4. DISCUSSION
Seat patch water potentials were related to terrestriality (including arboreality) in the six frog species studied, with species that were more aquatic having higher (less negative) seat patch water potentials than those that were more terrestrial or arboreal. Similarly, differences between water potential of the seat patch and of the blood were related to terrestriality insofar as the aquatic species and the terrestrial R. pipiens had different osmotic potentials of the blood in comparison with their seat patch water potentials, but the terrestrial toad and arboreal species had osmotic potentials of the blood that were similar to those of their seat patches.
The basic physiology of water uptake through the ventral skin of amphibians has been studied extensively (reviewed by Jørgensen 1997b) working under the assumption that blood osmotic water potential drives water exchange. Thus, the prevailing paradigm neither explains how aquatic frogs avoid taking up too much water, nor does it explain the disequilibrium between the seat patch water potential and blood water potential in species that live in or near water (Fig. 3). Studies of water uptake in amphibians have emphasized the behavioral, physiological and ecological mechanisms that allow “thirsty” terrestrial frogs to take up water rapidly (Tracy 1976; Jørgensen 1997b) with little emphasis on how frogs in aquatic environments avoid taking up too much water.
Our results suggest that species that are more aquatic appear to control their seat‑patch water potentials to be higher (less concentrated) than the osmotic potential of their blood, but species that are more terrestrial appear unable to do so. This ability to control the water potential of the seat patch is potentially beneficial to aquatic species, as it would allow them to limit the influx of water. Unrestricted water influx into frogs that spend considerable time exposed to water could create metabolic and electrolytic costs associated with ridding the body of excess water.
Terrestrial species face challenges to balancing their water budgets different from challenges faced by species that are primarily aquatic. For example, many terrestrial anurans absorb water from moist soils, and may be the primary source of water for some terrestrial species. The water potentials of soils depend upon soil composition and water content of the soil (Rose 1966; Tracy 1976). Thus, terrestrial frogs would benefit from having a lower (more concentrated) seat-patch water potential because that would allow absorption of water from soils with a wider range of water potentials, including relatively dry soils.
Although circulatory changes are associated with water uptake through the seat patch (Viborg and Hillyard 2005; Viborg et al. 2006), it seems unlikely that circulatory adaptations alone could account for the disequilibrium between the seat patch and blood water potentials that we observed in the more aquatic species. The mechanisms by which frogs could regulate their seat patch water potential may involve hormonal control of water channel proteins called aquaporins (Preston and Agre 1991) that are located in the bladder, in the tissues of the ventral skin, and possibly around the capillaries that connect the seat patch with the rest of the body’s circulatory system. A number of different types of aquaporins have been described from various amphibian tissues, and there are differences among species with respect to aquaporin types and expression (Suzuki et al. 2015). We suggest that the conductance of the walls of the capillaries that connect the seat patch with the rest of the body’s circulatory system can change in response to hormonal signals acting on aquaporins. This change in conductance regulates water uptake into the body’s circulatory system. Furthermore, this regulation can, under some conditions (see below), result in a disequilibrium between the water potentials of the seat patch and the blood.
The rates of water exchange between a frog and its environments can be modelled with the equations in Fig. 4 (Tracy 1976; Tracy and Rubink 1978; Tracy 1982). If all properties, or system states, of EquationA (viz., A v, K sp,Ψ sp) are constant, then equation Adescribes a straight line where the slope of the line is (A v * K sp), and the x‑intercept is where Ψ sp =Ψ en. However, in experiments in which the environmental (sucrose) water potential was a variable, our results (in Fig. 1) indicate that the lines, for most species, are not uniformly straight lines, but that they change slope relative to environmental water potentials. Thus, K sp is a constant at more negative environmental water potentials (closed circles in Fig. 1), but in environments with osmotic potentials closer to pure water there was a variable conductance of the frog (K sp) sufficient to offset the changing environmental (sucrose) water potential, leading to the observed leveling of the relationship between water uptake and environmental water potential (open circles in Fig. 1). It seems likely that this pattern is due to changes in aquaporin activity that prevents uptake of excess water in dilute environments. This indicates that water potential of seat patches (likely due to hormonally-controlled aquaporin activity), and hydric conductances of seat patches are variables that appear to act differently among different species, related to their ecological habits.
If water uptake rates are slowed in environments with higher water potentials (i.e., water potentials closer to that of pure water) by the actions of aquaporins located between the seat patch and the circulatory system (Equation B in Fig. 4), this would provide a mechanism that would explain the data designated by open circles in Fig. 1. Five of the six species (all except P. cadaverina ) maintained a relatively similar water uptake rate while immersed in a range of high water potentials. Thus, if the flow of water from the seat patch to the circulatory system was restricted by the controlling action of aquaporins, the seat patch water potential (Ψ sp) would be high and similar to the environmental water potential, but the water potential of the blood would remain low (i.e., the blood would maintain its osmotic concentration) rather than becoming excessively dilute. This (possibly aquaporin‑mediated) regulation would be particularly beneficial for amphibians that spend a lot of time in water, and is consistent with our results (Figs. 1, 2 and 3) that show that the more aquatic species have fluids in the seat patch that are significantly more dilute than blood. Such a disequilibrium cannot be explained simply by the water potential gradients between the seat patch and the blood, but it could result from an aquaporin-mediated decrease in hydric conductance between the seat patch and blood (K bl). As mentioned above, there is evidence that the influx of water across the skin is independent of cutaneous blood flow (Viborg and Hillyard 2005; Viborg et al. 2006). However, Burggren and Viborg (2005) have raised the possibility that blood flow may play a role in regulating water influx by changing the “functional” surface area (i.e., Abl in Equation B in Fig. 4). Thus, it is possible that blood flow may play a role in regulating rates of water uptake in addition to changes in conductance (by way of aquaporins).
Word and Hillman (2005) provide evidence that water taken across the skin of cane toads (Bufo marinus ) is taken directly into the capillaries rather than accumulating in lymphatic spaces. As we have shown in Fig. 3, for the boreal toad, the water potentials of the blood and seat patch are not different. Thus, there is no barrier preventing water being taken into the capillaries, so we would not expect an accumulation of fluids in the lymph. Therefore, the result for cane toads is wholly consistent with our results for boreal toads and with the model in Fig. 4. However, our model predicts that there would be an accumulation of dilute fluids in the ventral tissues in aquatic frogs such as Xenopus – a hypothesis that could be tested using the techniques of Word and Hillman (2005). An alternative hypothesis would be that the relatively impermeable skin of Xenopus could be explained by a lack of AVT-stimulated aquaporins in this species. However, the lack of an AVT response does not explain what mechanism would prevent osmotic dilution of blood over time. More importantly, in the context of the results shown in Fig. 1, a generalized low permeability of Xenopus skin might explain the low rate of water uptake, but it would not explain the less negative water potential (compared to the other species in our study) represented by the 0 uptake point on the X-axis intercept. This less negative water potential suggests that there is an accumulation of dilute fluid inXenopus . Thus, the aquaporin-mediated model in Fig. 4 is consistent with both the ability of terrestrial frogs to take up water from a relatively dry environment, and the ability of aquatic frogs to prevent an excessive water influx from a fresh water environment.
What currently is known about the types and roles of aquaporins in frog skin, bladders, and kidneys has been reviewed by Suzuki et al. (2007), Suzuki & Tanaka (2009), Ogushi et al. (2010) and Suzuki et al. (2015). These reviews, based on data from relatively few species of frogs, present a mechanistic aquaporin model for the transport of water across frog epidermis that involves hormonal control of three or more different aquaporin types. Although the model describes the transport of water from the environment across the epithelium, it does not explain the control of water transport into the sub-epidermal capillaries that link the seat patch fluids with the circulatory system of the body. Thus, the aquaporin model for cutaneous water uptake is incomplete. Much like the historical paradigm of water uptake that assumed that osmotic water potential of blood drives water exchange, the current aquaporin model explains how terrestrial frogs take up water, but it does not explain how aquatic species avoid taking up too much water. To address this fundamental question, we need to know more about aquaporin types from a wider range of species, and more about the mechanisms that control their levels of expression. Furthermore, to achieve a comprehensive understanding of frog water balance, the physiological and molecular studies of aquaporins should be designed within an ecological context by carefully selecting the species studied.
Evaluating the mechanisms by which the biophysical variables illustrated in Fig. 4 change to produce an osmotic disequilibrium in different parts of a frog is beyond the scope of this report, but it represents an exciting and logical next step in investigating the physiology of water uptake in frogs. An integrated approach involving physiological, molecular and ecological principles is needed to understand the evolutionary adaptations for water exchange between frogs and their environments.