Metabolic rates and thermoregulatory characteristics of Akodon azarae ( Rodentia : Sigmodontinae )

In this work, we evaluated the energetic status, activity daily metabolic rate, thermoregulatory capacity and the existence of torpor in Akodon azarae (Fisher 1829). Basal metabolic rate (BMR) was similar to the expected value from Kleiber' s, and also according to the BMR of phy logenetically related species. No statistical difference in BMR between sexes was detected. For both males and females, activity daily metabolic rate was 235 % of BMR. Body temperature was not statistically different, neither at different ambient temperatures nor between sexes. In the thermoneutral zone, body temperature was 36.1 ± 1.5 ac. Thermal conductance was lower than the expected for mammals whereas thermoregulatory capacity was 136 % of the expected. Akodon azarae did not show evidence of torpor.


INTRODUCTION
Endothermy is very expensive (McNab 1974) and endotherms expend most of the energy consumed to produce heat.Endotherms have sufficient fur insulation and thermogenic capacity to survive at low temperatures, but they need a regular food supply to do so.In consequence, endothermy allows mammals to maintain high body temperatures and independence of environmental conditions, but it becomes prohibitive in both unproductive and/or cold habitats (French 1992).
Body size measured as body weight (W), is the physical translator between the environment and the organism.Consequently, it constitutes the primary factor that determines the basal metabolic rate (BMR) and thermal conductance of individuals (C).Most vertebrates follow the Kleiber (1961) relation down to a "critical" W, below which the scaling of BMR must be changed to ensure the maintenance of endothermy.According to McNab (1983 ), the minimal W for endothermy in species that have a level of BMR similar to the Kleiber relation, is namely about 20 g.Contreras (1989), reported that, the minimal W to have a level of BMR similar to the Kleiber, is above 50 g.Therefore, a combination of a small body mass and a low BMR would cause animals enter in torpor.Kleiber's equation has been discredited as a reference curve (Hayssen & Lacey 1985).
(Received May 17, 1999; accepted October 13, 1999; managed by F. Bozinovic) However, Kleiber's equation has been used countless quantity of times in most of the research on energetics, for which it has a great comparative value.McNab (1988) discussed that Kleiber's recommended curve have an acceptable level.Moreover, Bozinovic & Rosenmann (1988b) confirm the universality of the ( -0.25) massspecific allometric value reported by Kleiber and favor its use as a standard reference.On the other hand, independent of the influence of W, in mammals, food habits exert an important effect on BMR (McNab 1986).
Some endothermic species, when facing low temperatures, reduce substantially their heat production, entering torpor and thus balancing their total energy requirements.Thus, during daily torpor, animals drop their body temperatures between approximately 1 0°C and 25 oc and torpor bouts of less than 24 h (Hudson 1973 ).This mechanism, which allows species to inhabit both unproductive and/or cold areas, has been found in five species of South American rodents (Bozinovic & Contreras 1990, Caviedes-Vidal et al. 1990, Bozinovic & Marquet 1991).Bozinovic & Rosenmann (1988b ), based on preliminary data set, hypothesized that other South American rodents such as Akodon azarae (Fisher 1829), Calomys ducilla, C. musculinus and Eligmodontia typus would show evidence of torpor at low ambient temperatures.
Other energetic studies have been accomplished on South American rodents (Caviedes-Vidal et al. 1987, Bozinovic & Rosenmann 1988a).For A. azarae, Dalby & Heath (1976) obtained BMR values on a single observation, not allowing any type of statistical comparison on BMR and thermoregulatory capacity nor between the sexes.In the present study, BMR, activity daily metabolic rate, thermoregulatory characteristics, and the existence of torpor in Akodon azarae are evaluated.For small omnivores as A. azarae an intermediate BMR (90 to 110% of Kleiber) would be expected.Furthermore, rodents as A. azarae, which tolerate low temperatures during winter having both, low BMR and small body mass, should show torpor or low thermal conductance.

Animals
Animals of both sexes (26 females and 27 males) were collected using Sherman live traps at Necochea, Buenos Aires Province, Argentina (38°29' S, 58°50' W).This region is known as the "Pampeana" biogeographic province (Cabrera & Willink 1973 ).Akodon azarae is a mouse of 24 g, strongly associated to natural grasslands, particularly to open formations (Bonaventura et al. 1992, Redford & Eisenberg 1992).This species, which is the most numerically representative rodent of the pampas grasslands, is found from southern Brazil to central Argentina (Redford & Eisenberg 1992).
Captured rodents were carried to the laboratory and housed individually in animal cages (0.3 x 0.22 x 0.15 m).Wood shaving and cotton for nesting material were placed on cage floors.All animals were maintained on a natural photoperiod (L:D = 10:14).Ambient tempeniture (T.) ranged from 19 to 25 oc, whereas relative ambient humidity ranged from 50 to 70 %.Animals were fed with pellets ad libitum (composition in %: minimum protein: 21, maximum fiber: 4.5, minimum fat: 8, average calcium: 1.8, phosphorous: 1.1, maximum ashes: 8).Tap water was provided ad libitum.

Energetics
The energetics variables were individually estimated by measurements of oxygen consumption (V0 2 ) in metabolic chambers at different T.'s using a modified closed automatic system, based on the design of Kalabukhov-Skvortsof (Gorecki 1975).The metabolic chamber was a double wall aluminum cage with polyurethane between walls.Animals were put inside the chamber, which was closed with a 20 mm thick acrylic window-door to observe the animals regularly.A thermocouple was used to measure temperature inside the chamber.Heat was interchanged with the outside chamber automatically by two Pelttier intercoolers (Melcor, model CP-1.4-127-061)connected to a PC and controlled by software.Ambient temperature inside the chamber was maintained with an accuracy of 0.1 °C.An internal and an external pressure sensor monitored air pressure.Oxygen influx was assessed through two electrovalves.Pressure sensors and electrovalves were connected to a PC and controlled by a software.C0 2 and Hp were removed by ascarite and drierite.In order to achieve a post-absortive state, metabolic rates (MR) were measured in 24 h food deprived animals.Animals were weighted before each V0 2 measure.V0 2 was always measured in the morning.Data were discarded when activity was observed.To minimize the effect of stress, vo2 records during the first 30 min were discarded.Basal metabolic rate was estimated from the minimal metabolic rate when it showed independence of T..For measuring activity daily metabolic rates (ADMR), animals were disposed into the respirometer during 24 h with food at thermoneutrality, provided with a treadmill.
V0 2 is reported as mass-specific metabolic rate (mllgh).All results are presented as means± standard deviation (SD).All measures of V0 2 were corrected to standard temperature and pressure of dry air.Nickerson et al. (1989) method, that identifies the best fit for a continuous n-phase regression, was used to represent the relationship between T and MR.This method estimates the a point at which the relationship between independent and dependent variables change (i.e.threshold point).The best n-phase regression model was selected by comparing from, a three-phase regression models to a simple regression model.

Body temperature and thermal variables
Body (rectal) temperatures (T b) and ambient temperature (Ta) were measured using a YSI probe (model93k73545-402) connected to a Cole-Parmer instrument thermistor (model8402-l0).Body temperature, was recorded at the beginning (T bb) and at the end (T be) of each V0 2 measurement by inserting a thermocouple 2 cm into the rectum of the mouse for a period no longer than 30 s. Decrease in Tb after each V0 2 measurement was estimated as DT b = T be -T bb' The slope of the first phase of the best fitted n-phase regression that related MR and T was used as an estimate of minimal thermal

Daily torpor
Since in the field sampling, during winter, mortality of 80 % was observed when animals were in the traps for more than 4 hat Ta lower than 10 °C, V0 2 was measured in individuals at 15 oc during 4 h.

Statistics
ANOV A was used both to test the null hypothesis of no differences among metabolic rates at different T between sexes and to test the null a hypothesis of no differences among DTb neither at different Ta nor between sexes.A posteriori Tukey's test was used to identify differences among DT b' both at different T a and between sexes.
A t-test was used either to evaluate the null hypothesis of no differences between observed BMR and the expected one by Kleiber's equation; to evaluate the null hypothesis of no differences between observed BMR and the expected one by Hayssen & Lacey's equation (1985); to test the null hypothesis of no differences in ADMR between sexes; to test the null hypothesis of no differences between the BMR observed and the expected one by McNab's equation on the euthermic limit; and finally for animals reared in animals room to test the null hypothesis of no differences in T b between sexes.A t-test for comparisons between two slopes was used to test the null hypothesis of no differences between observed and expected Cm;n• An analysis of sensitivity was used to solve the possible meaningless of SD values of estimated BMR and cm;n' since these parameters reflect the variance of W between the animals.

Metabolic measurement
No statistical differences among MR of males and females were detected at any Ta (F = 0.71, n = 70, df = 7, P > 0.5).To analyze the relationship between MR and T a' data from both sexes were pooled.Mean MR was lowest at 35 oc (best fit: was a two-phase lineal regression, break point = 35 °C, Fig. 1, Table 1 ).Estimated BMR was not statistically different to the expected for a 22.5 g mammal neither by the Kleiber relation nor by the Hayssen & Lasey relation (Table 2).Sensitive analysis of the observed/expected BMR of A. azarae, shows that a bias of 347 %from Kleiber (t obs = t crit = 2.57, BMR expected = BMR observed± 0.57, df = 5, P < 0.001) and a bias of 435 % from Hayssen & Lacey (t obs.= t crit.= 2.57, BMR expected = BMR observed ± 0.57, df = 5, P < 0.05) would be necessary to alter the latest conclusion.
The relation between V0 2 and T. was (Fig. 1  Statistical comparisons between either, one phase; two phases and three phases lineal regression model that related rate of metabolism and ambient temperature in Akodon azarae.Arrow indicates increase of the parameters number.Sse: squared sum of error, Fobs: observed F-statistic, Fcrit: critical F-statistic, b 1 and b2: slopes of the first phase and second phase of the regression model, respectively Comparaciones estadfsticas entre Ios niveles de ajuste de modelos de regresi6n lineal de una, dos y tres fases que relacionan, en Akodon azarae, la tasa metab6lica y la temperatura ambiente.La flecha indica el incremento en el numero de panimetros.Sse: suma del cuadrado de Ios errores, Fobs: estadfstico-F obsevado, Fcrit: estadfstico-F crftico, b I y b2: pendientes de la primera y segunda fase, respectivamente, del modelo de regresi6n

DISCUSSION
The inadequacy of W alone to account for the variation in standard energetic led some biologists to examine the effect of both, environment and food habits on mammalian energy expenditure (McNab 1992).The type of diet is crucial since it places limitations on the net amount of energy that can be assimilated and therefore affects metabolic rate.Akodon azarae is an omnivore rodent eating different parts of plants and including insects in its diet (Bilenca 1993 ).Then, the value of BMR found for A. azarae was according either, to the expectation of the Kleiber relation and, to the expectation of the Hayssen & Lacy (1985) relation for murid rodents.Males and females of A. azarae, had the same• BMR.According to Heusner (1982) similarity in oxygen consumption between males and females would imply similarity in body composition.This prediction is in agreement with the fact that both, males and females of A. az(lrae have the same body composition (del Valle personal communication).
Furthermore, in the same sense, similarities in ADMR for males and females of Akodon azarae in non-reproductive condition could imply similar selection pressures on the amount of energy allocated to behavioral activities.In accordance, Bilenca ( 1993) has reported that males and females of A. azarae do not have differences in habitat use.Activity daily metabolic rate in A. azarae was 235 % higher than BMR, suggesting that the cost of locomotion is as expensive as the reproductive cost of lactation which, as stated by Sikes ( 1995), could be 66 to more than 200% over non reproductive requirements.
Akodon azarae is a good thermoregulator.As micro tine rodents (Hart 1971 ), its thermoregulatory capacity was higher than the expected.Moreover, it has a lower than expected minimal thermal conductance and in consequence, its differential in body temperature did not show differences with changes in ambient temperature.As it was found a high mortality in traps at low field temperatures, torpor was evaluated in periods shorter than 24 h.We found that this species did not evidence torpor states, as happen with other Southamerican species such as Calomys musculinus, Eligmodontia typus, C. venitstus, and Phyllotis darwini rupestris (Bozinovic & Rosenmann 1988a, b, Caviedes-Vidal et al. 1990, Bozinovic & Marquet 1991 ).In spite of the short time that individuals were exposed to cold temperatures, ours results are consistent with both, the high thermal insulation observed and with a BMR not lower than the expected for an euthermic mammal.The high field mortality found in traps, could be due to the impossibility to support in winter, a high cost of termorregulation for long times at low temperatures with limited food resource.Thus, Bozinovic & Rosenman (1988b) predictions, about the possible existence of torpor in Akodon azarae are not supported by our data.
Burrow construction (Hodara et al. 1997) and temporal selectivity in habitat use (Antinuchi 1996) in A. azarae has been discussed as a mechanism that allow the avoidance of extreme low ambient temperatures.Although this behaviour would contribute to diminish the quantity of energy assigned to thermorregulation, it may also be important in minimizing the risk of predation.On the other hand observed decreasing field activity of A. azarae during warm hours during summer (Bilenca 1993) could be attributed mainly to thermal characteristics of this species.The avoidance of high T. in the field was in accordance with the fact that, MR responses to ambient temperatures above TNZ was abrupt and reached its highest value at 40 °C. A. azarae is the most numerically important species from pampas grassland and the above mentioned adaptations would contribute to the succes of this species.
Further studies on energetic of the reproduction, thermoregulatory acquisition, and metabolism in the development of newborn of A. azarae will allow a better understanding of the relationship among energetics, life history traits, and behavior in this species.
oc was estimated on only one individual since other mice exhibited thermal shock at this T .a

Fig. 2 .
Fig. 2. Relationship between metabolic rate and time at 15 °C in Akodon azarae.The line shows the limit of euthermia.