. 1). In each experiment the slugs were initially 100% hydrated. Each point in the curve of normal slugs ○——○ is the mean ( ± S.D.) of 12 measurements, and in the curve of ‘open pneumostome’ slugs, ●——●, 6 measurements. The inset compares the final measurements (% IBW) of each slug in each condition. Measurements were made at 18–21 °C, 70–80 %RH.

Fig. 8.

Evaporative water loss (% IBW) against time. Three measurements were made from each slug (N = 6), two in which normal pneumostome activity was possible and one when the pneumostome was kept open with a grommet (see Fig. 1). In each experiment the slugs were initially 100% hydrated. Each point in the curve of normal slugs ○——○ is the mean ( ± S.D.) of 12 measurements, and in the curve of ‘open pneumostome’ slugs, ●——●, 6 measurements. The inset compares the final measurements (% IBW) of each slug in each condition. Measurements were made at 18–21 °C, 70–80 %RH.

Effect of pneumostome activity on O2 uptake

The responses of the pneumostome to dehydration are well suited for reducing exposure of the lung and thereby evaporative water loss. This, however, also reduces respiratory gas exchange. We examined this apparent conflict by determining to what extent pneumostome activity affects respiratory function. Because integumental O2 uptake occurs in slugs (Schuurmans-Stekhoven, 1920), we partitioned respiratory function by measuring the relative contributions of the lung and integument to total oxygen uptake. O2 uptake measurements were made on slugs with grommets implanted in their pneumostomes (see Fig. 1). With a grommet open, the lung surface was completely exposed so that O2 uptake measurements included the integument and the lung. With a grommet plugged, the measurements were considered to be the O2 uptake of the integument.

When the slugs were introduced into the respirometer flasks they moved about for several minutes, after which many of them remained quiet for the entire measurement period (‘inactive’ slugs, N=5). Those slugs that continued to move about were considered to be ‘active’ slugs (N = 7). Of the total O2 uptake by inactive slugs ( ± S.D. = 4·53 ± 0·42 μO25 min−1 gm−1; range = 4·20–5·35), 20% can be attributed to the lung and 80 % to the integument. During activity, however, the contribution of the lung was 50% of the total O2 uptake (total O2 uptake = 9·44 ± 2·13 μl 5 min−1 gm−1 ; range = 7·08–13·08) and represented a five-to six-fold increase over the O2 uptake of the lung in inactive slugs. In contrast, there was only a small difference between integumental O2 uptake in inactive (3·70 ± 0·33 μlO25 min−1 gm−1) and active slugs (4·47 ± 1·09 μlO25 min−1 gm−1).

The difference in O2 uptake between plugged and unplugged active slugs might suggest that the plugged animals could have experienced an increase in haemolymph Although measurements of haemolymph or lung were not made, the maintenance of regular locomotor activity and reflex responsiveness in plugged slugs suggest that CO2 anaesthesia did not occur. Also it is useful to note that during integumental respiration in the lungless salamander, which is analogous to a plugged slug, the blood was found to be exceptionally low for a terrestrial vertebrate (; Gatz, Crawford & Piiper, 1974). This seems to be due to the high diffusion coefficient of CO2 and the lack of air convection resistance between the integument and the atmosphere. These observations also suggest that our measurements of integumental O2 uptake were not significantly affected by potential increases in haemolymph

The considerable respiratory range of the lung system in Umax further emphasizes the importance of the pneumostome in regulating lung exposure. It is clear that the pneumostome, in its role of regulating lung exposure, can affect both water balance and respiratory function.

As in other moist-skinned terrestrial animals, slugs can rapidly lose water in dry conditions (Dainton, 1954a; Machin, 1975). Machin (1975) has pointed out that water will evaporate from all surfaces of a slug whenever the ambient humidity is below about 99·5 % RH (20 °C). It is therefore not surprising that there exists an array of behavioural responses that are involved in maintaining body hydration. Even so, body hydration can vary significantly depending upon the environmental conditions.

100 % hydrated slugs

The uncertainty about the level of body hydration in terrestrial snails and slugs has made it difficult to assess properly the physiological and behavioural responses involved in the regulation of water balance (see Machin, 1975 for a review). We have circumvented this problem with a fasting procedure that establishes a reliable level of study hydration, as indicated by the regularity of haemolymph osmolality (Fig. 5), body weight (Figs 2, 3) and whole animal and tissue water content (Table 1). Even with this fasting procedure, the daily weight (i.e. hydration) cycle (Fig. 3) makes it necessary that experiments be initiated at approximately the same time of day. Thus by using our fasting procedure and making the initial experimental measurements at 8.00–9.00 a.m., it is possible to relate our observations to standard 100% hydrated slugs, and use % of initial body weight’ as an indication of the level of hydration. Unless slugs are hydrated (or dehydrated) in a controlled manner, any measurements related to water balance will necessarily vary to a great extent. For example, the unknown level of initial body hydration and the exponential relationship between haemolymph osmolality and dehydration (Fig. 5) may well explain the variation in measurements of haemolymph osmolality on ‘normal’ slugs and snails (e.g. Bailey, 1971, 349-409 mosmol kg H2O−1; Burton, 1964, 129-333 mosmol kg H2O−1; Roach, 1963, 97-231 mosmol kg H2O−1.

Changes in osmolality are known to have dramatic effects on numerous cellular functions, among the most sensitive of which is neurone excitability (Prior & Pierce, 1978, 1981 ; Treheme, 1980). Changes in the osmotic pressure or ionic composition of the saline bathing an isolated Umax brain (with intact sheath) can cause rapid (5–20s) changes in the patterned activity of specific neurones (Prior, 1981). This observation, together with other electrophysiological and ultrastructural data, suggest that the sheaths that encapsulate the ganglia of molluscs do not significantly impede the movement of water and ions (Lane & Treherne, 1972; Mellon & Treherne, 1969; Sattelle & Lane, 1972; Mirolli & Gorman, 1973; Sattelle, 1973). Thus molluscan neurones are exposed to the variation in haemolymph osmotic pressure and ionic concentration that occurs during dehydration.

Changes in haemolymph osmotic pressure may in fact be responsible for the initiation of water regulatory behaviour in slugs. It is already well established that increases in blood osmotic pressure can elicit drinking behaviour in a variety of vertebrate species (Fitzsimons, 1979). In similar types of experiments we have recently shown that the dehydration-induced reduction of feeding responsiveness in Limax can be mimicked by injections of hyperosmotic mannitol solutions (Phifer & Prior, 1982; Prior, 1983). Experiments are in progress that are designed to test the generality of this response.

Behavioural responses to dehydration

Huddling behaviour

A common feature of the behavioural responses we have examined is that they result in reduction of the total area of exposed evaporating surface. In the case of huddling behaviour, the reduction of water loss is probably due to a combination of two factors : (1) evaporation from a crowd of slugs generating a high humidity microenvironment which could reduce further water loss (see Materials and Methods for measurements) and (2) reduction of surface area by direct contact between slugs in a huddle (see Cook, 1981). Our results indicate that huddling is initiated by drying conditions (Table 2). It is likely that the response is simply a result of the preference of slugs for regions of high humidity. Comparable results have recently been obtained using both L. pseudoflavus and L. maximus (A. Cook, personal communication).

There is at least one report in the literature in which huddling in natural conditions was observed. South (1965) found slugs aggregated in suitable resting sites during the day. In this situation however the actual huddling of the slugs was probably less crucial for water balance than the high humidity of the site itself.

Pneumostome response

Dehydration also results in pneumostome responses that reduce the exposure of the lung surface. These include initiation of the opening and closing rhythm, increased duration of the closed periods, and reduced open diameter (Figs 6, 7). This is particularly clear in moderately dehydrated slugs (80–65% IBW), in which the pneumostome is kept closed most of the time (Fig. 7). These changes in pneumostome activity reduce the total exposure of the lung and thereby evaporative water loss from this surface. The cross-sectional area of the open pneumostome is small compared to the surface area of the body wall. It would be expected therefore that evaporation from the lung would account for a small percentage of total water loss. Our measurements indicate that at moderate levels of dehydration, closure of the pneumostome can reduce the total amount of water lost, albeit only by about 7 % (Fig. 8). Pneumostome closure may actually be more important during more severe dehydration when the integument begins to dry. In this condition, were it not for closure of the pneumostome, the only exposed moist surface would be the lung. Thus the pneumostome would provide an even greater level of protection.

Spiracular closure in insects is closely analogous to pneumostomal closure in slugs. It has been shown that there is more evaporative water loss in insects when the spiracles are kept open and that dehydration leads to increased duration of closures (see Edney, 1977, for a review). Thus, the array of water regulatory responses in both insects and slugs includes reduction of the exposure of respiratory surfaces. As indicated in the Introduction, this results in the possible problem of simultaneously reducing respiratory function.

Our examination of the effects of pneumostome closure on respiration revealed that the lung system has a considerable response range. In inactive slugs, the lung was responsible for 20 % of the total O2 uptake, whereas in active slugs, the contribution of the lung was 50%. In an earlier attempt to assess the contribution of the integument to total respiration in slugs, it was found that occlusion of the lung cavity with melted paraffin resulted in a 43 % reduction in O2 uptake (Schuurmans-Stekhoven, 1920; cited in Runham & Hunter, 1970). This estimate (56%) of the contribution of the integument to total O2 uptake is similar to our measurements from active slugs. The increased metabolic demand associated with activity resulted in an approximately 500 % increase in O2 uptake by the lung, whereas the integumental O2 uptake increased by only 17 %. The comparatively slight increase in integumental O2 uptake during increased activity suggests that gas exchange across this surface is diffusion limited. In contrast, the lung displays considerable capacity for response to increases in O2 demand. This may be due, at least in part, to the 15 % increase in heartrate that accompanies locomotion (Grega, 1982; MacKay & Gelperin, 1972). An increase in heartrate could result in both increased perfusion and ventilation of the lung. Because the pericardial sac forms a wall of the lung cavity, rhythmic contractions of the heart result in pulmonary ventilation. Simultaneous ECG and lung cavity pressure measurements have shown that each contraction of the heart results in a pressure change in the lung cavity (D. J. Prior, unpublished observations). Although it has been suggested that ventilatory movements in quiet snails have no respiratory significance (Krogh, 1941), ventilation may be involved in the overall response to the increased O2 demand associated with locomotion.

The responses of Limax to changes in body hydration include the initiation of discrete behavioural patterns and the modification of behavioural responsiveness. In slugs, which have neither external shells nor water reservoirs, such as pallial water, (Smith, 1981 ; Blinn, 1964) for protection against dehydration, the demand for effective behavioural responses is great. The view that there is simply a decrease in the activity of slugs in dry conditions is misleading. Instead of just a ‘general reduction in activity’, there occurs an orderly modification of several behavioural activities that together serve to minimize the effects of the dehydration stress. These modifications can be regarded as an active change in the behavioural pattern that is properly matched to the particular environmental condition.

This work was supported by The National Science Foundation, an Alfred P. Sloan Foundation Fellowship and The Whitehall Foundation. D.J.P. is a recipient of a National Institutes of Health R.C.D.A. Fig. 1 was drawn by Ilyse Atema. This is contribution 191 from The Tallahassee, Sopchoppy and Gulf Coast Marine Biological Association.

Andersson
,
B.
(
1978
).
Regulation of water intake
.
Physiol. Rev.
58
(
3
),
582
603
.
Bailey
,
T. G.
(
1971
).
Osmotic pressure and pH of slug haemolymph
.
Comp. Biochem. Physiol.
40A
,
83
88
.
Blinn
,
W. C.
(
1964
).
Water in the mantle cavity of land snails
.
Physiol. Zool.
37
,
329
337
.
Burton
,
R. F.
(
1964
).
Variations in the volume and concentration of the blood of the snail, Helix pomatia L., in relation to the water content of the body
.
Can. J. Zool.
42
,
1085
1097
.
Canzlani
,
G. A.
&
Canata
,
M. A.
(
1980
).
Water balance in Ceratophrys ornata from two different environments
.
Comp. Biochem. Physiol.
66A
,
599
603
.
Carley
,
W. W.
(
1978
).
Water economy of the earthworm Lumbricus terrestris L., coping with the terrestrial environment
.
J. exp. Zool.
205
,
71
78
.
Cook
,
A.
(
1981
).
Huddling and the control of water loss by the slug, Limax pseudoflavus Evans
.
Anim. Behav.
29
,
289
298
.
Dainton
,
B. H.
(
1954a
).
The activity of slugs. I. The induction of activity by changing temperatures
.
J. exp. Biol.
31
,
165
187
.
Dainton
,
B. H.
(
1954b
).
The activity of slugs. II. The effect of light and air currents
.
J. exp. Biol.
31
,
188
197
.
Edney
,
E. B.
(
1977
).
“Water Balance in Lond Arthropods
.
Berlin
:
Springer-Verlag
.
Fitzsimons
,
J. T.
(
1972
).
Thirst
.
Physiol. Rev.
52
(
2
),
468
561
.
Fitzsimons
,
J. T.
(
1979
).
The Physiology of Thirst and Sodium Appetite
.
Monographs of The Physiological Society, No
.
35
,
Cambridge
:
Cambridge University Press
.
Gatz
,
R. N.
,
Crawford
,
E. C.
, Jr
. &
Purer
,
J.
(
1974
).
Respiratory properties of the blood of a lungless and gill-less salamander, Desmognathus fuscus
.
Resp. Physiol.
20
,
33
41
.
Greca
,
D. S.
(
1982
).
Central and peripheral cardiac regulation in the pulmonate mollusc, Limax maximus
.
Ph.D. dissertation
,
University of Kentucky
.
Howes
,
N. H.
&
Wells
,
G. P.
(
1934a
).
The water relations of snails and slugs. I. Weight rhythms in Helix pomatia, L
.
J. exp. Biol.
11
(
4
),
327
343
.
Howes
,
N. H.
&
Wells
,
G. P.
(
1934b
).
The water relations of snails and slugs. II. Weight rhythms in Arion ater L. and Limax flavus L
.
J. exp. Biol.
11
(
4
),
344
351
.
Krogh
,
A.
(
1941
).
The Comparative Physiology of Respiratory Mechanisms
.
University of Pennsylvania Press
.
Lane
,
N. J.
&
Treherne
,
J. E.
(
1972
).
Accessibility of the central nervous connectives of Anodonta cygnea to a compound of large molecular weight
.
J. exp. Biol.
56
,
493
499
.
Mcalister
,
R. O.
&
Fisher
,
F. M.
(
1968
).
Responses of the false limpet, Siphonaria pectinata Linnaeus (Gastropoda, Pulmonata) to osmotic stress
.
Biol. Bull. mar. biol. Lab., Woods Hole
134
,
96
117
.
Machin
,
J.
(
1975
).
Water relations in Pulmonates
, Vol.
1
, (eds
V.
Fretter
&
J.
Peake
), pp.
105
163
.
New York
:
Academic Press
.
Mackay
,
A. R.
&
Gelperin
,
A.
(
1972
).
Pharmacology and reflex responsiveness of the heart of the giant garden slug, Limax maximus
.
Comp. Biochem. Physiol.
43
,
877
896
.
Mellon
,
Def.
&
Treherne
,
J. E.
(
1969
).
Exchanges of sodium ions in the central nervous system of Anodonta cygnea
.
J. exp. Biol.
51
,
287
296
.
Mirolli
,
M.
&
Gorman
,
A. L. F.
(
1973
).
The extracellular space of a simple molluscan nervous system and its permeability to potassium
.
J. exp. Biol.
58
,
423
435
.
Phifer
,
C. B.
&
Prior
,
D. J.
(
1982
).
Dehydration induced modification of feeding and its neural correlate in the slug, Limax maximus
.
Neurosci. Abst.
8
(
2
),
901
.
Prior
,
D. J.
(
1969
).
Observations on behavioural control of water balance and respiration in the terrestrial slug, Limax maximus
.
The Physiologist
22
(
4
),
103
.
Prior
,
D. J.
(
1981
).
Hydration related behaviour and the effects of osmotic stress on motor function in the slugs, Limax maximus and Limax pseudoflavus. In Adv. Phystol. Set
., Vol.
23
,
Neurobiology of Invertebrates
, (ed.
J.
Salanki
), pp.
131
145
.
Oxford
:
Pergamon Press
.
Prior
,
D. J.
(
1983
).
Hydration-induced modulation of feeding responsiveness in terrestrial slugs (in the press)
.
Prior
,
D. J.
&
Pierce
,
S. K.
(
1978
).
Response of Limulus cardiac ganglion neurons to osmotic stress. Am
.
Zool.
18
(
3
),
657
.
Prior
,
D. J.
&
Pierce
,
S. K.
(
1981
).
Adaption and tolerance of invertebrate nervous systems to osmotic stress
.
J. exp. Zool.
215
,
237
245
.
Roach
,
D. K.
(
1963
).
Analysis of haemolymph of Arion ater (Gastropoda: Pulmonata
).
J. exp. Biol.
40
,
613
623
.
Roots
,
B. I.
(
1956
).
The water relations of earthworms. II. Resistance to desiccation and immersion, and behaviour when submerged and when allowed a choice of environment
.
J. exp. Biol.
33
,
29
44
.
Runham
,
N. W.
&
Hunter
,
P. J.
(
1970
).
Terrestrial Slugs
.
London
:
Hutchinson University Library
.
Sattelle
,
D. B.
&
Lane
,
N. J.
(
1972
).
Architecture of gastropod central nervous tissues in relation to ionic movements
.
Tissue and Cell
4
(
2
),
253
270
.
Sattelle
,
D. B.
(
1973
).
Potassium movements in a central nervous ganglion of Limnaea stagnalis (L.) (Gastropoda: Pulmonata
).
J. exp. Biol.
58
,
15
28
.
Schuurmans-Stekhoven
,
J. H.
(
1920
).
Über die Atmung der Schnecken Limax agrestis and Helix pomatia
.
Tijdschr. ned. dterk. Vereen.
18
,
1
43
.
Smith
,
L. H.
(
1981
).
Quantified aspects of pallial fluid and its effect on the duration of locomotor activity in the terrestrial gastropod Triodopsis albolabris
.
Physiol. Zool.
54
(
4
),
407
414
.
South
,
A.
(
1965
).
Biology and ecology of Agriolimax veticulatus (Mull.) and other slugs: spatial distribution
.
J. anim. Ecol.
34
,
403
419
.
Treherne
,
J. E.
(
1980
).
Neuronal adaptations to osmotic and ionic stress
.
Comp. Biochem. Physiol.
67B
,
455
463
.
Wolf
,
A. V.
(
1938
).
Studies on the behaviour of L. terrestris to dehydration and evidence for a dehydration tropism
.
Ecology
19
,
233
242
.