Received 9 July 2002 Accepted 16 September 2002 Published online 10 December 2002

Adaptation of metabolism and evaporative water loss along an aridity gradient

B. Irene Tieleman 1 * , Joseph B. Williams2 and Paulette Bloomer3
Zoological Laboratory, University of Groningen, PO Box 14, 9750 AA Haren, The Netherlands Department of Evolution, Ecology and Organismal Biology, Ohio State University, 1735 Neil Avenue, Columbus, OH 43210, USA 3 Molecular Ecology and Evolution Program, Department of Genetics, University of Pretoria, 0002 Pretoria, South Africa
1 2

Broad-scale comparisons of birds indicate the possibility of adaptive modification of basal metabolic rate (BMR) and total evaporative water loss (TEWL) in species from desert environments, but these might be confounded by phylogeny or phenotypic plasticity. This study relates variation in avian BMR and TEWL to a continuously varying measure of environment, aridity. We test the hypotheses that BMR and TEWL are reduced along an aridity gradient within the lark family (Alaudidae), and investigate the role of phylogenetic inertia. For 12 species of lark, BMR and TEWL decreased along a gradient of increasing aridity, a finding consistent with our proposals. We constructed a phylogeny for 22 species of lark based on sequences of two mitochondrial genes, and investigated whether phylogenetic affinity played a part in the correlation of phenotype and environment. A test for serial independence of the data for masscorrected TEWL and aridity showed no influence of phylogeny on our findings. However, we did discover a significant phylogenetic effect in mass-corrected data for BMR, a result attributable to common phylogenetic history or to common ecological factors. A test of the relationship between BMR and aridity using phylogenetic independent constrasts was consistent with our previous analysis: BMR decreased with increasing aridity. Keywords: basal metabolic rate; total evaporative water loss; aridity; larks; Alaudidae; phylogeny 1. INTRODUCTION Terrestrial vertebrates continuously expend energy to carry out chemical processes necessary to sustain life, and constantly lose water through respiration, cutaneous evaporation and excretion. Rates of energy expenditure and water loss vary considerably among and within vertebrate taxa even under standard laboratory conditions. The reasons underlying this variation have been the focus of intensive research and are still only partially understood (Kleiber 1961; Bartholomew & Cade 1963; Crawford & Lasiewski 1968; Aschoff & Pohl 1970; McNab 1986; Nagy 1987; Bennett 1988; Williams 1996; Mueller & Diamond 2001). Basal metabolic rate (BMR), the minimum energy expenditure of a fasted endotherm in its rest phase at thermoneutral temperatures (King 1974), integrates all catabolic energy transformations required for bodily maintenance. BMR correlates with the energy expenditure of free-living animals (Nagy 1987; Daan et al. 1990; Ricklefs et al. 1996) and with life-history attributes, such as growth and reproduction (Kleiber 1961; Bennett 1988; Trevelyan et al. 1990; Harvey et al. 1991; Hulbert & Else 2000). Total evaporative water loss (TEWL), the sum of respiratory and cutaneous water losses, constitutes a significant proportion of an animals total water loss, up to 7080% in small birds when measured at 25 C (Lee & SchmidtNielsen 1971; Bartholomew 1972; Dawson 1982; Williams 1996). Both BMR and TEWL appear, in part, to be genetically determined, suggesting that natural selection could operate on these variables (Furuyama & Ohara 1993; Konarzewski & Diamond 1995). After correcting for body mass and higher level taxonomic affiliations, a large variation in metabolism and TEWL remains (Kleiber 1961; Crawford & Lasiewski 1968; Dawson 1982; McNab 1988; Williams 1996). Differences in BMR and TEWL have been reported between species or populations differing in diet, altitude, latitude, temperature and aridity (Dawson & Bennett 1973; Weathers 1979; Nagy 1987; McNab 1988; Nagy et al. 1999; Williams & Tieleman 2001), but a unifying environmental characteristic that determines the metabolic physiology of animals has not been identified. Mueller & Diamond (2001) reported that primary productivity explains 88% of the variation in BMR between five species of Peromyscus mice, and suggested that food availability, a correlate of primary productivity, might be a unifying explanatory variable for metabolic physiology. One might also predict that water availability is a major factor explaining variation in TEWL. Previous work on BMR and TEWL has compared species from disparate environments, such as the tropics versus temperate or desert versus non-desert (Scholander et al. 1950; Bartholomew & Cade 1963; Dawson & SchmidtNielsen 1964; Dawson & Bennett 1973; Weathers 1979; Dawson 1984; Williams 1996; Tieleman & Williams 2000). However, the use of a dichotomous categorization of environments might obscure useful biological information. Climatologists have long recognized that environments form a continuum with respect to meteorological parameters, and have emphasized that, for example, the environment of a given desert depends on the interaction 207
2002 The Royal Society

Author for correspondence ([email protected]).

Proc. R. Soc. Lond. B (2003) 270 , 207214 DOI 10.1098/rspb.2002.2205

208 B. I. Tieleman and others Metabolism and evaporative water loss of larks

of a number of variables, including temperature, amount and timing of rainfall, relative humidity and wind (Thornthwaite 1948; Meigs 1953). To the best of our knowledge, our study is the first to relate variation in avian physiological variables to a continuously varying measure of environment, aridity. Aridity is directly related to primary productivity (Emberger 1955) and provides a proxy for the selection pressures that animals experience with increasing aridity, including decreasing water and food availability and increasing air temperatures. The challenge in understanding interspecific correlations between phenotype and environment is to distinguish the contributions of the various processes that underlie the pattern, including genetic adaptation, phylogenetic inertia (Gould & Lewontin 1979; Westoby et al. 1995; Hansen 1997) and phenotypic responses (acclimatization). Our limited understanding of the time-scale at which natural selection and other evolutionary processes operate has resulted in two main approaches to questions in evolutionary biology, the phylogenetic comparative method and the optimality approach. Comparative methods that take into account potential effects of phylogeny emphasize the historical component of adaptation (Gould & Vrba 1982) and implicitly rule out the possibility that a trait is maintained as an adaptation by stabilizing selection (Frumhoff & Reeve 1994; Westoby et al. 1995; Hansen 1997). By contrast, optimality studies assume that traits with ample genetic variation, such as most quantitative characters (Houle 1992; Lynch 1988), are probably maintained by stabilizing selection (Hansen 1997). A comparative approach with an optimality viewpoint of adaptation would assume that each species is at an optimum (Hansen 1997). One way to reduce problems relating to differences in phylogenetic history is to compare traits among closely related species (Coddington 1988; Bennett 1988; Price 1991; Leroi 1994). The family of larks (Alaudidae) occurs over a wide range of environments and continents, with representatives in habitats ranging from hyperarid deserts to mesic grasslands (Cramp 1988; Pa tzold 1994). Because all larks eat similar foods, a mixture of insects and seeds, diet is not a confounding variable in our analyses. This family provides a model to investigate physiological adaptation to environment. This study tests the proposal that lower BMRs and TEWLs in birds are correlated with increasing aridity, and investigates the role of phylogenetic inertia in shaping the relationship between physiology and environment. We constructed a phylogeny of larks based on two mitochondrial genes, and investigated whether phylogeny could explain the variation in BMR and TEWL. We used conventional and, when appropriate, phylogenetically corrected analyses to examine the relationships between aridity, BMR and TEWL. 2. MATERIAL AND METHODS (a) Animals
We mist-netted hoopoe larks (Alaemon alaudipes), Dunns larks (Eremalauda dunni ), desert larks (Ammomanes deserti ), black-crowned finchlarks (Eremopterix nigriceps) and crested larks (Galerida cristata) at various sites in the west-central Arabian Desert (table 1), and housed them in outdoor aviaries at the
Proc. R. Soc. Lond. B (2003)

National Wildlife Research Center (NWRC), near Taif, Saudi Arabia. Calandra larks (Melanocorypha calandra) from Iran were transported to Saudi Arabia and kept at the NWRC. We caught skylarks (Alauda arvensis) and woodlarks (Lullula arborea) in the province of Drenthe, The Netherlands, and kept them in outdoor aviaries at the Zoological Laboratory of the University of Groningen. We measured BMR and TEWL of all birds in Saudi Arabia and The Netherlands between June and August of 1998 2001. Birds were measured in the post-absorptive state during their nocturnal phase by standard flow-through respirometry and hygrometry methods (Williams & Tieleman 2000; Tieleman et al. 2002). In addition, we used data for spike-heeled larks (Chersomanes albofasciata) from Kimberley, South Africa (C. Brown, unpublished data), for short-toed larks (Calandrella brachydactyla) and lesser short-toed larks (Calandrella rufescens) from the North Caspian region, Russian Federation (Shishkin 1980), for Starks lark (Eremalauda starki ) and grey-backed finchlarks (Eremopterix verticalis) from the Namib Desert (Willoughby 1968) and for horned larks (Eremophila alpestris) from North America (Trost 1972). For all these species, birds were measured during spring or summer, in the post-absorptive state and during their rest phase.

(b) Environmental aridity

We calculated an aridity index as Q = P /((Tm a x 1 Tm in )(Tm a x 2 Tm i n )) 1000, where P is the average annual precipitation (mm), Tm a x is the mean maximum temperature of the hottest month (C) and Tm i n is the mean minimum temperature of the coldest month (C) (Emberger 1955). Although not perhaps intuitively straightforward, this index was empirically derived to describe primary productivity in arid and semi-arid areas (Emberger 1955). The index is low in hot dry deserts and high in cool wet areas. We collected climatic data from local or national meteorological institutes, from the literature (Walter & Lieth 1967; Williams 2001) and from http://www.worldclimate. com/ and http://www.onlineweather.com/ (table 1). Because Q increases rapidly when environments become more mesic, we avoided unequal weighing of data for mesic species by using log Q in our analyses (table 1).

(c) Phylogeny of larks

The geographical origins of the DNA samples of all larks in our phylogeny are available on request from the authors. DNA was extracted from blood or tissue samples using standard protocols (50 mM of Tris, pH 7.6; 100 mM of NaCl; 1 mM of ethylenediamine tetra-acetic acid (EDTA), pH 8.0; 0.5% sodium dodecyl sulphate (SDS), 1 mg ml2 1 of proteinase K and 0.1 mg of Rnase A). Impurities were removed by phenol chloroform extraction, and total genomic DNA was ethanolprecipitated and eluted in sterile distilled water. Polymerase chain reaction (PCR) amplification (Saiki et al. 1988) followed standard protocols (Kocher et al. 1989). The cytochrome b gene was amplified using two primer sets: L14990 (shortened primer L14841 of Kocher et al. (1989)) and H15696 (primer H15547 of Edwards et al. (1991)); and L15245 (modified primer CB4a-L of Palumbi et al. (1991)) and H16064 (located in the tRNA t h r ). A portion of the 16S rRNA gene was amplified using primers L2313 and H4015 (Lee et al. 1997). Successful amplicons were purified using a High Pure PCR Purification kit (Roche Diagnostics). We sequenced DNA using the four PCR primers for the cytochrome b gene, and H4015 and an internal primer, L2925, designed for passerines in our laboratory (59 AGCCATCAA-

Metabolism and evaporative water loss of larks

B. I. Tieleman and others 209

Table 1. Geographical origin and environmental aridity for 14 species of lark. The environmental aridity index (Q) was calculated following Emberger (1955), based on precipitation (P ), maximum temperature (Tm a x ) and minimum temperature (Tm i n ). species hoopoe lark a Dunns lark a desert lark a Starks lark b grey-backed finchlarkb crested lark a black-crowned finchlark a calandra lark c ,d horned lark e short-toed lark c ,f lesser short-toed lark c ,f spike-heeled lark c ,f woodlark g skylarkg latitude 22159 N 22159 N 22559 N 2334 9 S 2334 9 S 21159 N 21159 N 35009 N 33179 N/ 34359 N 49259 N 49259 N 2848 9 S 52529 N 52529 N longitude 4150 9 E 4150 9 E 4109 9 E 1503 9 E 1503 9 E 4142 9 E 4142 9 E 5100 9 E 116409 W/ 11806 9 W 4651 9 E 4651 9 E 2442 9 E 5252 9 E 5252 9 E log Q 1.78 1.78 1.78 1.76 1.76 2.24 2.24 2.26 2.41 2.59 2.59 2.60 3.20 3.20 P (mm) 89.6 89.6 89.6 57.2 57.2 209.1 209.1 250.0 309.9 281.0 281.0 420.4 750.0 750.0 Tm a x (C) 40.2 40.2 40.2 33.0 33.0 35.7 35.7 37.0 34.8 30.2 30.2 32.6 21.7 21.7 Tm i n (C) 10.7 10.7 10.7 10.0 10.0 8.1 8.1 22.4 21.2 213.7 213.7 2.7 20.7 20.7

Climate data sources: a National Wildlife Research Center, Taif, Saudi Arabia; b Williams (2001), includes fog precipitation; c Walter & Lieth (1967); d http://www.onlineweather.com; e US National Weather Service; f http://www.worldclimate.com; and g Koninklijk Nederlands Meteorologisch Instituut.
CAAAGAGTGCG 39) for the 16S rRNA gene. Sequences of the light and heavy strands were determined using dye-terminator cycle sequencing (Big Dye DNA sequencing kit, Applied Biosystems) and an ABI377 or ABI3100 automated DNA sequencer (Applied Biosystems). Sequences for each taxon were proofread in Sequence Navigator, and complete sequences were aligned in Clustal X (Thompson et al. 1997). We analysed the aligned sequences with Paup 4.0 using multiple heuristic searches with default settings (Swofford 1998). We assessed the resolution of internal nodes using 1000 bootstrap replicates with random replacement (Felsenstein 1985a). Phylogenetic signal was determined by evaluating the tree-length distribution of 1000 randomly generated trees (Hillis & Huelsenbeck 1992). In addition to using unweighted parsimony analysis, we attempted to reduce homoplasy by downweighting characters based on their consistency indices (Farris 1969). Pairwise estimates of nucleotide sequence divergence were calculated using the HKY85 model (Hasegawa et al. 1985). Gene sequences are deposited in GenBank (accession numbers AY165123AY165147 (16S rRNA) and AY165148AY165172 (cyt b)). for serial independence is more suitable than other phylogenetic autocorrelation methods for small datasets (Cheverud et al. 1985; Gittleman & Kot 1990; Martins & Hansen 1996; Abouheif 1999). If no phylogenetic effect exists, then incorporating phylogeny in statistical methods is unnecessary (Gittleman & Kot 1990; Bjo rklund 1997; Abouheif 1999). If a phylogenetic effect does exist, it may be attributable to phylogenetic constraint or to ecological factors, and corrections for phylogenetic relationships may or may not be appropriate (Westoby et al. 1995).

(e) Statistics
We performed analyses of variance and regression analyses with SPSS 10.0, and calculated phylogenetic independent contrasts (Felsenstein 1985b) with the PDTree module in the computer program PDAP (Garland et al. 1992). We calculated the degrees of freedom as N 2 P u , where N is the number of independent contrasts and Pu is the number of unresolved polytomies (Purvis & Garland 1993).

3. RESULTS (a) Physiology and environment For 12 species of lark, BMR was related to body mass as log BMR (kJ day2 1 ) = 0.225 1 0.901 log mass (g) (r2 = 0.53, d.f. = 11, s.e.s lo p e = 0.269, p = 0.007; table 2). A multiple regression analysis with log BMR as the dependent variable and log mass and log Q (aridity) as independent variables showed that both body mass (t = 6.17, p , 0.0001) and aridity (t = 6.08, p , 0.0001) had a significant effect on BMR: log BMR = 20.194 1 0.845 log mass 1 0.208 aridity (r2 = 0.91, d.f. = 11, p , 0.0001). The BMR of larks increased as the environment became more mesic (figure 1a). TEWL among larks was related to body mass as log TEWL (g day2 1 ) = 20.814 1 0.816 log mass (g) (r2 = 0.72, d.f. = 11, s.e.s lo p e = 0.162, p = 0.001; table 2). We used multiple regression analysis to assess the effect of aridity on TEWL: log TEWL = 20.903 1 0.684 log

(d ) Phylogenetic effect
Correlations between species may be statistically biased if sister taxa tend to be similar to one another as a result of common ancestry (Felsenstein 1985b; Cheverud et al. 1985; Harvey & Pagel 1991). To evaluate whether a phylogenetic effect (sensu Grafen (1989) and Harvey & Pagel (1991)) exists among the larks in this study, we used tests for serial independence to determine whether there was a significant positive autocorrelation for mass-corrected BMR, mass-corrected TEWL or aridity (Abouheif 1999; Reeve & Abouheif 1999). In each simulation the topology was randomly rotated 2000 times per iteration and the original data were shuffled 2000 times to provide the null hypothesis sampling distribution (Reeve & Abouheif 1999). We calculated mass-corrected BMR and mass-corrected TEWL by dividing BMR and TEWL, respectively, by massx , where x is the slope of the allometric equations relating log BMR and log TEWL to log body mass in 12 lark species (see 3). The test
Proc. R. Soc. Lond. B (2003)

210 B. I. Tieleman and others Metabolism and evaporative water loss of larks
Table 2. Basal metabolic rate (BMR) and total evaporative water loss (TEWL) for 14 species of larks. mass (g) 21.5 20.9 36.9 15.2 31.2 50.6 26.0 23.6 24.0 25.7 32.0 25.6
b

species grey-backed finchlarka Starks lark a desert lark Dunns lark hoopoe lark black-crowned finchlark crested lark calandra lark horned lark b lesser short-toed lark c short-toed lark c spike-heeled lark d skylark woodlark Data sources: a Willoughby (1968); e Day-time values.

BMR s.d. BMR (kJ day2 1 ) (kJ day2 1 ) 20.1 24.7 32.8 16.5 32.2 49.5 28.6 31.6 35.6 29.1 62.4 49.4
c

n 6 22 21 6 6 2 27 8 20 29 20
d

mass (g) 15.1 15.6 21.5 20.5 36.9 15.2 31.2 50.6 26.0 25.0 31.7 25.5

TEWL (g day2 1 ) 1.31e 1.31e 1.60 1.69 2.59 1.34 2.44 3.03 2.08 3.33 3.47 2.41

s.d. TEWL (g day2 1 ) 0.26 0.49 0.62 0.36 0.24 0.57 0.53 0.88 0.70

n 6 16 21 6 6 2 10 15 14

2.46 2.61 4.45 1.10 2.30 1.07 3.08 6.73 4.98 8.43 9.96 Shishkin (1980);

Trost (1972);

C. Brown (unpublished data).

a) BMR kJ d1 mass 0.845)

3.5 3.0 2.5 2.0 1.5 1.0

as independent variables. The models with the best fit to the data (see 3a) included log mass and aridity, but not latitude (BMR: t = 1.85, p = 0.10; TEWL: t = 1.17, p = 0.28). Hence, aridity explained more of the variation in BMR and TEWL than did latitude after body mass was taken into account. (b) Phylogeny of larks Cytochrome b (975 bases) and 16S rRNA (566 bp) sequences were generated for 22 species. In addition to the species for which ecophysiological data were available, we included other African species in an attempt to improve the resolution of the phylogenetic placement of our focal taxa. A heuristic search of the combined dataset yielded two equally most-parsimonious trees (length 1424 steps, CI = 0.407, RI = 0.492, g1 = 21.018, p , 0.001). One round of reweighting (using a base weight of 1) yielded a single tree of 579.27 steps (figure 2; CI = 0.490, RI = 0.559). Out of the 378 parsimonyinformative characters, 41 had a weight of 1 and 337 had a weight of less than 1, reflecting homoplasy in the dataset. We performed bootstrap analysis with 1000 iterations using the hoopoe lark as an outgroup (figure 2). Several highly supported clades (more than 70% bootstrap support) were consistently retrieved (also in separate analyses of the two genes, trees not shown): basal placement of spike-heeled lark clade; finchlarks/desert larks clade; singing bush lark/monotonous lark clade; lesser short-toed/Athi short-toed lark clade; and skylark/ woodlark clade. Several currently recognized genera appear to be polyphyletic (e.g. Ammomanes, Certhilauda, Mirafra). A lack of resolution of some of the terminal nodes may be caused by the relatively short internal branch lengths compared with the long terminal branch lengths. The divergence among these lark clades is also indicated by genetic distances ranging from 7 to 19% based on the cytochrome b gene and between 2.5 and 8% based on the 16S rRNA fragment. We did not have DNA sequence information for the calandra lark, horned lark and short-toed lark. To apply the test for serial independence to all species for which we

b) TEWL g d1 mass 0.684)

0.40 0.35 0.30 0.25 0.20 0.15 1.5 2.0 2.5 3.0 aridity index log Q) 3.5

Figure 1. Mass-adjusted (a) BMR and (b) TEWL of 12 species of lark as a function of environmental aridity. Symbols are species averages.

mass 1 0.121 aridity (r2 = 0.87, d.f. = 11, p , 0.0001). Mass and aridity both had a significant effect on TEWL (log mass: t = 5.50, p , 0.0001; aridity: t = 3.12, p = 0.011), a finding consistent with the hypothesis that larks have a lower TEWL with increasing aridity (figure 1b). Because aridity was correlated with latitude (table 1; r = 0.84, n = 9, p = 0.005), we performed stepwise multiple regression analyses with log BMR or log TEWL as the dependent variable and log mass, latitude and aridity
Proc. R. Soc. Lond. B (2003)

Metabolism and evaporative water loss of larks 12.588 15.263 24.365 22.878 15.120 15.981 14.790 21.340 crested lark large-billed lark skylark

B. I. Tieleman and others 211

23.294 100) 16.037 100) 10.532 99) 6.175 92)

black-crowned finchlark grey-backed finchlark bar-tailed desert lark desert lark

2.137 52) 6.564 95) 3.594 63) 5.661 88) 2.000 100) 8.207 88)

10.515 88) 26.150 6.735 86)

15.869 19.361 14.307

7.236 90) 22.360

woodlark , Temminck s horned lark , Stark s lark , Sclater s lark lesser short-toed lark Athi short-toed lark , Dunn s lark

19.017

10.234 99) 13.840 98) 18.664 21.547 23.588

14.032 21.383

singing bushlark monotonous lark dune lark sabota lark spike-heeled lark , Gray s lark long-billed lark hoopoe lark

6.430 96) 27.317 10.955 100) 24.838 32.882

Figure 2. Phylogenetic tree of 22 species of lark based on cytochrome b and 16S rRNA sequences, and analysed using maximum parsimony criteria. Numbers above the branches indicate reweighted branch lengths and, in parentheses, per cent bootstrap recovery in 1000 replications.

had ecophysiological data, we placed the calandra lark in the polytomy with dune lark/sabota lark, the finchlark/desert lark clade and a third clade containing Galerida, Lullula, Calandrella and others. We placed the horned lark as sister species to the Temmincks horned lark. We assumed that the short-toed lark was closely related to the Athi short-toed lark. (c) Phylogenetic constraint on BMR and TEWL? We found no significant phylogenetic autocorrelation in the data for mass-corrected TEWL or aridity, but did find a significant autocorrelation in the data for mass-corrected BMR (mass-corrected TEWL: p = 0.44; aridity: p = 0.09; mass-corrected BMR: p = 0.01). Therefore, the positive association between TEWL and aridity was not influenced by phylogenetic constraint. The significant phylogenetic autocorrelation in the mass-corrected BMR data could be attributable to phylogenetic constraint or to ecological factors, indistinguishable alternatives (Westoby et al. 1995). Therefore, in addition to our conventional analysis, we tested whether the relationship between mass-corrected BMR and aridity could be confirmed using phylogenetic independent contrasts (Felsenstein 1985b). A stepwise multiple regression analysis with contrasts of log BMR as the dependent variable, and contrasts of log mass and of aridity as independent variables, showed significant effects of log mass and aridity on log BMR (r2 = 0.88, d.f. = 9, F = 32.36, p , 0.001; log mass: t = 6.07, p , 0.001; aridProc. R. Soc. Lond. B (2003)

ity: t = 4.83, p = 0.001). Hence, the result of the phylogenetic analysis was consistent with that of the conventional analysis: the BMR of larks decreases with increasing aridity. 4. DISCUSSION The BMR (n = 12) and TEWL (n = 12) of lark decreased along a gradient of increasing aridity, consistent with our proposal. These results support the conclusions of previous studies that found reduced BMR and TEWL in desert birds compared with species from mesic habitats (Dawson & Bennett 1973; Weathers 1979; Arad & Marder 1982; Williams 1996; Tieleman & Williams 2000), but the use of a continuous environmental classification in this study makes the argument more compelling. Although body mass alone explained 53% of the interspecific variation in BMR, adding aridity increased the explained variance to 91%. Similarly, body mass explained 72% of the variance in TEWL, while adding aridity increased the explained variance to 87%. A combination of low BMR and low TEWL could be favourable in birds from dry hot environments because it reduces food and water requirements and minimizes heat production. The value of comparisons within a group of closely related species is illustrated by the deviations of the allometries for BMR and TEWL of larks from those of all

212 B. I. Tieleman and others Metabolism and evaporative water loss of larks

birds. The equation relating log body mass to log BMR in larks deviated from an allometry for all birds by 29% for a 15 g lark and by 125% for a 50 g lark (Tieleman & Williams 2000). The BMR of desert larks was close to allometric predictions, in contrast to the BMR of mesic larks, which far exceeded predictions. Predictions of TEWL based on the lark allometry were lower than those based on an allometry including all birds (Williams 1996) by 226% for a 15 g lark and by 212% for a 50 g lark. Although the TEWL of a lark from the desert was below allometric predictions, as expected, larks in general had a low TEWL. Hence, comparing the BMR or TEWL of a single lark species with the all-bird allometry would have led to erroneous conclusions about the adaptive significance of these traits. The phylogeny of larks was characterized by long terminal branches and short internal branch lengths (figure 2). This, combined with the occurrence of larks in a wide array of habitats, may indicate that lark species rapidly radiated into different environments and lived for a significant part of their evolutionary history in diverse habitats. One might predict that phylogenetic constraints are small, natural selection has had ample time to eliminate suboptimal phenotypes and the current trait values are of adaptive significance in the current environment. The lack of phylogenetic effect in the mass-corrected TEWL data implies that phylogenetic relatedness is not a major evolutionary factor explaining interspecific differences in TEWL. The positive phylogenetic autocorrelation in the mass-corrected BMR data could indicate either a phylogenetic constraint or closely related species experiencing common ecological factors (Westoby et al. 1995). The correlation between BMR and aridity is not the result of phylogenetic constraint, because the results of the contrast analysis are consistent with those of the conventional statistics. Therefore, evolutionary forces other than phylogenetic constraint are likely to underlie the correlations between BMR, TEWL and environmental aridity. Interspecific phenotypeenvironment correlations can indicate either genetic differences brought about by natural selection or phenotypically plastic responses to environmental conditions. Plastic responses include changes in adult phenotype depending on the environment (acclimation) and differences between phenotypes resulting from developmental conditions (ontogenetic plasticity). In a separate study, we found that the decrease in BMR and TEWL in larks along an aridity gradient cannot be attributed to acclimation to thermal environment, food availability or photoperiod (Tieleman et al. 2003). In summary, decreasing levels of BMR and TEWL in larks correlate with increasing aridity. These physiological traits may have adaptive significance in the current environment, and natural selection is a likely process to explain our findings. Identifying evolutionary processes that cause correlative associations between traits remains difficult (Leroi 1994), but we have eliminated phylogenetic constraint and acclimation as likely alternative explanatory processes. The results of this study together with the decrease in BMR in mice along an environmental gradient of decreasing primary productivity (Mueller & Diamond 2001) suggest that the decrease in energy and water requirements with decreasing food and water availability is a general pattern found in birds and mammals.
Proc. R. Soc. Lond. B (2003)

The authors thank P. Paillat, A. Khoja, P. Seddon, M. Shobrak, S. Ostrowski, J.-Y. Cardona and the other staff at the National Wildlife Research Centre (NWRC), Taif, Saudi Arabia, for logistic support throughout this study. Wildlife research programmes at the NWRC are possible through the generous support of HRH Prince Saud al Faisal and under the guidance of A. Abuzinada of the National Commission of Wildlife Conservation and Development. We are grateful to C. Brown and K. Barnes for providing unpublished data on the spike-heeled lark and Athi short-toed lark, respectively, and to A. Kuzmenko for translating a Russian article. We are grateful to W. Delport for assistance with DNA sequencing. G. Overkamp and the animal caretakers at the Zoological Laboratory provided valuable help and advice. We thank S. Daan, R. Ricklefs and two anonymous referees for commenting on a previous draft. Financial support for this study was made available by the Schuurman Schimmel van Outeren Foundation, the Schure Beijerinck Popping Foundation, the National Wildlife Research Centre, the University of Groningen, the Ohio State University and the National Science Foundation (IBN0212092).

REFERENCES
Abouheif, E. 1999 A method for testing the assumption of phylogenetic independence in comparative data. Evol. Ecol. Res. 1, 895909. Arad, Z. & Marder, J. 1982 Comparative thermoregulation of four breeds of fowls (Gallus domesticus) exposed to gradual increase of ambient temperatures. Comp. Biochem. Physiol. A 72, 179184. Aschoff, J. & Pohl, H. 1970 Rhythmic variations in energy metabolism. Fed. Proc. 29, 15411552. Bartholomew, G. A. 1972 The water economy of seed-eating birds that survive without drinking. In Proc. 15th Int. Ornithol. Congr., The Hague (ed. K. H. Voous), pp. 237254. Leiden: E. J. Brill. Bartholomew, G. A. & Cade, T. J. 1963 The water economy of land birds. Auk 80, 504539. Bennett, A. F. 1988 Structural and functional determinates of metabolic rate. Am. Zool. 28, 699708. Bjo rklund, M. 1997 Are comparative methods always necessary? Oikos 80, 607612. Cheverud, J. M., Dow, M. M. & Leutenegger, W. 1985 The quantitative assessment of phylogenetic constraints in comparative analyses: sexual dimorphism in body weight among primates. Evolution 39, 13351351. Coddington, J. A. 1988 Cladistic tests of adaptational hypotheses. Cladistics 4, 322. Cramp, S. 1988 Handbook of the birds of Europe, the Middle East and North Africa. Oxford University Press. Crawford, E. C. & Lasiewski, R. C. 1968 Oxygen consumption and respiratory evaporation of the emu and rhea. Condor 70, 333339. Daan, S., Masman, D. & Groenewold, A. 1990 Avian basal metabolic rates: their association with body composition and energy expenditure in nature. Am. J. Physiol. 259, R333 R340. Dawson, W. R. 1982 Evaporative losses of water by birds. Comp. Biochem. Physiol. A 71, 495509. Dawson, W. R. 1984 Physiological studies of desert birds: present and future considerations. J. Arid Environ. 7, 133 155. Dawson, W. R. & Bennett, A. F. 1973 Roles of metabolic level and temperature regulation in the adjustment of western plumed pigeons (Lophophaps ferruginea) to desert conditions. Comp. Biochem. Physiol. A 44, 249266. Dawson, W. R. & Schmidt-Nielsen, K. 1964 Terrestrial animals in dry heat: desert birds. In Handbook of physiology: adaptation to the environment (ed. C. G. Wilber, E. F.

Metabolism and evaporative water loss of larks

Adolph & D. B. Dill), pp. 481492. Washington, DC: American Physiological Society. Edwards, S. V., Arctander, P. & Wilson, A. C. 1991 Mitochondrial DNA resolution of a deep branch in the genealogical tree of perching birds. Proc. R. Soc. Lond. B 243, 99107. Emberger, L. 1955 Afrique du Nord-Ouest. In Plant ecology: reviews of research (ed. UNESCO), pp. 219249. Paris, France: UNESCO. Farris, J. S. 1969 A successive approximation approach to character weighting. Syst. Zool. 18, 374385. Felsenstein, J. 1985a Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783791. Felsenstein, J. 1985b Phylogenies and the comparative method. Am. Nat. 125, 115. Frumhoff, P. C. & Reeve, H. K. 1994 Using phylogenies to test hypotheses of adaptation: a critique of some current proposals. Evolution 48, 172180. Furuyama, F. & Ohara, K. 1993 Genetic development of an inbred rat strain with increased resistance adaptation to a hot environment. Am. J. Physiol. 265, R957R962. Garland Jr, T., Harvey, P. H. & Ives, A. R. 1992 Procedures for the analysis of comparative data using phylogenetically independent contrasts. Syst. Biol. 41, 1832. Gittleman, J. L. & Kot, M. 1990 Adaptation: statistics and a null model for estimating phylogenetic effects. Syst. Zool. 39, 227241. Gould, S. J. & Lewontin, R. C. 1979 The Spandrels of San Marco and the Panglossian paradigm. A critique of the adaptationist program. Proc. R. Soc. Lond. B 205, 581598. Gould, S. J. & Vrba, E. S. 1982 Exaptationa missing term in the science of form. Paleobiology 8, 415. Grafen, A. 1989 The phylogenetic regression. Phil. Trans. R. Soc. Lond. B 326, 199257. Hansen, T. F. 1997 Stabilizing selection and the comparative analysis of adaptation. Evolution 51, 13411351. Harvey, P. H. & Pagel, M. D. 1991 The comparative method in evolutionary biology. Oxford University Press. Harvey, P. H., Pagel, M. D. & Rees, J. A. 1991 Mammalian metabolism and life histories. Am. Nat. 137, 556566. Hasegawa, M., Kishino, H. & Yano, T. 1985 Dating of the human-ape splitting by the molecular clock of mitochondrial DNA. J. Mol. Evol. 22, 160174. Hillis, D. M. & Huelsenbeck, J. P. 1992 Signal, noise, and reliability in molecular phylogenetic analysis. J. Heredity 83, 189195. Houle, D. 1992 Comparing evolvability and variability of quantitative traits. Genetics 130, 195204. Hulbert, A. J. & Else, P. L. 2000 Mechanisms underlying the cost of living in animals. A. Rev. Physiol. 62, 207235. King, J. R. 1974 Seasonal allocation of time and energy resources in birds. In Avian energetics (ed. R. A. Paynter Jr), pp. 485. Cambridge, MA: Nuttall Ornithological Club. Kleiber, M. 1961 The fire of life: an introduction to animal energetics. New York: Wiley. Kocher, T. D., Thomas, W. K., Meyer, A., Edwards, S. V., Paabo, S., Villablanca, F. X. & Wilson, A. C. 1989 Dynamics of mitochondrial DNA evolution in animals: amplification and sequencing with conserved primers. Proc. Natl Acad. Sci. USA 86, 61966200. Konarzewski, M. & Diamond, J. 1995 Evolution of basal metabolic rate and organ masses in laboratory mice. Evolution 49, 12391248. Lee, K., Feinstein, J. & Cracraft, J. 1997 The phylogeny of ratite birds: resolving conflicts between molecular and morphological data sets. In Avian molecular evolution and systematics (ed. D. P. Mindell), pp. 173211. San Diego, CA: Academic Press.
Proc. R. Soc. Lond. B (2003)

B. I. Tieleman and others 213

Lee, P. & Schmidt-Nielsen, K. 1971 Respiratory and cutaneous evaporation in the zebra finch: effect on water balance. Am. J. Physiol. 220, 15981605. Leroi, A. M. 1994 What does the comparative method reveal about adaptation? Am. Nat. 143, 381402. Lynch, M. 1988 The rate of polygenic mutation. Genet. Res. 51, 137148. McNab, B. K. 1986 The influence of food habits on the energetics of eutherian mammals. Ecol. Monogr. 56, 119. McNab, B. K. 1988 Food habits and the basal rate of metabolism in birds. Oecologia 77, 343349. Martins, E. P. & Hansen, T. F. 1996 The statistical analysis of interspecific data: a review and evaluation of phylogenetic comparative methods. In Phylogenies and the comparative method in animal behavior (ed. E. P. Martins), pp. 2275. New York: Oxford University Press. Meigs, P. 1953 Reviews of research on arid zone hydrology. Paris, France: UNESCO. Mueller, P. & Diamond, J. 2001 Metabolic rate and environmental productivity: well-provisioned animals evolved to run and idle fast. Proc. Natl Acad. Sci. USA 98, 12 55012 554. Nagy, K. A. 1987 Field metabolic rate and food requirement scaling in mammals and birds. Ecol. Monogr. 57, 111128. Nagy, K. A., Girard, I. A. & Brown, T. K. 1999 Energetics of free-ranging mammals, reptiles and birds. A. Rev. Nutr. 19, 247277. Palumbi, S., Martin, A., Romano, S., McMillan, W. O., Stice, L. & Grabowski, G. 1991 The simple fools guide to PCR, version 2. Honolulu: Department of Zoology and Kewalo Marine Laboratory, University of Hawaii. Pa tzold, R. 1994 Die Lerchen der Welt. Magdeburg, Germany Die Neue Brehm-Bu cherei, Westarp Wissenschaften. Price, T. 1991 Morphology and ecology of breeding warblers along an altitudinal gradient in Kashmir, India. J. Anim. Ecol. 60, 643664. Purvis, A. & Garland Jr, T. 1993 Polytomies in comparative analyses of continuous characters. Syst. Biol. 42, 569575. Reeve, J. & Abouheif, E. 1999 Phylogenetic independence. Version 1.1. [1.1]. Department of Ecology and Evolution, SUNY at Stony Brook. Ricklefs, R. E., Konarzewski, M. & Daan, S. 1996 The relationship between basal metabolic rate and daily energy expenditure in birds and mammals. Am. Nat. 147, 1047 1071. Saiki, R. K., Gelfand, D. H., Stoffel, S. J., Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B. & Erlich, H. A. 1988 Enzyme directed enzymatic amplification of DNA with a thermostable polymerase. Science 239, 487491. Scholander, P. F., Hock, R., Walters, V. & Irving, L. 1950 Adaptation to cold in arctic and tropical mammals and birds in relation to body temperature, insulation and basal metabolic rate. Biol. Bull. 99, 259271. Shishkin, V. S. 1980 The energetic estimation of trophic relations of the larks (Alaudidae) in the Caspian semi-desert. Zool. Zh. 59, 12041216. Swofford, D. L. 1998 Phylogenetic analysis using parsimony, version 4. Sunderland, MA: Sinauer. Thompson, J. D., Gibson, T. J., Plewniak, T. F., Jeanmougin, F. & Higgins, D. G. 1997 The Clustal X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 24, 48764882. Thornthwaite, C. W. 1948 An approach toward a rational classification of climate. Geogr. Rev. 38, 5594. Tieleman, B. I. & Williams, J. B. 2000 The adjustment of avian metabolic rates and water fluxes to desert environments. Physiol. Biochem. Zool. 73, 461479.

214 B. I. Tieleman and others Metabolism and evaporative water loss of larks
Tieleman, B. I., Williams, J. B. & Buschur, M. E. 2002 Physiological adjustments to arid and mesic environments in larks (Alaudidae). Physiol. Biochem. Zool. 75, 305313. Tieleman, B. I., Williams, J. B., Buschur, M. E. & Brown, C. R. 2003 Phenotypic variation of larks along an aridity gradient: are desert birds more flexible? Ecology (In the press). Trevelyan, R., Harvey, P. H. & Pagel, M. D. 1990 Metabolic rates and life histories in birds. Funct. Ecol. 4, 135142. Trost, C. H. 1972 Adaptations of horned larks (Eremophila alpestris ) to hot environments. Auk 89, 506527. Walter, H. & Lieth, H. 1967 KlimadiagrammWeltatlas . Jena: VEB Gustav Fischer. Weathers, W. W. 1979 Climatic adaptation in avian standard metabolic rate. Oecologia 42, 8189. Westoby, M., Leishman, M. R. & Lord, J. M. 1995 On misinterpreting the phylogenetic correction. J. Ecol. 83, 531534. Williams, J. B. 1996 A phylogenetic perspective of evaporative water loss in birds. Auk 113, 457472. Williams, J. B. 2001 Energy expenditure and water flux of freeliving dune larks in the Namib desert: a test of the re-allocation hypothesis. Funct. Ecol. 15, 175185. Williams, J. B. & Tieleman, B. I. 2000 Flexibility in basal metabolic rate and evaporative water loss among hoopoe larks exposed to different environmental temperatures. J. Exp. Biol. 203, 31533159. Williams, J. B. & Tieleman, B. I. 2001 Physiological ecology and behavior of desert birds. In Current ornithology, vol. 16 (ed. V. Nolan & C. F. Thompson), pp. 299353. New York: Plenum Press. Willoughby, E. J. 1968 Water economy of the Starks lark and grey-backed finch-lark from the Namib desert of South West Africa. Comp. Biochem. Physiol. 27, 723745.

As this paper exceeds the maximum length normally permitted, the authors have agreed to contribute to production costs.

Proc. R. Soc. Lond. B (2003)