Regulation of Transpiration in the Clover Mite Bryobia Praetiosa Koch (Acarina: Tetranychidae)

The rigorous limitation of evaporative water loss is essential for the survival of small arthropods normally exposed to low humidities. Such animals, particularly the Acarina, contain limited amounts of water that can be lost and a relatively large surface area from which to lose it. Evaporation is limited primarily by physical barriers in the exoskeleton, the chitin-protein complex of the endo- and exocuticles (Beament, 1961) and a thin surface layer of oriented lipid molecules (Wigglesworth, 1945 ; Beament, 1945). The intact lipid layer provides a high degree of waterproofing but is characterized by loss of effectiveness above critical temperatures or upon treatment with solvents (Beament, 1959).

In addition, active mechanisms in the respiratory system and within the living portions of the cuticle may further reduce and control water loss (Edney, 1957). Of these, control of spiracular opening has been shown by many workers to be highly effective for water conservation in insects. Mellanby (1935) greatly increased water loss in Tenebrio by preventing spiracular closure with excess CO₂, while it has recently been shown that the degree of opening of these organs is influenced directly by humidity in tsetse flies (Bursell, 1957) and by water balance in dragon flies (Miller, 1964). Browning (1954) and McEnroe (1961) showed the importance of the mechanism to ixodid ticks and to the spider mite Tetranychus telarius.

Another active principle, a restrictive mechanism in the general body cuticle, seems to be of secondary significance in most insects (Wigglesworth, 1945), but it was shown in ticks (Lees, 1946; Browning, 1954) to be equally as important as respiratory control. It has been called ‘active retention’ by Edney (1957) and others; and Lees’ (1947) hypothesis that the epidermal cells are able to secrete water inward seems to offer the best explanation, although it has not yet been experimentally confirmed. As the result of this proposal, Edney (1957) and Beament (1961) suggested that the mechanism might be one which would result in the uptake of atmospheric water at high humidities. It would remove the need for a different mechanism to explain this puzzling phenomenon which is known to occur in a limited number of arthropods, e.g. T. molitor (Buxton, 1930) and many ticks (Lees, 1946).

These active mechanisms in the cuticle and in the respiratory system may also contribute to at least one of the two forms of regulation that are apparent in the water balance of some hardy insects and arachnids. Certain insects are able to regulate their percentage of body water, maintaining it at essentially constant levels both with time and over a wide range of humidities, e.g. T. molitor (Buxton, 1930). In addition a very few animals, e.g. nymphs of the grasshopper Chortaphaga viridifasciata (Ludwig, 1937) and a locust Oedipha coeruliens (Jakovlev & Krüger, 1953), have been shown to regulate their water loss, keeping it at nearly constant rates over a substantial portion of the humidity range. The only specific studies on the mechanisms by which these regulatory functions are accomplished have been on the first type in which it has been shown, especially in Tribolium confusum and Dermestes vulpinus (Fraenkel & Blewett, 1944), that metabolism, and thereby the production of metabolic water, is adjusted to differences in humidity. More food is consumed in dry air than in moist and the extra water produced is enough to maintain the water level in the body. Each of these controls is such as to keep an internal variable at an essentially constant level in the face of changing external conditions, thus conforming to the usage of the term ‘regulation’ by Prosser & Brown (1961).

The clover mite Bryobia praetiosa is active on warm walls and tree trunks from late September to early June in Colorado, and the ability of these very small animals (< 1 mm.) to thrive at the low humidities commonly measured on such exposed surfaces prompted studies on their humidity tolerances and behaviour (Winston, 1963 a). These have been followed in the present work by the measurement of water loss in both living and dead mites over a wide range of humidities. Because an active mechanism was indicated by the earlier studies, it could be expected that dead mites would show a higher rate of loss than living ones throughout the humidity range, Furthermore, it seemed likely that a comparison of weight changes in living mites at various humidities would reveal the existence of mechanisms for the control and restriction of water loss and for the uptake of atmospheric water in different portions of the humidity range. Since preliminary studies on water loss in living mites (Winston, 1964) indicated the ability to regulate over a substantial part of the range, the nature and location of the underlying mechanisms was made the primary objective of this investigation.

The mite used in this study is the Bryobia praetiosa Koch of Anderson & Morgan (1958), and a brief description of its life-cycle in Colorado appears in Winston (1963b). For most of the winter and spring, a plentiful supply of these parthenogenetic animals from a natural population was maintained on bean plants growing on flats in the windows of the Laboratory. A ‘conditioning’ or pre-experimental treatment was adopted to assure a more uniform experimental animal than could be obtained directly from a wild population. Adult mites were taken from the leaves rather than from the windows, ensuring that the majority had fed for some time. They were then held without food or water for 48 hr. at 29% R.H. and 25°C. This allowed ample time for defecation of the copious, watery faeces and for most of the egg-laying. No faeces were found after tests that had been preceded by this conditioning and there were usually fewer than two eggs per mite.

Groups of sixteen mites were lightly anaesthetized, weighed as a group, and placed in screen-topped vials for exposure to air of a specified humidity. This treatment was carried out in a water-bath at 25 ±0·5° C. in small chambers in which the humidity was controlled by saturated salts as described by Winston & Bates (1960). The mites were exposed to humidities of o, 12, 29, 43, 52, 62, 76, 85, and 93%; none higher was used because the mites easily drown in condensed droplets and because their activity is greatly reduced at very high humidities. Because all the work was done at one temperature, relative humidity (R.H.) and saturation deficit are equivalent, and the former term will be used throughout this paper. After 24 hr., later 18, the mites were again anaesthetized, weighed, dried for 24 hr. at 90°C., and weighed again. These treatment periods were long enough to result in measurable weight-changes at high humidities, but short enough to reduce mortality to a minimum at the low. No differences in the calculated rates of loss were noted between the two testing periods. The water and dry-matter content of both feeding and conditioned mites were measured at intervals to check for any seasonal effect on these variables and none was found. Feeding animals were removed from the plants, anaesthetized, weighed and dried to constant weight. The conditioned animals were treated similarly except for the 48 hr. starvation period at 29% R.H. and 25°C. prior to weighing.

A correction factor of 1·5 μg. each was obtained for the few eggs laid during the tests by calculating the volume of an egg from the diameter and assuming the density to be about 1·02. Measurements of many eggs with an ocular micrometer showed them to be of nearly uniform size, so a reasonably accurate correction could be made for the loss of weight by counting the number of eggs laid. For the dry weights a correction of 0·4 μg. per egg was used. These factors were multiplied by the number of eggs laid per mite in each group and the result added to the average weight. Any water loss resulting from the process of oviposition was so slight as to be undetectable with the methods employed.

Ether was used instead of other standard anaesthetics throughout these experiments because its ill-effects were much less than any others tested. The possible effect of this lipid solvent on the water-proofing layer was of some concern to us, but tests indicated that stronger doses of the anaesthetic than that used in practice did not increase mortality even in dry air.

Water loss in dead mites was measured after they had been killed by exposure for 2 hr. to hydrogen cyanide fumes in a water-saturated atmosphere. The greatly increased rate of loss at death made it necessary to expose these animals for only 4 hr., to prevent low water content and decay from influencing the results.

For measurement of the effects of high concentrations of CO₂ on weight loss the mites were exposed to atmospheres containing 10% of the gas at various humidities. An approximation to the lowest concentration which would produce a maximum effect was obtained by determining the water loss in 5, 10, and 15% mixtures of the gas in air at one humidity. The two higher concentrations produced essentially the same loss while at 5% it was significantly lower; the 10% mixture was therefore used throughout the tests. No anaesthetic effect was observed at these concentrations. Humidity in this series of experiments was controlled by appropriate concentrations of sulphuric acid (Solomon, 1951) because, unlike many of the saturated salts, these solutions absorb very little CO₂.

A Cahn Electrobalance, model Mio, was modified to obtain a sensitivity of less than 1 μg. with a range of 1 mg. Changes in weight of individual mites were quite small, and to obtain greater accuracy we weighed them in groups of sixteen and calculated the average loss per mite. Standard weights approximating the weight of a group of mites were weighed frequently to check for drift in the balance circuit. Single adult mites averaged 30 +11μg. in weight; and, to counteract the wide variations found, groups were made up of animals as close to the same size as possible.

Standard deviations and standard errors of the means of all values were calculated by the methods of Arkin & Colton (1955) and the significance of the difference between means of adjacent values was tested by the use of the Z-test of Fisher (1950).

In these experiments weight loss was determined on 560 groups of 16 mites, about 9000 individuals. Each group was weighed at least twice, and most three times, to obtain an initial weight, a final weight, and a final dry weight.

The water content of the mites was calculated as a percentage of the final total weight and plotted against R.H. in Fig. 1. The freshly fed animals have a much higher water content than that of conditioned ones, a result to be expected from the large amount of water in the plant cells on which they feed. Much of this water is lost rapidly soon after feeding, but the remaining water is only slightly reduced during the subsequent treatment periods of 18 or 24 hr., indicating a form of regulation in relation to time. It can also be seen that there is very little effect of R.H. on the water content of control mites throughout the humidity range. They are thus able to regulate the percentage of body water despite different rates of water loss over a wide range of humidities.

The water content of CO₂-treated mites falls below that of the controls (Fig. 1) and is essentially proportional to the R.H. AS the dry weights are the same in both groups, the reduction in water is probably due to loss through the opened spiracles, which would indicate that the mechanism for regulation is based in the control of these structures. In addition, the differences in rate of loss of solids in control mites were so slight (Fig. 4) that it does not appear possible that maintenance of a constant water content could be the result of adjustments in metabolic rate in response to water loss. This problem is really tangential to the present one of regulation of water loss, however, and further study must be deferred for the present.

Though the object of this study was to determine water loss in groups of mites, it was obviously possible to obtain only the loss in total weight. The final weight was easily resolved into solids and water by obtaining the dry weight and then subtracting this value from the total to get the water. For the initial weight, however, these had to be calculated by using the percentage of water and of solids in the conditioned mites. Water and dry-weight losses were then obtained by subtracting the final from the initial values. The results of these calculations are plotted in Fig, 4 and show that water loss follows total weight loss quite faithfully, and that the latter is not affected by the slight changes in dry weight. We will, therefore, use both terms interchangeably in the remainder of this paper even though the values presented will actually be of total weight loss. These animals vary so much in size that only relative comparisons are possible, and because of this we have expressed weight changes as percentages of the initial total weights, in percentage per hour per mite.

There are four avenues for the escape of water in Bryobia-. urine and faeces, oviposition, cuticular transpiration, and spiracular loss. In our work the 48 hr. conditioning period eliminated urine, faeces, and most of oviposition as factors; and only water loss through the cuticle and through the spiracles remained to account for that which was measured.

The nature of the barriers to transpiration from these two areas was studied by the measurement of water loss in living mites, in those killed by HCN gas, and in others killed in chloroform vapour (Fig. 2). As the lowest rates of loss were from living mites, rates above these can be considered due to the removal or disruption of one or more of these barriers. By far the highest rates of loss were from the chloroformtreated animals, demonstrating changes in the lipid waterproofing layer. Gibbs & Morrison (1959) recently showed the same kind of layer for another tetranychid, the spider mite Tetranychus telarius. In both this species and Bryobia the cuticle underlying the wax layer is so thin that it is difficult to see how it could be much of a barrier, but Beament (1961) has shown that, though very permeable, such a layer would reduce evaporation considerably below that from a free-water surface. It can be seen that the passive barriers provide the major reduction of evaporation, but the substantially higher rates of loss in dead mites over those in living ones suggest the presence of an active mechanism which further restricts water loss.

The respiratory system is considered to be a primary site of evaporative water loss in many terrestrial arthropods (Wigglesworth, 1953), and the relatively simple technique needed to demonstrate this offered an obvious starting-point for the problem in Bryobia. Since CO₂ at concentrations of 5−10% is known to maintain the spiracles in an open position in most arthropods (Wigglesworth, 1953) including ticks (Browning, 1954), the rates of loss of CO₂-treated and untreated mites were compared (Fig, 3). Tests at five humidities indicated that a CO₂-sensitive mechanism, presumably control of the respiratory openings, would account for about half the difference in the rate of water loss between live, untreated animals and dead ones; thus about half the loss is considered to be through the spiracles and half through the cuticle. Lees (1946) and Browning (1954) found a similar relationship in ticks.

There is no doubt of the existence of organs acting as spiracles in clover mites and many other Acarina, but little is known of their behaviour and controls (Winston, 1964). Bryobia is usually very active in dry air and one would expect this to result in high rates of water loss through necessarily opened spiracles. There is higher loss at 0% R.H. than in more humid air, but it is so small that these animals can survive for relatively long periods under dry conditions. It must be that they, like the tsetse fly (Bursell, 1957), can obtain enough oxygen for activity through nearly closed spiracles and thus make gas exchange subordinate to control of water loss in these organs. It has been reported that the respiratory organs of spider mites are exposed to the air to a greater or lesser extent in response to changing levels of activity (Blauvelt, 1945), humidity, and water content (McEnroe, 1961). Observations in this laboratory have failed, however, to show any visible positional changes of these organs in Bryobia that would indicate a control function. They have been seen to be exposed only when the mites were actively feeding.

Since the differences between living and killed mites cannot be explained wholly on the basis of a CO₂-sensitive respiratory mechanism an additional one is indicated.

This is most likely some form of active retention of water by the general body cuticle, though Beament’s (1961) suggestion of a rapid change in the cuticle at death still remains a remote possibility.

The better to show important deflexions in the curves for water loss, the rates for living mites are given on a larger scale in Fig. 4. The general tendency of the curve, as shown by the broken line between 0% and 85% R.H., indicates an overall inverse relationship between water loss and R.H. The deviation from this straight line between 53% and 85%, however, indicates that the mites are able to modify the relationship in this part of the range. Between these humidities, loss is maintained at a nearly constant rate, showing that the mites can regulate water loss in the face of considerable differences in the evaporating power of the air.

Despite the slight influence of humidity in the zone of regulation, the differences between adjacent means were found to be not significant to the 0-05 level, though that between the two extremes was close to it. At humidities above and below this region the differences between adjacent means were found to be highly significant, at least to the 0·01 level. It is to be noted that cyanide-treated mites do not exhibit this plateau, and so it can be assumed to be the result of an active process. The sharp breaks at either end of this zone of regulation are indications that some distinct changes in the active process take place between 85 and 93% and between 43 and 53% R.H.

In contrast, the greater slope of the line between 43 and 0% shows that loss depends more on the humidity, but this slope is still much less than was found in dead mites. Thus the sudden increase in rate of loss from 53 to 43% is not to a condition which is completely dependent on the evaporating power of the air. Instead, there is evidence of an active mechanism functioning at low humidities that restricts, but does not regulate, water loss.

The apparently insignificant effect of R.H. on dry-weight losses (Fig. 4) does not support the hypothesis that energy is expended for the retention of water, as any change in respiratory activity should produce a corresponding change in the loss of solids. Other evidence for an active principle is too strong to be ignored, however. Furthermore, it is quite probable that the amount of energy used is too small to be shown by measuring the very slight changes in solid matter in these animals.

The mechanism for this regulation could be based either in the respiratory system or in an active component of the cuticle. If it were in the cuticle the regulatory pattern should still be evident after CO₂ treatment, but if it were in the spiracles, which presumably were wide open, the line should show inverse proportionality to the humidity over its entire length. It is evident that the line for average rates of loss in the treated mites (Fig. 3) shows the same deviation as that for the control animals, indicating a zone of regulation. This is good evidence that the mechanism is indeed a part of the cuticular activity rather than being based in the respiratory system. Treated animals showed a much greater water loss than did the controls, but this is to be expected when the spiracles were kept open. Moreover, if this were a spiracular mechanism, one would not anticipate the sharp breaks at either end of the zone of regulation found in untreated mites. Spiracular control would probably require hygroreceptors of some sort, and such abrupt changes are not characteristic of receptors in general.

The restriction of water loss by active work of some mechanism in the cuticle, possibly the epidermal cells, has been shown to be an important factor in only a few other animals, notably in several species of both ixodid and argasid ticks (Lees, 1946). This mechanism is thought to be of only slight importance in most insects (Wiggle-worth, 1945), and it may be that it is more significant in the Acarina because of the relatively greater surface area for water loss in these small animals. Little has been done on the very small hardy insects, however, and this approach might lead to other ideas on the subject.

It would be convenient to assume that the active cuticular restriction of loss below 50% R.H. and the regulation above it are part of the same mechanism, if for no other reason than simplicity. There is little or no evidence for either one mechanism or two at this time, and a second ought not to be assumed if it is not necessary. This can be better justified by postulating that above 50% R.H. the mechanism can regulate water loss, causing the deviation from proportionality to humidity that we have shown. In all humidities below this zone of regulation, though, it would be working at the same rate, its full capacity. Hence, differences would be due to variations in the drying power of the air. The mechanism would only be restricting transpiration in the lower humidities.

The mechanism by which regulation and restriction of water loss are accomplished is unknown, but Edney’s (1957) and Beament’s (1961) suggestions, that it might be the same as the one that produces active uptake of atmospheric water at high humidities in some arthropods, should be considered. Beament (1954) hypothesized that uptake is based on the active transport of water; and, to follow this same line of thought, regulation and restriction would then be based on active transport. If this were so, one would expect a sharp change such as that between 43% and 53% R.H., indicating a breakdown of the mechanism at lower vapour pressures. It would mean that the regulation of transpiration at humidities below 50% R.H. requires more energy than is available to the system, and the rate of loss rises at 43 % to become proportional to the evaporating power of the air. Such an abrupt change between these two humidities is also typical of several other aspects of the humidity relations of B. praetiosa, especially of survival (Winston, 1963 a); and Edney (1945) and Beament, Noble-Nesbitt & Watson (1964) found that the lower limits for uptake were near 50% R.H. in prepupae of rat fleas and in the common firebrat. It is possible that these changes are the result of the breakdown of a mechanism common to many species.

Though regulation is apparent, we have been unable to demonstrate active uptake in the clover mite by any of the techniques used successfully on other acarines and insects. One might expect to find it in this mite, though it is the first tetranychid to be studied in this way, because the other Acarina which have been properly tested have shown the phenomenon quite readily—for example, ticks of several species (Lees, 1946 ; Belozerov & Seravin, 1960), grain mites (Knülle, 1962; Solomon, 1962), spiny rat mites (Wharton & Kanungo, 1962) and rabbit ticks (Camin, 1963). None of these is as well adapted to dry air as are clover mites, however, except for some of the argasid ticks, and it may be that uptake is not present in some hardy forms such as Bryobia. Water loss is slow enough to make it possible for them to replenish their water supply by drinking or feeding well before their water content has dropped to acute levels, and uptake would not be of any particular survival value. Thus the function may have either been lost completely or masked so that the usual techniques would not reveal it.

To show that there is regulation of water loss one must expose the experimental animals to a wide range of humidities, and very few workers have used this approach. Of these, only Ludwig (1937) and Jakovlev & Krüger (1953), working with two orthopterans, were able to demonstrate regulation. It is probably only because so few animals have been tested in this way that the regulation of transpiration is not more commonly known among terrestrial arthropods.

Such a regulatory system would be a major factor in the maintenance of a stable internal environment over a major portion of the humidity range. Even though many arthropods are apparently able to withstand wide variations in their internal medium, there must be an advantage to having as much constancy of the blood as possible. The development of homeostatic mechanisms has usually brought advances for the fortunate species that had them, and, in general, fluctuating humidities represent much more of a stress factor for small animals than for large ones. This ability to control and restrict evaporative water loss in all but the highest humidities, coupled with broad temperature tolerances (Anderson & Morgan, 1958), may be the primary factors that make it possible for these mites to be almost the only arthropods active in numbers above the soil surface during the winter months when competition and predation are essentially nil. They carry on their life-cycle in normal fashion during a long period of the year in which daytime humidities in this region may range down to 5% or below and in which temperatures commonly vary between 25° and — 10° C. They are to be found on sun-warmed walls all during this period, limited in their activity by low light intensities and extremely low temperatures.

The humidities encountered by this mite over most of its distribution in the North and South Temperate Zones (Morgan, 1960) are almost always within or above the zone of regulation. Presumably, they evolved under such conditions where life would seem to be easiest for them from the standpoint of humidity. There is no question, though, from the data presented in this paper, that the mites can cope without difficulty with the occasional periods below 50% R.H. which would be experienced. Drier areas, such as this one at Boulder, Colorado, where the humidity during the day is almost always below the zone of regulation and the evaporating power of the air is high, represent nearly marginal habitats that must tax their control mechanisms considerably. Nevertheless, they are able to thrive under such conditions.

The authors wish to acknowledge support of this research by the University of Colorado and by Research Grant no. G14517 to the senior author from the National Science Foundation.