Food Habits of Procyonids.—Food habits of six procyonids for which metabolic data are available are presented inTable 9. All six species clearly have mixed diets. Compared to other species,Procyon lotoris highly catholic in its diet, taking food from almost twice as many categories asNasua narica, three times as many asProcyon cancrivorus,Nasua nasua, andBassariscus astutus, and nine times as many asPotos flavus.
For those species for which food habit data are quantified, we used Eisenberg's (1981:247-251) substrate/feeding matrix method, where "substrate" is analogous to McNab's (1986a) "behavior," to construct the following feeding categories that are based on the major food groups utilized by each species (Table 9).
1.Potos flavus:(1) arboreal/frugivore, insectivore.2.Procyon cancrivorus:(1) semiaquatic/crustacivore, molluscivore, insectivore, piscivore, carnivore.3.Nasua nasua:(1) terrestrial/insectivore, arachnidivore, carnivore, frugivore.4.Bassariscus astutus:(1) terrestrial/carnivore, insectivore, frugivore.5.Procyon lotor:(1) terrestrial/carnivore, granivore, frugivore, insectivore; and (2) semiaquatic/crustacivore, molluscivore, insectivore, piscivore, carnivore.
1.Potos flavus:(1) arboreal/frugivore, insectivore.2.Procyon cancrivorus:(1) semiaquatic/crustacivore, molluscivore, insectivore, piscivore, carnivore.3.Nasua nasua:(1) terrestrial/insectivore, arachnidivore, carnivore, frugivore.4.Bassariscus astutus:(1) terrestrial/carnivore, insectivore, frugivore.5.Procyon lotor:(1) terrestrial/carnivore, granivore, frugivore, insectivore; and (2) semiaquatic/crustacivore, molluscivore, insectivore, piscivore, carnivore.
Food Habits and Basal Metabolism.—The most important foods in the diet ofProcyon lotorare vertebrates, nuts, seeds, and fruits (Table 9). These are the same foods that are eaten by those dietary specialists that have Ḣb's equivalent to, or higher than, values predicted for them by the Kleiber equation (McNab, 1986a). The most important foods in the diets ofPotos flavus,Procyon cancrivorus, andNasua nasuaare invertebrates and fruit (Table 9), and these foods are eaten by dietary specialists that have lower than predicted Ḣb's (McNab, 1986a). Major foods in the diet ofBassariscus astutusare terrestrial vertebrates, insects, and fruit (Table 9). Dietary specialists that eat terrestrial vertebrates have higher than predicted Ḣb's, whereas those that feed on insects have Ḣb's that are lower than predicted (McNab, 1986a). Year-round utilization of vertebrates byBassariscus astutussuggests that it also should have a metabolic rate that is equivalent to or higher than predicted, rather than lower (McNab, 1986a). However, perhaps year-round inclusion of insects in its diet (Martin et al., 1951; Taylor, 1954; Wood, 1954; Toweill and Teer, 1977; Trapp, 1978), plus water-and energy-conserving advantages of a low metabolic rate, each exert a stronger selective influence on Ḣbthan do vertebrates in its diet.
Summary.—The basal metabolic rate of these procyonids does appear to be influenced by diet. But, it is apparent from this family's evolutionary history and tropical origins that climate also has had a profound influence on its member's metabolism. The history of the family and the data presented here (Table 7) suggest that lower than predicted Ḣbis a feature that evolved very early as the primary metabolic adjustment to a tropical climate. From this perspective, it could be argued that climate would have been the major selective force determining Ḣb, whereas food habits would have had a secondary influence.
Basal Metabolism and Intrinsic Rate of Natural Increase
Background.—McNab (1980a) suggested that if food is not restricted during an animal's reproductive period, the factor that will limit growth and reproduction will be the rate at which energy can be used in growth and development. Under these conditions, an increase in Ḣbwould actually increase rmaxbecause it would provide a higher rate of biosynthesis, a faster growth rate, and a shorter generation time. Hennemann (1983) tested McNab's (1980a) premise and found a significant correlation between rmaxand metabolic rate, independent of body size, for 44 mammal species. A low correlation coefficient for this relationship, however, indicated to him (Hennemann, 1983) that factors such as (1) food supply, (2) thermal characteristics of the environment, and (3) brain size also contribute toward shaping a species' reproductive potential, particularly when these factors strongly influence rates of biosynthesis or growth or for some reason alter generation time. Results of our estimates of rmaxfor procyonids are presented inTable 10.
Procyon lotor.—This species had the highest Ḣband Dd, and also had the highest rmax(1.34;Table 10). Such a high rmaxmay infer that this trait evolved under conditions where food and temperature were not limiting to reproduction. Under these conditions selection could have favored those reproductive characteristics sensitive to a higher Ḣb(biosynthesis, growth, and generation time; McNab, 1980a).Procyon lotor's high reproductive potential is due to its early age of first female reproduction and its large litter size, characteristics that may reflect metabolically driven increases in both biosynthesis and growth.
Bassariscus astutus.—This species has a low Ḣbbut an rmaxthat was 124% of expected (Table 10). This suggests that rmaxevolved under conditions where food and temperature were not limiting to reproduction. Reduced litter size should restrict this species' reproductive potential and may be a reflection of its low Ḣb. The factor that is responsible for increasing its reproductive potential, however, is its early age of first female reproduction.Bassariscus astutusis the smallest of these procyonids, and even though it has a low Ḣb, its small mass may contribute to its ability to reach adult size and sexual maturity in its first year. The high quality of its diet (a high proportion of small vertebrates;Table 9) also may be a factor that is permissive to early female reproduction. Thus, small body size and diet may be factors that have allowed this species to evolve a higher than expected reproductive potential in spite of its low Ḣb.
Nasua narica.—This species is one of the largest procyonids (Table 7), and it possesses characteristics that should limit its reproductive potential: lower than predicted Ḣb(Table 7), a relatively low-quality diet (Kaufmann, 1962:182-198;Table 9), and delayed time of first reproduction (Table 10). In spite of this,Nasua naricahas a higher than expected rmax(111% of predicted;Table 10). The life history feature that enhancesNasua narica's reproductive potential, and increases rmaxbeyond expected, is its large litter size. In this species females live in bands. Each year just before their young are born these bands break up, and each female seeks out a den for herself andher litter. Once the young are able to leave the den (approximately five weeks), bands reform. In this situation, females not only care for their own young but also for those of other females in the band (Kaufmann, 1962:157-159, 1982, 1987; Russell, 1983). This social structure may contribute to this species' ability to produce large litters and in this way increase its reproductive potential.
Table 10.—Intrinsic rate of natural increase (rmax) of several procyonids. (a = potential age of females producing first young; b = potential annual birth rate of female young (= average litter size/2; average litter size was calculated from the published range of litter sizes for each species); n = potential age of females producing their final young; rmaxe= intrinsic rate of natural increase expected from body mass (Hennemann, 1983); rmaxr= ratio of calculated to expected intrinsic rate of natural increase (rmax/rmaxe).)
[a]rmaxe= 4.9·m0.2622, where m is body mass in grams.[b]Regression of rmaxon body mass (m). Assume rmax= 1.02 forProcyon cancrivorus: rmax= 0.00005·m + 0.623; R = 0.19; R2= 0.03; Regression of rmaxr(Table 10) on Hbr(Table 7); assumeNasua nasuahas the same rmaxrasNasua narica: rmaxr= 3.35·Hbr- 1.11; R = 0.93; R2= 0.86.[c]Estimate based on females reproducing in their first (a = 0.83) or second (a = 1.75) year.
Nasua nasua.—Unfortunately, there is not enough reproductive data to allow calculation of rmaxforNasua nasua(Table 10), therefore, it is not possible to compare the reproductive potential of this South American coati with its North American relative,Nasua narica. Given its low Ḣband relatively low-quality diet of fruit and terrestrial invertebrates (Table 9), however, rmaxofNasua nasuamay be very similar to that ofNasua narica.
Procyon cancrivorus.—The age of first female reproduction forProcyon cancrivorushas not been reported. However, if one assumes females can reproduce in their first year, rmaxforProcyon cancrivoruswould be 1.02 (132% of expected;Table 10). If, on the other hand, first female reproduction is delayed until the second year, rmaxwould be 0.65 (84% of predicted;Table 10).Procyon cancrivorushas a low Ḣb, reduced litter size, and small body mass. Its low Ḣbmay limit litter size, but as withBassariscus astutus, the quality of its diet (a high percentage of small vertebrates;Table 9) and its small body size may make it possible for females to reproduce in their first year and thus increase the species' reproductive potential. This reasoning would argue thatProcyon cancrivorusprobably enjoys higher, rather than lower, than expected rmax.
Potos flavus.—In addition to a low Ḣb, this species possesses other characteristics that limit its reproductive potential: low-quality diet, delayed reproduction, and birth of a single young each year. Because there does not appear to be any other feature of its life history that can counteract the influence of these factors, rmaxinPotos flavushas evolved to be only 48% of expected (0.30;Table 10). Its close relative, the olingo,Bassaricyon gabbii, appears to share the same condition (Table 10).
Summary.—This brief survey illustrates that, with the exception ofPotos flavus, procyonids tend to have values of rmaxthat are higher than those predicted for them on the basis of mass (Table 10). Regression analysis indicates that, within the family, body mass accounts for only a small amount (3%) of the variation in rmax, whereas the positive slope of the correlation between rmaxrand Hbr(R = 0.93) suggests that low metabolism has a limiting effect on rmax(seeTable 10, footnote b). The implication here is that low Ḣbwould be associated with a lower rate of biosynthesis, a slower growth rate, and a longer generation time. Procyonids with low Ḣbbut higher than expected rmaxmust possess other traits that serve to offset the effects of low metabolism. Our survey indicates that the following features compensate for low Ḣband help increase rmax: (1) a high-quality diet may make biosynthesis and growth more efficient, thus optimizing the time element associatedwith each of these processes; (2) larger litter sizes and cooperation in care of the young may increase survivorship in spite of a slower growth rate; and (3) an early age of first reproduction, a long reproductive life span, and moderate-size litters (two to four young) may in the long run add as many individuals to the population as a shortened generation time. Our survey also suggests that, at the other extreme, factors such as a low-quality diet, reduced litter size, absence of cooperative care of the young, delayed age of first reproduction, and shortened reproductive life span all serve to decrease rmax. Thus, it is obvious that diet, litter size, social structure, reproductive strategy, and reproductive life span can operate synergistically with Ḣbto magnify its influence on rmax(as withProcyon lotorandPotos flavus), or they can function in opposition to Ḣbto change the direction of its influence on rmax(as withBassariscus astutus,Procyon cancrivorus,Nasua narica, and perhapsNasua nasua).
Basal Metabolism and Climatic Distribution
Procyon lotor.—The evolution of a higher Ḣb(Tables 7,8) may have been the physiological cornerstone that enabledProcyon lotorto break out of the mold being exploited by other procyonids and to generalize its use of habitats and climates. Once this basic physiological change was in place, selection for appropriate alterations in thermal conductance, capacity for evaporative cooling, diversity of diet, and energy storage would have provided this species with the suite of adaptations needed to extend its distribution into other habitats and climates. Support for this concept follows from the fact that high levels of Ḣbare associated with (1) cold-hardiness in mammals that live in cold-temperate and arctic climates (Scholander et al., 1950c; Irving et al., 1955; Irving, 1972:115, 116; Shield, 1972; Vogel, 1980; Golightly and Ohmart, 1983); (2) the ability to utilize a wide variety of food resources and to occupy a large number of different environments and habitats (McNab, 1980a); and (3) a high intrinsic rate of natural increase (McNab, 1980a; Hennemann, 1983; Lillegraven et al., 1987; Nicoll and Thompson, 1987; Thompson, 1987).
Other Procyonids.—Other procyonids (Potos flavus,Procyon cancrivorus,Nasua narica, andNasua nasua) have lower than predicted Ḣb's (Table 7), a characteristic that is considered to be an energy-saving adaptation for those that live in relatively stable tropical and subtropical habitats (Müller and Kulzer, 1977; Chevillard-Hugot et al., 1980; Müller and Rost, 1983). However,Bassariscus astutusis found in tropical, subtropical, and temperate climates. This species is found from tropical Mexico to temperate regions of the western United States (Kaufmann, 1982, 1987; Nowak and Paradiso, 1983:979). In the northern part of its distribution,Bassariscus astutuslives in habitats that are unstable (arid regions), that are low in productivity, and that characteristically have marked seasonal changes in temperature. Its lower than predicted Ḣbcould be an important water-conserving adaptation at times when temperatures are high (McNab and Morrison, 1963; McNab, 1966; MacMillen and Lee, 1970; Noll-Banholzer, 1979) and an important energy-conserving mechanism when cold weather may limit food availability and hunting time (Scholander et al., 1950c; Wang et al., 1973). As will be seen later,Bassariscus astutusis unique among procyonids with lower than predicted Ḣb's in that it also has a lower than predicted Cmw(Table 7). This allows it to use less energy than expected for thermoregulation at low temperatures. Another species with a similar set of adaptations (lower than predicted Ḣband Cmw) is the arctic hare,Lepus arcticus(Wang et al., 1973), which lives in one of the coldest and least-productive regions on earth. Wang et al. (1973) suggest that this combination of adaptations allowsLepus arcticusto better match its energy requirements to the low productivity of its environment. A similar relationship may hold forBassariscus astutus, particularly in colder arid portions of its distribution, and may be the reason that it, but not other procyonids with low Ḣb's, has been able to inhabit temperate climates.
Minimum Thermal Conductance
Background
Thermal conductance is a measure of the ease with which heat is passively transferred to or from a body through its tissues and pelt. Within Tn, a mammal is able to vary its thermal conductance over a wide range of values by changing heat transfer characteristics of both of these layers. Minimum thermal conductance occurs when total heat transfer through these layers is reduced to its lowest possible rate. This minimum value, which is the reciprocal of maximum resistance, occurs, theoretically, but not always practically (see McNab, 1988b), at the animal's Tlcand is best estimated under standard conditions in a metabolism chamber (McNab, 1980b; Aschoff, 1981). Minimum thermal conductance scales to body mass (McNab and Morrison, 1963; Herreid and Kessel, 1967; McNab, 1970, 1979b; Bradley and Deavers, 1980; Aschoff, 1981). Therefore, to make comparisons between species of various sizes, we scaled out body mass by expressing Cmwas the ratio of measured to predicted values (Cmwr;Table 7). These ratios were used to make comparisons of heat-transfer characteristics between species that occupy different habitats or climates.
Effect of Molt on Thermal Conductance
In summer, Tlc's of male and femaleProcyon lotor(Figure 2) were very similar to those of other procyonids (22°C-26°C;Table 7). In winter, Tlcof both sexes shifted downward to 11°C (Figure 3). This seasonal shift in Tlcoccurred as the result of a seasonal change in minimum thermal conductance (Table 3). For many northern mammals, a seasonal change in thermal conductance is partly mediated via cyclic changes in the insulative quality of their pelt (Scholander et al., 1950a; Irving et al., 1955; Hart, 1956, 1957; Irving, 1972:165).
Procyon lotorbegins to shed its heavy winter coat about the time its young are born. Molt progresses through summer and by late August the new coat is complete (Stuewer, 1942). During its summer molt,Procyon lotor's Cmwincreased by about 49% over the value for female raccoons in winter (Table 3). In summer, therefore, it had the highest mass specific Cmwof those procyonids considered (Cmwr= 1.77 and 1.79;Table 7). An increase in thermal conductance facilitates passive heat loss for temperate and arctic species, and this serves as an important thermoregulatory adaptation during warm summer months (Scholander et al., 1950c; Irving et al., 1955; Hart, 1956, 1957; Irving, 1972:165). This adaptation is particularly important to those temperate- and arctic-zone species (including raccoons) whose Ḣb's do not decrease during summer (Irving et al., 1955). From August on, the fur ofProcyon lotorbecomes increasingly longer and heavier, with peak, or prime, condition occurring in late fall and early winter (Stuewer, 1942). Minimum conductance of our captive raccoons was lowest in winter (Cmwr= 1.15) when their pelts were in prime condition (Tables 3,7). Because "primeness" of raccoon pelts varies geographically, thicker pelts being associated with colder climates (Goldman, 1950:21; Whitney and Underwood, 1952:24-41), the degree of seasonal change in Cmwmust also vary geographically.
The only other procyonid for which a seasonal molt has been described isBassariscus astutus. Molt in this species extends from late summer to late fall (Toweill and Toweill, 1978). How molt effects thermal conductance inBassariscus astutusis not known because metabolic data for this species (Table 7) apparently were collected only when their pelts were in prime condition (Chevalier, 1985).
Goldman (1950:20) reports thatProcyon cancrivorusdoes not have a seasonal molt. Like other tropical procyonids,Procyon cancrivoruslives in an environment that has the following characteristics: high even temperatures throughout the year (1°C-13°C difference in monthly mean temperature), a greater range in temperature between day and night than in mean monthly temperature throughout the year, uniform lengths of day and night, seasonal variation in rainfall, and lowest temperatures during the rainy season(s) (Kendeigh, 1961:340). In such a stable environment there would be no advantage to a sharply defined seasonal molt cycle that could place an animal in thermoregulatory jeopardy by increasing its thermal conductance. This would be particularly true for animals like tropical procyonids that have lower than predicted Ḣb's but that maintain typical eutherian body temperatures (Table 7). Consequently, molt in all tropical procyonids may either be prolonged or continuous. This is a feature of their biology that needs to be examined in more detail.
Comparison of Thermal Conductances
Procyon lotorversus Tropical Procyonids.—CmwrforProcyon lotorin winter was 1.15, which is similar to the values forPotos flavusandProcyon cancrivorus, 1.02 and 1.25, respectively (Table 7). These two tropical species, therefore, have Cmw's that are similar on a mass specific basis to the value forProcyon lotorin winter. However, at their Tlc's, the thermal gradient sustained by these tropical animals is only about 11°C, whereas forProcyon lotorin winter it was 26.5°C. Examination ofEq. 4with respect to these thermal gradients suggests that tropical procyonids achieve such low Cmw's by virtue of their lower than predicted Ḣb's rather than by having pelts that are exceptionally good insulators. In fact, the insulation afforded by the pelts of these tropical procyonids is about the same as that of the 50 g arctic lemming,Dicrostonyx groenlandicus rubricatus, whose coat has an insulative value that is about half that of the hare,Lepus americanus, red fox,Vulpes fulva alascensis, and pine martin,Martes americana, animals comparable in size to these procyonids (Scholander et al., 1950a). Therefore, pelts of these tropical procyonids do not have the same insulative value as the prime winter coat ofProcyon lotor.
Nasua naricaandNasua nasuahave tropical and subtropical distributions and they are the only procyonids that are diurnal (Kaufmann, 1962:103-105, 1982, 1987). Because they are active during the day they experience a more extreme thermal environment (higher Ta's and solar radiation) than their nocturnal cousins. Values of CmwrforNasua narica(1.45 and 1.55) andNasua nasua(1.24 and 1.65) are higher than those forProcyon cancrivorusorPotos flavus(Table 7). Thus, these coatis have higher mass specific Cmw's than their nocturnal tropical cousins. A high Cmwreduces the cost of thermoregulation in hot environments because it increases an animal's ability to lose excess heat passively. The higher Cmw's of these coatis serve as an adaptation that contributes to the success of their diurnal life style as well as their ability to expand their habitat use to areas with less thermal stability, such as oak and pine woodlands and deserts.
Bassariscus astutus.—This species has the lowest mass specific Cmwof these procyonids (Cmwr= 0.85;Table 7), which indicates that its pelt has a greater insulative value than the coats ofPotos flavus,Procyon cancrivorus,Nasua nasua, orNasua narica. This, coupled with a lower than predicted Ḣb, allowsBassariscus astutusto maintain Tbwith less energy expenditure than is possible for any other procyonid of comparable size; and this combination of adaptations providesBassariscus astutuswith a distinct energy advantage in environments that have low productivity (Wang et al., 1973). The evolution of a pelt that provides better insulation must be considered an, important contributing factor for the spread of this species into desert regions of the western United States.
Thermoregulation and Use of Stored Fat at Low Temperatures
Background
Thermoregulation.—At temperatures below a mammal's Tn, heat loss exceeds Ḣb. To maintain Tbunder theseconditions, metabolic rate must be increased (Eq. 4).Procyon lotorin summer during its annual molt (Table 5;Figure 2),Bassariscus astutus(Chevalier, 1985),Nasua nasua(Chevillard-Hugot et al., 1980; Mugaas et al., in prep.),Nasua narica(Scholander et al., 1950b; Mugaas et al., in prep.), andPotos flavus(Müller and Kulzer, 1977; Müller and Rost, 1983) all are able to elevate their metabolic rates by 130% above basal when they are exposed to Ta= 0°C.Procyon cancrivorusresponds to 0°C with an increase in metabolic rate of 257% above basal (Scholander et al., 1950b). All animals listed have about the same Tlcand Tb, so the temperature differential producing this response is about the same for each species. Metabolic ability to defend body temperature against low ambient temperatures, therefore, is well developed in these procyonids. Such large increases in metabolic rate are energetically expensive, and if these animals were routinely exposed to Ta= 0°C, it would be difficult for them to acquire enough food each day to maintain endothermy. Raccoons in winter pelage, however, need only elevate their metabolic rate by 47% above basal to maintain endothermy at Ta= 0°C (Table 5;Figure 3). Each year at the completion of its molt, the raccoon's highly insulative pelt is renewed. This lowers their Tlcby 9°C to 15°C below that measured for them in summer (Figure 3) and decreases their cost of thermoregulation at low temperatures. The increased insulative capacity of their pelt is one of the primary adaptations that has allowedProcyon lotorto extend its distribution into cold climates.
Stored Fat.—Cyclic fattening is an integral and important part of a raccoon's annual cycle (Mugaas and Seidensticker, ms); however, it has not been reported for other procyonids. During winter in parts of the United States and Canada, raccoons are confined to their dens for variable periods of time (days to months) depending on the severity of the weather (Stuewer, 1943:223-225; Whitney and Underwood, 1952:108-116; Sharp and Sharp, 1956; Mech et al., 1968; Schneider et al., 1971). During this confinement, they do not hibernate but rather enter a state of "dormancy" and become inactive. While dormant they remain endothermic (Tb> 35°C; Thorkelson, 1972:87-90) and derive most of their energy requirement from fat reserves accumulated during fall. The rate at which fat stores are consumed during winter dormancy depends on the thermoregulatory requirement imposed on them by local weather conditions, the insulative quality of their pelt, and any advantage they may gain by seeking shelter in a den.
Thermal Model of the Raccoon and Its Den
Heat transfer between an animal and its environment is a function of the interaction of its body temperature and thermal conductance with various environmental variables (air temperature, wind speed, vapor pressure, and thermal radiation). When a raccoon is outside its den, its thermal conductance (Cmw) is the only barrier to heat transfer with the external environment. However, when it enters a tree den, a raccoon imposes two other thermal barriers between itself and the external environment: (1) conductance of the air space between its fur and the den's walls (Ca) and (2) conductance of the den's walls (Cd; Thorkelson, 1972:59-63; Thorkelson and Maxwell, 1974). Thorkelson and Maxwell (1974) modeled heat transfer of a simulated raccoon (a water-filled aluminum cylinder equipped with a heater and covered with a raccoon pelt) in a closed tree den. In their system, 65% of resistance to heat flux was attributable to the pelt, whereas the remainder (35%) was due to Caand Cd. Because resistance is the inverse of conductance, and resistances for the raccoon and its den are arranged in series, we can estimate total conductance (Ct) of this system withEq. 7.
Minimum thermal conductance Cmwfor raccoons in winter was 0.0172 mL O2·g-1·h-1·°C-1(Table 3). Based on Thorkelson and Maxwell's (1974) model we let 1/Cmw= 0.65(1/Ct) = 1/0.0172 mL O2·g-1·h-1·°C-1, and 1/Ca+ 1/Cd= 0.35(1/Ct). Substituting these values intoEq. 7and solving for Ctyields 0.0112 mL O2·g-1·h-1·°C-1, a value that is 35% lower than that of the animal alone. Substituting this value and the value for basal metabolism of winter raccoons (0.47 mL O2·g-1·h-1;Table 7) intoEq. 4and solving for (Tb- Ta) yields a new temperature differential of 42°C. Therefore, by using tree dens, raccoons in north central Virginia, with Tb= 37°C (Figure 7), could effectively reduce their Tlcfrom 11°C to -5°C and markedly reduce their metabolic cost of thermoregulation.