Correlation of Metabolism with Tissue Carbon and Nitrogen Turnover Rate in Small Mammals

Oecologia (2006) 150:190–201 DOI 10.1007/s00442-006-0522-0

ECOPHYSIOLOGY

Correlation of metabolism with tissue carbon and nitrogen turnover rate in small mammals

Stephen E. MacAvoy Æ Lynne S. Arneson Æ Ethan Bassett

Received: 20 February 2006 / Accepted: 18 July 2006 / Published online: 12 September 2006 Ó Springer-Verlag 2006

Abstract Stable isotopes have proven to be a useful but the literature for birds shows that mass and tool for deciphering food webs, examining migration whole-body metabolic rates in birds scale logarithmi- patterns and determining nutrient resource allocation. cally with tissue turnover. Interestingly, the mamma- In order to increase the descriptive power of isotopes, lian data graph separately from the bird data on a an increasing number of studies are using them to turnover versus metabolic rate plot. Both mice and model tissue turnover. However, these studies have, rat tissue in this study exhibited a slower turnover mostly by necessity, been largely limited to laboratory rate compared to metabolic rate than for birds. These experiments and the demand for an easier method of data suggest that metabolic rate may be used to estimating tissue turnover in the ﬁeld for a large estimate tissue turnover rate when working with variety of organisms remains. In this study, we have organisms in the ﬁeld, but that a different relationship determined the turnover rate of blood in mice and between tissue turnover and metabolism may exist for rats using stable isotope analysis, and compared these different classes of organisms. rates to the metabolic rates of the animals. Rats (Rattus norvegicus) (n=4) and mice (Mus musculus) Keywords Stable isotope Æ Turnover Æ Metabolism Æ (n=4) were switched between isotopically distinct Blood Æ Mammal diets, and the rate of change of d13C and d15N in whole blood was determined. Basal metabolic rates

(as CO2 output and O2 consumption per unit time, Introduction normalized for mass) were determined for the rats and mice. Rats, which were an order of magnitude Stable isotope analysis has become an important tool larger and had a slower metabolic rate per unit mass used by ecologists when determining food webs, re- than mice (0.02 vs. 0.14 O2/min/g), had a slower blood source use, migration patterns, and species interac- 13 turnover than mice for C (t1/2 =24.8 and 17.3 days, tions. The basic premise behind using stable isotopes in 15 respectively) and N (t1/2 =27.7 and 15.4 days, res- dietary studies is that a consumer’s tissues will isoto- pectively). A positive correlation between metabolic pically resemble what is consumed (Fry and Sherr rate and blood isotopic turnover rate was found. 1984; Minagawa and Wada 1984; Peterson and These are the only such data for mammals available, Howarth 1987). Tissues can only be built from available nutrients, such as carbohydrates, proteins, and lipids, although the extent to which a tissue resembles Communicated by Jim Ehleringer. the different dietary components depends on the percent composition of the food as well as the type of S. E. MacAvoy Æ L. S. Arneson (&) Æ E. Bassett tissue examined (Hobson et al. 1995; MacAvoy et al. Department of Biology, American University, 2005; Tieszen et al. 1983; Tieszen and Farge 1993). Hurst Hall 101, 4400 Massachusetts Ave NW, Washington, DC 202-885-2186, USA Stable isotope analysis can be a valuable tool in e-mail: [email protected] differentiating the relative importance of protein and

123 Oecologia (2006) 150:190–201 191 carbohydrate in diet as long as the isotope ratios of the isotopes to characterize food webs could evaluate various nutrient components are known (Arneson and when the organisms could be expected to be in isotope MacAvoy 2005; Hobson and Bairlein 2003; Koch equilibrium. and Phillips 2002; Phillips and Gregg 2003; Phillips and Metabolism refers to the sum of all catabolic Koch 2002). By utilizing stable isotopes, studies have (degrading) and anabolic (synthesizing) processes that shown that protein is the primary source material used occur in a living system. Metabolic processes produce for building new tissues whereas carbohydrates are energy by consuming O2 in order to break down mol- used mainly as an energy source (Hobson et al. 2000; ecules that store energy, producing CO2 as a waste MacAvoy et al. 2005). Because organisms resemble the product. Anabolic metabolic processes use this energy isotopic signatures of their diets, many ecological to produce macromolecules, carbohydrates, proteins or studies use stable isotopes to gain insights into the food lipids, in order to either increase tissue mass (resulting web of a particular community or ecosystem (see in growth) or to replace macromolecules that have Hobson 1999; Hobson and Wassenaar 1999 for re- been degraded (resulting in tissue replacement). Thus, views). These studies often assume that consumers are metabolic processes contribute to both components of in isotopic equilibrium with their diet. However, this tissue turnover as measured by stable isotope analysis, may not always be the case in the event of a shifting growth and tissue replacement (MacAvoy et al. 2005). diet or migration of consumer or prey (Gannes et al. Metabolic rate, which can be measured by oxygen 1997; Hobson 1999). consumption per unit time, correlates to the body mass The isotopic turnover rates of various tissues have of the organism according to the ‘‘3/4 rule’’ (Kleiber been published for several animals, including birds, 1932, 1947), in which metabolic rate is proportional to mammals, and marine organisms. For example, it has M0.75, where M is body mass. been shown that garden warblers (Sylvia borin) and In this study we examine the relationship between Japanese quails (Coturnix japonica) have similar tissue turnover rate and resting metabolic rate for two blood d13C half-lives at 5.4 and 11.4 days, respec- species of small mammal, rats (Rattus norvegicus) and tively (Hobson and Bairlein 2003; Hobson and Clark mice (Mus musculus). We also compile existing 1992), while larger birds such as the American crow information on metabolic rate and tissue turnover for (Corvus brachyrhynchos) and canvasback duck a variety of avian species in order to broadly describe (Aythya valisineria) have longer blood half-lives: the overall relationship between the two variables for approximately 30 and 21 days, respectively (Haramis birds. We found that rats, which were an order of et al. 2001; Hobson and Clark 1993). MacAvoy et al. magnitude larger and had a slower metabolic rate per (2005) found that blood d13C turnover in mice (Mus unit mass than mice, had a slower blood carbon and musculus) has a half-life of approximately 17 days, nitrogen turnover than mice (carbon t1/2 = 24.8 vs. 13 while Voigt et al. (2003) found blood d C half-lives 17.3 days; nitrogen t1/2 = 27.8 and 15.4 days). Meta- in bats were much longer, 113 or 120 days in two bolic rates and blood turnover rates of different bird different species. Turnover in poikilotherms is much species show that mass and whole-body metabolic slower than that of homeotherms, which is likely due rates scale logarithmically with tissue turnover. to low metabolic rates (MacAvoy et al. 2001). In However, mice and rats both had a slower blood cases where turnover in poikilotherms is relatively isotopic turnover rate versus MR/gram than birds of fast, it has been shown that the rapid turnover is due similar body mass, suggesting that the relationship mostly (and in some cases exclusively) to growth between these two variables may differ between classes (Tieszen and Farge 1993; Frazer et al. 1997; Herzka of organism. and Holt 2000; Maruyama et al. 2001; Vander Zanden et al. 1997). Studies continue to show that large and signiﬁcant Methods differences exist among organisms in how quickly they incorporate the isotopic signature of their diet. Al- Experimental design and tissue sampling though an increasing number of studies are appearing in the literature, measurements of isotope turnover are Adult female Sprague-Dawley rats (R. norvegicus) still limited to a few species. However, if a relationship (n=6) and adult female BALB/c mice (M. musculus) between isotope turnover and a more common, well- (n=6) were allowed to equilibrate to a control diet documented parameter, such as metabolic rate, could for approximately 120 days (2018, Harlan Teklad, be established, then estimates of isotope turnover rates Madison, WI, USA). Four of each species began an could be more widely applied. Researchers using isotopically distinct experimental diet, while two

123 192 Oecologia (2006) 150:190–201 remained on the control diet. The control diet con- Metabolic rates (MR) tains both corn and wheat, and contains 18.9% protein. The experimental diet is 21% protein from CO2 output was measured using a Qubit Systems casein, 59% carbohydrate from beet sugar and 7% (Kingston, ON, USA) S153 CO2 Analyzer attached to soybean oil, with the remainder of the diet composed an open air respiration chamber. Two readings were of ﬁber, vitamins and minerals (Arneson and Mac- taken per animal during the equilibration phase, and Avoy 2005). Carbon and nitrogen isotope values for two more were taken prior to blood sampling each diet are given in Table 1. Four blood samples throughout the course of the experiment. In this study, (lateral tail vein bleeding, ~100 ll) were taken from MR is measured as O2/min. each animal during the equilibration period, and every seven days (R. norvegicus) and every ten days Growth rates (M. musculus) during the experimental period, which lasted for 80 days. Blood samples were placed in a Masses of individual animals were taken prior to each drying oven at 60 C for three days and then reﬂuxed ° blood collection and metabolic measurement. The for 35 min in dichloromethane for lipid extraction growth rate constant, k, for each group was determined (Knoff et al. 2002). After air drying, 1 mg of tissue ~ using was measured and placed in tin capsules in prepa- ration for stable isotope analysis. Blood tissue sam- k ln MS=M0 =t; 2 ples were analyzed for d13C and d15N using a Europa ¼ ð Þ ð Þ

Hydra 20/20 (University of California, Davis, CA, where M0 is the initial mass in grams and MS is the USA) stable isotope ratio mass spectrometer (IRMS) mass in grams on day t of the experiment. This was 13 15 to obtain d C and d N. Standards were run in done so that the contribution of growth to tissue duplicate every twelve measurements (within a run turnover could be ascertained (see below, Modeling of 100 samples, which included 15 standards, the turnover). standard deviation of the standard material was 0.2& for nitrogen and 0.1& for carbon). Modeling turnover

Stable isotope analysis The rate of isotope turnover in blood can best be described by the following equation When determining isotopic signatures, the heavy to light isotopic ratio of a particular element in the dC k m C C ; 3 sample is measured relative to that of a stan- dt ¼ Àð þ Þ Â ð À EÞ ð Þ dard. The equation for an isotopic signature is as which describes the change in isotope value over time. follows: C is the signature at day t, CE is the signature in

H L equilibrium with the new diet, k is the speciﬁc growth X X sample dHX 1 1000: 1 rate, m is the metabolic tissue replacement rate, and k ¼ HX=LX À Â ð Þ can be measured directly as a function of time (see Eq. "ðÀ ÞÁstandard # 2 above). Integrating equation 3 results in Here X is any element, and H and L are the heavy and light mass numbers, respectively; units are in per mil k m t C CE aeÀð þ Þ ; 4 (&). À ¼ ð Þ

Table 1 Average tissue signatures at t0 and tend and diet-to-tissue discrimination averages for each diet. ±SD is the standard deviation and Frac stands for ‘‘fractionation,’’ the name given to the diet-to-tissue discrimination value

d13C ±SD Frac d15N ±SD Frac

Control diet (N=10) –21.4& 0.7 2.8& 0.6 M. musculus (N=15) –19.2 0.1 +2.3& 6.1 0.1 +3.3& R. norvegicus (N=6) –19.0 0.1 +2.4 6.1 0.1 +3.3 Experimental diet (N=8) –26.8 0.1 5.5 0.4 M. musculus (N=8) –25.6 0.1 +1.2 8.5 0.1 +3.0 R. norvegicus (N=12) –25.2 0.1 +1.6 8.4 0.1 +2.9

123 Oecologia (2006) 150:190–201 193 where a is the difference between the initial and ﬁnal diets similar to those in this study observed the time isotope signatures. Therefore, Eq. 4 becomes to isotopic equilibrium to be approximately 70 days. Therefore, the experimental mice were fed for

k m t 85 days, and we expected equilibrium to be reached C t C C C eÀð þ Þ ; 5 ð Þ ¼ E þ ð 0 À EÞ ð Þ during this time, although it was some months before where C0 is the initial signature. This is an equation all isotope data could be gathered and analyzed. describing isotope change over time as used by Hess- Although we did not have isotope equilibrium time lein et al. (1993); m can be calculated by rearranging estimates for rats, weekly sampling was halted at Eq. 5 so that 72 days. However, the experimental group continued to be fed the new diet, and a final blood sample was taken at 176 days to ensure there was no difference C CE ln À C0 CE between the last weekly measurement and the one m À k : 6 13 ¼ À2 t þ 3 ð Þ 100 days later. No significant difference (d C –25.1 vs. –25.1& and d15N 8.3 vs. 8.4&, respectively) was 4 5 observed between the 72- and 176-day samples. The relative importance of k to m changes during different life stages. For young, fast-growing organisms, the m component is negligible relative to k (Frazer Half-life et al. 1997; Fry and Arnold 1982; Herzka and Holt 2000). The opposite is true for very slow-growing or In this study, half-life refers to the time it takes for half adult organisms. In these cases it has been shown that of the existing tissue to resemble the isotopic signature metabolic tissue replacement is an important driver of of the new diet. Half-life is calculated by the following isotopic change. MacAvoy et al. (2005) found that, in equation: adult mice, the rate of change is almost entirely ln 2 (>90%) due to m, as little or no growth occurs. It is also t : 7 1=2 ¼ m k ð Þ clear that m (metabolic tissue replacement rate) is þ apparently higher in animals with high metabolic rates (measured as oxygen consumption or carbon dioxide Metabolic rate and m consumption per gram). Using the growth and metabolic tissue replacement constants, the half-lives of Equations were developed relating the rate of blood various elements within a tissue can be calculated, turnover (m and t ) to average mass, metabolic rate yielding a measure of how quickly an organism 1/2 and metabolic rate per unit mass. To complete these resembles the isotopic signature of its diet. models, a literature review was also performed with When solving Eq. 5 for the contribution of growth to any published turnover data available. In addition to tissue turnover, m was set to zero. The growth using data obtained in this study with R. norvegicus and constant, k, used in Eq. 5 was the value obtained from M. musculus, data found in published papers on other the day equilibrium was reached, t . Any isotopic E species was also used to support/strengthen correla- turnover in excess of what was attributable to growth tions (birds: Bearhop et al. 2002; Chamberlain et al. was considered metabolic tissue replacement. The 1997; Evans-Ogden et al. 2004; Haramis et al. 2001; metabolic constant, m, was determined using observed Hobson and Clark 1992a, 1993; Pearson et al. 2003; data in Eq. 6 from day 0 to day t . The best estimate E mammals: Ayliffe et al. 2004; Tieszen et al. 1983). of m resulted in the least absolute sum of the differences between calculated and observed isotope values for each measurement up until the time of equilibrium (which was day 65 in all cases). We interpreted an Results approximate 0.1& fluctuation between measurements made at successive time intervals as being indicative of Growth rates the animals reaching isotopic equilibrium. The repro- ducibility of tissue isotope (d13C and d15N) measure- The animals used in this study were mature at the ments usually has a standard deviation of ±0.2&, so beginning of the experiment. Both the rats and the we accepted a 0.1& fluctuation as a reasonable mice were a minimum of eight months old on day zero approximation of isotope equilibrium. Earlier feeding of the experiment, and were sexually mature. The mass experiments with this particular strain of mouse and of each animal was determined every seven (rats) or

123 194 Oecologia (2006) 150:190–201 ten (mice) days to determine rate of growth (Fig. 1). (±0.1) in M. musculus (n=15), and –19.0& (±0.1) in R. Although some change in mass was seen in both the norvegicus (n=6; Table 1). The average blood d15N at control and experimental groups of rats and mice, it the beginning of the experiment was 6.1& (±0.1) in M. was minimal. R. norvegicus in the experimental group musculus (n=15), and 6.1& (±0.1) in R. norvegicus (n=4) had an average growth rate of 8.62·10–5 g/day (n=6; Table 1). The diet–tissue discrimination values at (±3.13·10–4), while the control group (n=2) grew at an the beginning of the experiment were 2.3 and 2.4& for average rate of 4.50·10–4 g/day. M. musculus on the carbon and 3.3 and 3.3& for nitrogen in M. musculus experimental diet (n=4) had an average growth rate of and R. norvegicus, respectively. –3.95·10–4 g/day (±6.89·10–4), while the control group Blood samples were collected every ten (mice, grew at an average rate of 3.64·10–4 g/day. Any change Fig. 2) or seven (rats, Fig. 3) days and carbon and in mass was most likely due to change in fat deposition nitrogen stable isotope ratios were analyzed. Blood from excess caloric uptake. This is suggested by the carbon and nitrogen turned over at similar rates relatively sharp decrease in mass after 15–24 h of within the same species. The half-life of tissue carbon fasting prior to metabolic testing (Fig. 1). and nitrogen was 17.3 and 15.4 days, respectively, for mice (Table 2). Blood tissue carbon and nitrogen Turnover rates turned over more slowly in rats, with a half-life of 24.8 and 27.7 days, respectively (Table 2). Diet–tissue dis- The animals used in this study were weaned on the crimination values at the end of the experiment were control diet, and were also fed this diet until they were 1.2 and 1.6& for carbon and 3.0 and 2.9& for nitro- switched to the experimental treatment. The animals gen in mice and rats, respectively (Table 1). For both were maintained on the control diet for an equilibration period of four months to ensure that the different lots of the commercially available diet did not affect the stable isotope ratio. Blood samples taken during a -16 the equilibrium phase prior to the beginning of the experiment from both species of animals indicated that -18 the carbon and nitrogen stable isotope ratios did not -20 change during this time (data not shown). -22 On day zero of the experiment, two mice and two C -24 rats were maintained on the control diet, while the 13 δ diets of four mice and four rats were changed to the -26 experimental diet. Blood samples were collected from -28 all animals in order to determine the starting carbon and nitrogen stable isotope ratios. The average blood -30 13 0 10 20 30 40 50 65 72 85 d C at the start of the experiment (t0) was –19.2& Days

R. norvegicus b 10 400 360 9 320 280 8 240 200 7

0 10 20 30 40 50 60 70 N 15

M. musculus δ 6

Mass (g) 35 5 30

25 4 0 10 20 30 40 50 65 72 85 20 0 10 20 30 40 50 60 70 Days Days Fig. 2a–b a Carbon and b nitrogen isotope change in Mus Fig. 1 Mass of individual Rattus norvegicus and Mus musculus musculus blood over time. Controls are squares and experimen- versus time tals are circles. Bars represent ±1 standard deviation

123 Oecologia (2006) 150:190–201 195

Metabolic rates a -16 -18 Metabolic rate averages were determined using data -20 from the experimental period, including day zero. Average whole-body metabolic rate for mice was -22 3.2 ml O2/min compared to 8.9 ml O2/min for rats.

C -24 However, as expected, metabolic rate per gram for 13 δ -26 mice was considerably higher than that for rats (6.84 ml O /h/g for mice compared to 1.84 ml O /h/g -28 2 2 for rats) (Fig. 4). -30 0 7 14 21 28 35 42 49 65 72 Days Discussion b 9 This study examined the blood turnover rate in mice

8 and rats using stable carbon and nitrogen isotopes. We have found that blood carbon and nitrogen turnover

7 was roughly equivalent in mice and in rats, and that mouse blood turned over slightly faster than rat blood. 6 We have also determined the resting metabolic rate N

15 (MR) of the animals used in this study. The measured δ 5 MR and the relationship between mass and metabolic rate are consistent with what other researchers have 4 observed (Fig. 5). Our ﬁndings are also in accordance 0 7 14 21 28 35 42 49 65 72 Days 1.2 Fig. 3a–b a Carbon and b nitrogen isotope change in Rattus norvegicus blood over time. Controls are squares and experi- 1.0

) Rats mentals are circles. Bars represent ±1 standard deviation ) h /

2 0.8 O

m 0.6 ( mice and rats, equilibrium of blood tissue with the Mice R 0.4

experimental diet occurred by day 65. Values at the M (

end of the experiment were used as the equilibrium g o 0.2 values in Eq. 5 to determine turnover constants. The L data do not allow the extent to which the animals 0.0 were actually at equilibrium to be determined, and we 0.0 0.5 1.0 1.5 2.0 2.5 3.0 cannot rule out the existence of a very long turnover pool (see Ayliffe et al. 2004). However, the values at 14.0 the end of the experiment are likely to be close 12.0 )

(within ±0.1&) to the true equilibrium values, and g /

h 10.0 /

calculated turnover rates are relatively insensitive to 2 O ( error in this parameter. 8.0 s

s Mice a 6.0 M /

R 4.0 Table 2 Average half lives (t ) and metabolic constants (m) for M 1/2 Rats each isotope in each species 2.0 m t (days) 0.0 1/2 0.0 0.5 1.0 1.5 2.0 2.5 3.0 M. musculus 13C 0.04 17.3 Log (Mass(g)) R. norvegicus 0.028 24.8 M. musculus 15N 0.0449 15.4 Fig. 4a–b a Log metabolic rate (O2/h) versus log mass (g) and b R. norvegicus 0.025 27.7 metabolic rate per gram (O2/h/g) versus log mass (g) for Rattus norvegicus and Mus musculus

123 196 Oecologia (2006) 150:190–201

0.06 contains both a growth and a metabolic component (Hesslein et al. 1993; MacAvoy et al. 2005). For adult 0.05 g / animals with relatively high metabolic rates and very ) n i 0.04 little mass gain over time, isotope turnover should be m / 2

O particularly highly correlated with MR. Indeed, Spon-

L 0.03

m heimer et al. (2006) suggest the existence of just such a (

R 0.02 relationship between alpacas and gerbils. However, our M results are in opposition to the ﬁndings of Carleton and 0.01 del Rio (2005), which conclude that metabolism and 0 isotope incorporation into tissues are not directly re- 0 1 2 3 4 5 6 lated in birds. Log (mass(g)) In this study we could not determine the type of

Fig. 5 Metabolic rates (ml O2/min)/g for mammals (ﬁlled dia- correlation between tissue turnover and metabolism monds) and all birds (open squares) in Tables 3 and 4 were plotted because two species were utilized and two data points versus average mass of the animal. Trend lines are shown for both cannot yield a predictive equation. Only blood was categories of animals. For mammals, y = –0.0265ln(x) + 0.0462, r2=0.9246, whereas for all birds, y = –0.0313ln(x) + 0.0529, sampled in this study, so we could not compare our r2=0.8854 isotopic turnover values with those reported for other mammals (horse, Ayliffe et al. 2004; gerbil, Tieszen et al. 1983), as these studies did not measure blood with the long-known relationship between size and metabolism (Kleiber 1932, 1947). The mice exhibited a faster MR per gram tissue but slower whole-body MR 30 relative to rats (Fig. 4), and our data indicate that faster isotope turnover is associated with higher MR/g Rat

(Fig. 7). 20 The carbon isotope turnover rate of the mice was Mouse C (days) 13 slightly slower in this study than that reported for the 1/2 same strain by MacAvoy et al. (2005) (half-lives for t 10 carbon: 17.3 vs. 16.9). However, the mice used in the previous study were younger than the adults used here and the shorter half-life likely reflects an increase in 0 mass superimposed on turnover from metabolic tissue 0 5 10 15 20 25 30 replacement. MacAvoy et al. (2005) estimated that growth accounted for approximately 10% of the ob- 30 served isotope turnover, whereas in the current study mass gain was negligible. Rat In this study, as in our previous studies (MacAvoy 20 et al. 2005; Arneson et al. 2006), the carbon and N Mouse (days) 15 nitrogen tissue turnover rates are roughly similar 1/2 within a tissue. This is in contrast to results from t 10 Carleton and del Rio (2005) in which they find that d13C incorporation is approximately 1.5 times faster than d15N in birds exposed to low temperatures. 0 Although we do frequently see differences between 0 5 10 15 20 25 30 nitrogen and carbon incorporation rates, we cannot MR (mLO2/min) apply significance to these differences. In the final section of this study, we compared the Fig. 6a–b Carbon (a) and nitrogen (b) stable isotope half-lives in blood for the birds given in Table 3 are plotted versus metabolic tissue turnover rate to MR in rats and mice, and found rate (closed diamonds). The trend lines for these relationships that there is a correlation between these two mea- are shown, and the equations are a y = 3.802ln(x) + 6.978, surements. Specifically, we found that a faster MR per r2=0.90 and b y = 3.2919ln(x) + 8.2191, r2=0.72. The metabolic gram tissue correlates with a faster rate of tissue rates for mouse and rat are overlaid for blood stable carbon (a) and nitrogen (b) isotope turnover. Note that values for gerbil and turnover (Fig. 7). This relationship is not unexpected, horse are not included, as blood isotopic turnovers were not as the equation describing stable isotope turnover rate determined for these species

123 Oecologia (2006) 150:190–201 197 isotope turnover. Therefore, we examined the litera- rats have longer blood tissue turnover rates per gram ture for other studies determining whole blood turn- body mass than birds (Figs. 6, 7). While we realize that over rate in species for which the metabolic rate has the relationship between passerines and nonpasserine been determined (Table 4). These data, exclusively birds and metabolic rate are not equivalent (McNabb from studies of birds, describe a logarithmic relation- 1988), when the birds are grouped as a whole, they ship between isotope turnover and metabolic rate tend to have faster metabolic rates per gram tissue than (Fig. 6) or metabolic rate per gram (Fig. 7) when mammals, especially in smaller animals (Fig. 5). Be- plotted. The various studies pooled to examine this cause mammals have a slower metabolic rate per gram relationship examined birds ranging in mass from 11.5 than birds, and metabolic rate per gram is positively to 1,250 g and metabolic rates ranging from 0.56 to correlated to tissue turnover rate, it is not surprising

27.5 O2/min (warbler and canvasback, respectively) that mice and rats have a slower turnover rate than (McKechnie and Wolf 2004). Yet even with the order similarly sized birds. Although the metabolic rate and of magnitude differences in these variables, the corre- blood turnover rate have only been correlated for two lation of metabolic rate per unit mass with 13C and 15N mammals in this study, the relationship seen in various turnover is very high (0.87 and 0.9, respectively, Fig. 7). avian species suggests that the correlation in mammals Unfortunately, almost all of the tissue turnover would also be logarithmic. However, it is important to studies published have only examined avian species. stress that additional data points are needed to verify While these data show that within a class of organism this relationship and construct a model. there is a high degree of predictability regarding how It should also be noted that the relationship between quickly an organism will resemble the isotope signa- isotope turnover and MR found in the study likely only ture of its diet that comes from knowledge of its holds for adults (where growth rate is effectively zero) metabolic rate, it is clear from Figs. 6 and 7 that and homeotherms. We postulate that in growing ani- mammalian MR versus isotope half-life are not well- mals or poikilotherms there would be a markedly dif- described by the relationship in birds. Both mice and ferent relationship and likely a weaker correlation between the two variables than what we show here. Determining the relationship between metabolic 30 rate and tissue turnover rate could allow researchers to 25 Rat predict the turnover rate of the tissue in question based 20 on knowledge of the metabolic rate of the animal being Mouse 15 studied. Currently, tissue turnover rates are deter-

C mined in laboratory studies by changing the diet of the (days) 13 10 organism and measuring the stable isotope ratios of the 1/2 t 5 tissues over time as they come into equilibrium with 0 the new diet. The relationship proposed in this study 0 0.01 0.02 0.03 0.04 0.05 0.06 will be valuable for researchers seeking a way to pre- 30 dict time to isotope equilibrium for species for which Rat metabolic rate is known but the isotope turnover is not. 25 If a relationship between metabolic rate and turnover 20 N Mouse rate can be modeled, the researcher could determine (days) 15 15 1/2

t the metabolic rate for the organism and use this mea- 10 surement to predict the isotope turnover rate.

5 The main caveat to this work is that metabolic rate can change, and therefore likely cause the tissue turn- 0 0 0.01 0.02 0.03 0.04 0.05 0.06 over rate to be altered. Most published metabolic rates are reported as basal metabolic rates (BMR), which MR/g (ml O2/min/g) require the measurement of metabolic rate under Fig. 7a–b Carbon (a) and nitrogen (b) stable isotope half-lives in standard conditions (McNabb 1988). These standard blood for the birds given in Table 3 are plotted versus metabolic rate per gram (closed diamonds). The trend lines for these conditions require the animal be (1) resting during its relationships are shown, and the equations are a y = –7.5508ln(x) normal time of rest (as in circadian rhythm); (2) in 2 2 – 16.536, r =0.87 and b y = –3.2376ln(x) – 1.6114, r =0.90. The thermoneutrality; (3) postabsorptive; and (4) an adult metabolic rates for mouse and rat are overlaid for blood stable (McNabb 1988). In contrast, metabolic rates measured carbon (a) and nitrogen (b) isotope turnover. Note that values for gerbil and horse are not included, as blood isotopic turnover in the ﬁeld may be affected by photoperiod, tempera- was not determined for these species ture (ambient and body), animal activity, age and diet.

123 198 Oecologia (2006) 150:190–201

13 15 Table 3 Mass, metabolic rate (MR in ml O2/min), and isotope ( C and N) turnover rates of various tissues in four mammals 13 15 Species Size (g) MR Tissue Ct1/2 Nt1/2 (1) Turnover reference (ml O2/min) (days) (days) (2) MR andmass reference

Mammalian: Rodents 101 g Gerbil 64.8 1.61 Liver 6.4 ND (1) Tieszen et al. (1983) Meriones unguiculates Fat 15.8 (2) Lovegrove (2003), Muscle 27.7 White and Seymour (2003) Brain 27.7 Hair 46.2 Mouse 18.55 0.84 Blood 16.9 19.3 (1) Macavoy et al. (2005) Mus musculus 18.55 0.84 Liver ND 7.3 (2) Lovegrove (2003), Muscle 16.5 24.8 Heusner (1991) Mouse 27.7 3.16 Blood 18.6 19.6 (1) This Study Mus musculus (2) Lovegrove (2003); Heusner (1991) Mammalian: Rodents 102g Rat 288 5.90 Blood 24.7 21.3 (1) This Study Rattus norvegicus (2) Hart (1971) Mammalian: 105g Horse 488,636 1125.22 Hair 136.0 ND (1) Ayliffe et al. (2004) Equus caballus 409,778 487.45 (2) Bromham et al. (1996), Equus asinus Heusner (1991)

13 15 Table 4 Mass, metabolic rate (in mL O2/min), and isotope ( C and N) turnover data for seven species of birds 13 15 Species Size (g) MR Tissue Ct1/2 Nt1/2 (1) Turnover reference (ml O2/min) (days) (days) (2) MR and mass reference

Avian: Passerine Canvasback 1,250 27.50 Blood (clam/corn Diet) 26.2 26.5 (1) Haramis et al. (2001) Aythya valisineria Blood (tuber diet) 19.8 25.4 (2) Woodin and Stephenson (1998) Blood (clam diet) 16.0 17.6 Average 20.7 23.2 Garden warbler 24.8 1.23 Blood (meal worm diet) 5.0 11.2 (1) Hobson and Bairlein (2003) Sylvia borin Blood (blk elderberry diet) 5.8 5.0 (2) Mckechnie and Wolf (2004) Average 5.4 8.1 Yellow-rumped warbler 11.5 0.56 Blood (49% insect diet) 3.9 7.5 (1) Pearson et al. (2003) Dendrorica coronta Blood (73% insect diet) 5.0 (2) Mckechnie and Wolf (2004) Blood (97% insect diet) 6.1 Average 5.0 American crow 384.8 9.70 Plasma 2.9 (1) Hobson and Clark (1993) Corvus brachyrhynchos Blood cells 30.1 (2) Mckechnie and Wolf (2004) Avian: Nonpasserine Japanese quail 115 2.89 Blood 11.2 (1) Hobson and Clark (1992) Coturni japonica Liver 2.5 (2) Roberts and Baudinette (1986) Muscle 12.4 Bone 173.3 Great skua 970 14.03 Blood 15.1 12.0 (1) Bearhop et al. (2002) Catharacta skua (2) Mckechnie and Wolf (2004) Dunlin 44 1.57 Blood 11.2 10.0 (1) Evans-Ogden et al. (2004) Calidris alpina paciﬁca (2) Lindstrom (1997)

It has been shown that the metabolic rate that can none with marsupials. The lack of evidence supporting be measured in the field (FMR) scales differently with a relationship between FMR and BMR is probably body size than BMR (Kojeta 1991). Kojeta (1991) and due to the fact that MR is highly variable in the field. Ricklefs et al. (1996) showed that a correlation does BMR, therefore, is important as a standardized tool exist between BMR and FMR with mammals (ro- that justifies comparisons between individuals and dents), but a weak correlation exists with birds, and species.

123 Oecologia (2006) 150:190–201 199

The potential difference between FMR and BMR mammals that consume low dietary nitrogen, as well as indicates that using BMR to predict the tissue turnover in animals experiencing starvation. rate could result in inaccurate estimates. However, A second caveat to this work is that studies have most current studies determining tissue turnover rate shown that different tissues turn over at different rates using stable isotope analysis do not conduct their (Arneson and MacAvoy 2005; Evans-Ogden et al. experiments in a field-type setting. Animals are usually 2004; MacAvoy et al. 2005; Tieszen et al. 1983; Voigt housed in a laboratory, with limited living space and et al. 2003). Therefore, the relationship between food supplied ad libitum, likely leading to decreased whole-body metabolic rate and blood carbon or activity levels. Laboratory temperatures and photope- nitrogen tissue turnover could not be used to predict riod are generally stably and optimally maintained. the turnover rates of other tissues. The relationship will Thus, the tissue turnover rates experimentally mea- likely need to be modeled for each type of tissue sured in laboratories are also likely obtained using studied before any predictions can be made. animals that do not exhibit metabolic rates similar to In a 2004 paper, Ogden, Hobson and Lank ob- those observed in the field. served that studies are needed to determine the However, recent data from Carleton and del Rio relationship between metabolic rate and isotope (2005) suggest that a cold-induced change in meta- turnover. They point out that variable metabolic rate, bolic rate in an avian species has a negligible effect on even among individuals of the same species, could carbon and nitrogen incorporation into red blood have a significant impact on isotope turnover rates. cells. The metabolic rates measured in this study are The study reported here presents the first examina- not basal MR, but are affected by environmental tion we are aware of relating tissue isotope turnover components that could be expected to affect MR in to metabolic rate in an effort to determine a predic- the field. These results suggest that FMR, although tive relationship between the two. While we under- different then BMR, may have only an indirect, and stand that directly applying these results to field perhaps negligible, effect on the isotopic tissue turn- situations where different types of animals experience over rate, allowing laboratory studies relating BMR variable physiological stresses would not be advisable and tissue turnover to apply to studies of animals in (Carleton and del Rio 2005), this examination has the field. been a useful step towards a better understanding of An additional interesting component of this work is the effects of metabolic rate on isotope turnover. that low nitrogen dietary intake may negate the rela- Given the strong and predictive relationship between tionship between MR and tissue turnover, as seen in metabolism and isotope turnover in adult homeo- bats (Voigt et al. 2003). Nectar-feeding bats have therms apparent from this study, we believe that higher mass-specific metabolic rates than similar sized knowledge of the metabolic rates of organisms within terrestrial mammals, and yet blood carbon turnover an ecosystem will allow researchers to make well- rates are considerably slower than those seen in ter- grounded assumptions about the isotope equilibrium restrial mammals of the same mass, seemingly negating status of each system studied. the relationship found earlier in this study. However, the bats were fed sugar-water, with little nitrogen Acknowledgments The authors would like to thank the COS- content (Voigt et al. 2003), and nectar-feeding bats in MOS Foundation and the Mellon Fund (American University) for partial funding of this study, and two anonymous reviewers the wild could also be expected to take in low levels of for their constructive comments. The experiments described in dietary nitrogen. As protein turnover and replacement this paper comply with the current laws of the United States. is a major driver in metabolism, and the bats are deficient in dietary nitrogen and thus would need to References recycle tissue components, it seems likely that both tissue nitrogen and carbon are being recycled, resulting Arneson LS, MacAvoy SE (2005) Carbon, nitrogen and sulfur in longer tissue carbon half-lives, despite the faster diet-tissue discrimination in mouse tissues. Can J Zool MR. As seen in other species (Hobson and Stirling 83:989–995 Arneson LS, MacAvoy SE, Bassett E (2006) Metabolic protein 1997; Arneson et al. 2006), the carbohydrate in the replacement drives tissue turnover in adult mice. Can J Zool nectar or sugar-water is likely primarily oxidized to 84:983–993 Ayliffe LK, Cerling TE, Robinson T, West A, Sponheimer M, provide energy and expelled as breath CO2 (Sponhei- mer et al. 2006), rather then being incorporated into Passey B, Hammer J, Roeder B, Dearing MD, Ehleringer JR (2004) Turnover of carbon isotopes in tail hair tissues. A similar disconnect between MR and tissue and breath CO2 of horses fed an isotopically varied diet. isotope turnover rate would also be expected in other Oecologia 139:11–22

123 200 Oecologia (2006) 150:190–201

Bearhop S, Waldron S, Votier SC, Furness RW (2002) Factors Hobson KA, Schell SM, Renouf D, Noseworthy E (1995) Stable that influence assimilation rates and fractionation of nitro- carbon and nitrogen isotopic fractionation between diet and gen and carbon stable isotopes in avian blood and feathers. tissues of captive seals: implications for dietary reconstruc- Physiol Biochem Zool 75:451–458 tions involving marine mammals. Can J Fish Aquat Sci Bronham L, Rambaut A, Harvey P (1996) Determinants of rate 53:528–533 variation in mammalian DNA sequence evolution. J Mol Kleiber M (1932) Body size and metabolism. Hildargia 6:315–353 Evol 43:411–439 Kleiber M (1947) Body size and metabolic rate. Physiol Rev Carleton SA, del Rio CM (2005) The effect of cold-induced in- 27:511–541 creased metabolic rate on the rate of 13C and 15N incorpo- Knoff AJ, Macko SA, Erwin RM, Brown KM (2002) Stable ration in house sparrows (Passer domesticus). Oecologia isotope analysis of temporal variation in the diets of pre- 144:226–232 fledged laughing gulls. Waterbirds 25:142–148 Chamberlain CP, Blum JD, Holmes RT, Feng X, Sherry TW, Koch PL, Phillips DL (2002) Incorporating concentration depen- Graves GR (1997) The use of isotope tracers for identifying dence in stable isotope mixing models: a reply to Robbins, populations of migratory birds. Physiol Biochem Zool Hilderbrand and Farley (2002). Oecologia 133:14–18 75:451–458 Kojeta P (1991) On the relation between basal and field meta- Evans-Ogden LJ, Hobson KA, Lank DB (2004) Blood isotopic bolic rates in birds and mammals. Funct Ecol 5:56–64 (d13C and d15N) turnover and diet-tissue fractionation factors Lindstrom A (1997) Basal metabolic rates of migrating waders in in captive dunlin (Calidris alpine pacifica). Auk 121:170–177 the Eurasian Artic. J Avian Biol 28:87–92 Frazer TK, Ross RM, Quetin LB, Montoya JP (1997) Turnover Lovegrove BG (2003) The influence of climate on the basal of carbon and nitrogen during growth of larval krill, metabolic rate of small mammals: a slow-fast metabolic Euphausia superba Dana: a stable isotope approach. J Exp continuum. J Comp Physiol B Biochem Environ Physiol Mar Biol Ecol 212:259–275 173:87–112 Fry B, Arnold C (1982) Rapid 13C/12C turnover during growth of MacAvoy SE, Macko SA, Garman GC (2001) Isotopic turnover brown shrimp (Penaeus aztecus). Oecologia 54:200–204 in aquatic predators: quantifying the exploitation of migra- Fry B, Sherr EB (1984) d13C measurements as indicators of tory prey. Can J Fish Aquat Sci 58:923–932 carbon flow in marine and freshwater ecosystems. Contrib MacAvoy SE, Macko SA, Arneson LS (2005) Growth versus Mar Sci 27:13–47 metabolic tissue replacement in mouse tissues determined Gannes LZ, O’Brien DM, del Rio CM (1997) Stable isotopes in by stable carbon and nitrogen isotope analysis. Can J Zool animal ecology: assumptions, caveats, and a call for more 83:631–641 laboratory experiments. Ecology 78:1271–1276 Maruyama A, Yamada Y, Rusuwa B, Yuma M (2001) Change in Haramis GM, Jorde DG, Macko SA, Walker JL (2001) Stable- stable nitrogen isotope ratio in the muscle tissue of a isotope analysis of canvasback winter diet in upper Chesa- migratory goby, Rhinogobius sp., in a natural setting. Can J peake Bay. Auk 118:1008–1017 Fish Aquat Sci 58:2125–2128 Hart JS (1971) Rodents. In: Whittow GG (ed) Comparative McKechnie AE, Wolf BO (2004) The allometry of avian basal physiology of thermoregulation, Academic, London, pp 1–149 metabolic rate: good predictions need good data. Physiol Herzka SZ, Holt GJ (2000) Changes in isotopic composition of Biochem Zool 77:502–521 red drum (Sciaenops ocellatus) larvae in response to dietary McNabb BK (1988) Complications inherent in scaling the basal shifts: potential applications to settlement studies. Can J rate of metabolism in mammals. Q Rev Biol 63:25–54 Fish Aquat Sci 57:137–147 Minagawa M, Wada E (1984) Stepwise enrichment of 15N along Hesslein RH, Hallard KA, Ramlal P (1993) Replacement of sulfur, food chains: further evidence and the relation between d15N carbon, and nitrogen in tissue of growing broad whitefish and animal. Geochim Cosmochim Acta 48:1135–1140 (Coregonus nasus) in response to a change in diet traced by Pearson SF, Levey DJ, Greenberg CH, del Rio CM (2003) d34S, d13C and d15N. Can J Fish Aquat Sci 50:2071–2076 Effects of elemental composition on the incorporation Heusner AA (1991) Size and power in mammals. J Exp Biol of dietary nitrogen and carbon isotopic signatures in 160:25–54 an omnivorous songbird. Oecologia 135:516–523 Hobson KA (1999) Tracing origins and migration of wildlife Peterson BJ, Howarth RW (1987) Sulfur, carbon, and nitrogen using stable isotopes: a review. Oecologia 120:314–326 isotopes used to trace organic matter flow in the salt-marsh Hobson KA, Bairlein F (2003) Isotope fractionation and turn- estuaries of Sapelo Island, Georgia. Limnol Oceanograph over in captive garden warblers (Sylvia borin): implications 32:1195–1213 for delineating dietary and migratory associations in wild Phillips DL, Gregg JW (2003) Source partitioning using stable iso- passerines. Can J Zool 81:1630–1635 topes: coping with too many sources. Oecologia 136:261–269 Hobson KA, Clark RG (1992) Assessing avian diets using stable Phillips DL, Koch PL (2002) Incorporating concentration isotopes I: turnover of 13C in tissues. Condor 94:181–188 dependence in stable isotope mixing models. Oecologia Hobson KA, Clark RG (1993) Turnover of 13C in cellular and 130:114–125 plasma reactions of blood: implications for nondestructive Ricklefs RE, Konarzweski M, Daan S (1996) The relationship sampling in avian dietary studies. Auk 110:638–641 between basal metabolic rate and daily energy expenditure Hobson KA, Stirling I (1997) Low variation in blood 13C among in birds and mammals. Am Nat 147:1047–1071 Hudson Bay polar bears: implications for metabolism and Roberts JR, Baudinette RV (1986) Thermoregulation, oxygen- tracing terrestrial foraging. Mar Mamm Sci 13:359–367 consumption and water turnover in stubble quail, Coturnix Hobson KA, Wassenaar LI (1999) Stable isotope ecology: an pectoralis, and king quail, Coturnix chinensis. Aust J Zool introduction. Oecologia 120:312–313 34:25–33 Hobson KA, McLellan BN, Woods JG (2000) Using stable car- Sponheimer M, Robinson TF, Cerling TE, Tegland L, Roeder bon (d13C) and nitrogen (d15N) isotope to infer trophic BL, Ayliffe L, Dearing MD, JR Erleringer (2006) Turnover relationships among black and grizzly bears in the upper of stable carbon isotopes in the muscle, liver, and breath Columbia River basin, British Columbia. Can J Zool CO2 of alpacas (Lama pacos). Rapid Commun Mass Spec- 78:1332–1339 trom 20:1395–1399

123 Oecologia (2006) 150:190–201 201

Tieszen LL, Farge T (1993) Effect of diet quality and composi- nitrogen isotope ratios (d15N) and literature dietary data. tion on the isotopic composition, CO2, bone collagen, bio- Can J Fish Aquat Sci 54:1142–1158 apatite, and soft tissues. In: Lambert J, Groupe G (eds) Voigt CC, Matt F, Michener R, Kunz TH (2003) Low turnover Molecular archeology of prehistoric human bone. Springer, rates of carbon isotopes in tissues of two nectar-feeding bat Berlin Heidelberg New York, pp 123–135 species. J Exp Biol 206:1419–1427 Tieszen LL, Boutton TW, Tesdahl KG, Slade NA (1983) Frac- White CR, Seymour RS (2003) Mammalian basal metabolic rate tionation and turnover of stable carbon isotopes in animal is proportional to body mass2/3. Proc Natl Acad Sci USA tissues: implications for d13C analysis of diet. Oecologia 100:4046–4049 57:32–37 Woodin M, Stephenson R (1998) Circadian rhythms in diving Vander Zanden MJ, Cabana G, Rasmussen JB (1997) Compar- behavior and ventilatory response to asphyxia in canvasback ing trophic position of freshwater ﬁsh calculated using stable ducks. AmJ Physiol 274:R686–R692

123