Geographic Variation in Avian Incubation Periods and Parental Influences on Embryonic Temperature

ORIGINAL ARTICLE

doi:10.1111/j.1558-5646.2007.00204.x

GEOGRAPHIC VARIATION IN AVIAN INCUBATION PERIODS AND PARENTAL INFLUENCES ON EMBRYONIC TEMPERATURE

Thomas E. Martin,1,2 Sonya K. Auer,1,3,4 Ronald D. Bassar,1,3,5 Alina M. Niklison,1,6 and Penn Lloyd1,7 1United States Geological Survey Montana Cooperative Wildlife Research Unit, University of Montana, Missoula, Montana 59812 2E-mail: [email protected] 4E-mail: [email protected] 5E-mail: [email protected] 6E-mail: [email protected] 7Percy FitzPatrick Institute of African Ornithology, DST/NRF Centre of Excellence, University of Cape Town, Rondebosch 7701, South Africa E-mail: [email protected]

Received March 1, 2007 Accepted June 13, 2007

Theory predicts shorter embryonic periods in species with greater embryo mortality risk and smaller body size. Field studies of 80 passerine species on three continents yielded data that largely conflicted with theory; incubation (embryonic) periods were longer rather than shorter in smaller species, and egg (embryo) mortality risk explained some variation within regions, but did not explain larger differences in incubation periods among geographic regions. Incubation behavior of parents seems to explain these discrepancies. Bird embryos are effectively ectothermic and depend on warmth provided by parents sitting on the eggs to attain proper temperatures for development. Parents of smaller species, plus tropical and southern hemisphere species, commonly exhibited lower nest attentiveness (percent of time spent on the nest incubating) than larger and northern hemisphere species. Lower nest attentiveness produced cooler minimum and average embryonic temperatures that were correlated with longer incubation periods independent of nest predation risk or body size. We experimentally tested this correlation by swapping eggs of species with cool incubation temperatures with eggs of species with warm incubation temperatures and similar egg mass. Incu- bation periods changed (shortened or lengthened) as expected and verified the importance of egg temperature on development rate. Slower development resulting from cooler temperatures may simply be a cost imposed on embryos by parents and may not enhance offspring quality. At the same time, incubation periods of transferred eggs did not match host species and reflect intrinsic differences among species that may result from nest predation and other selection pressures. Thus, geographic variation in embryonic development may reflect more complex interactions than previously recognized.

KEY WORDS: Development rates, developmental trade-offs, egg temperature, life history, nest attentiveness, nest predation, parental care, passerines.

3Present Address: Department of Biology, University of California, Riverside, CA 92521

C 2007 The Author(s). Journal compilation C 2007 The Society for the Study of Evolution. 1 Evolution MARTIN ET AL.

The rate of embryonic development is of great interest because it and Weimerskirch 2001). Incubation period is positively related can have critical fitness consequences, and interspecific variation to adult survival probability (Ricklefs 1993; Martin 2002), which is thought to be explained by three factors. Allometric constraints may reflect enhanced mechanisms for survival afforded by slower cause longer developmental periods in larger species (Rahn and development, although direct tests of such functional relationships Ar 1974; Case 1978; Arendt 1997). Shorter development peri- are lacking. ods, however, can be favored by greater predation risk (Case An alternative possibility is that longer-lived species, as com- 1978; Bosque and Bosque 1995; Martin 1995, 2002; Arendt 1997; mon in the tropics and southern hemisphere, invest less effort Remeˇs and Martin 2002). At the same time, faster development in incubation that reduces egg temperatures and slows develop- for a given mass can trade-off with intrinsic features (e.g., lo- ment (Martin 2002). Long-lived species should invest less effort comotor performance, immune function) of offspring that affect in reproduction than shorter-lived species to reduce adult mor- fitness of later stages, and may thereby favor slower development tality risk, even at a cost to offspring (Williams 1966; Charnov (Case 1978; Arendt 1997, 2003; Billerbeck et al. 2001; Brommer and Schaffer 1973; Law 1979; Michod 1979; Barbraud and 2003; Shine and Olsson 2003). Yet, the ability of allometry, mor- Weimerskirch 2001; Ghalambor and Martin 2001). Incubation is tality risk, and intrinsic trade-offs to explain geographic variation energetically expensive and can influence future reproduction and in development rates among species is relatively untested. survival (Bryan and Bryant 1999; Reid et al. 2000; Visser and Les- As typical of other taxa, incubation periods of birds gener- sels 2001). Thus, tropical and southern hemisphere bird species ally are longer in larger species across taxonomic orders (Rahn and may invest less time and energy in incubation and thereby slow de- Ar 1974; Ricklefs and Starck 1998). The longer time needed for velopment by causing cooler embryo temperatures (Martin 2002). proliferation of more and larger cells of larger species is a reason- Embryo temperature may be important because bird embryos able and well-documented expectation (Case 1978; Arendt 1997; are effectively ectothermic and depend on the presence of incu- Ricklefs and Starck 1998). Yet, incubation periods appeared not bating parents for warmth (White and Kinney 1974). Embryonic to be constrained by body or egg size within a sample of passer- temperature variation is characteristic of many true ectotherms be- ines from two latitudes (Martin 2002). The potential indepen- cause of variation in where parents place eggs in the environment. dence of incubation periods of passerines from body and egg size A critical consequence of such variation is that development takes (Martin 2002) suggest that their embryonic development rates longer for eggs placed in cooler incubation environments among within and among geographic regions are subject to influences conspecific ectotherms, and also yields smaller and/or lower qual- beyond simple allometric constraints. ity hatchlings (Webb and Cooper-Preston 1989; Shine et al. 1997; Increases in time-dependent mortality such as nest predation Qualls and Andrews 1999; Shine and Olsson 2003; Hare et al. favor shorter developmental periods (Case 1978). Egg predation 2004). If slow development in birds also results from cool embry- risk, however, is unlikely to explain broad geographic variation in onic temperatures, then offspring quality may be compromised passerine incubation periods (Martin 1996, 2002). Incubation pe- (Olson et al. 2006), rather than enhanced. The ability of parental riods are commonly longer for species in the tropics and southern behavior and egg temperatures to explain interspecific variation hemisphere compared with related species in the north temperate in incubation periods of birds has been debated (Boersma 1982; region (Martin 2002, Chalfoun and Martin 2007), although nest Ricklefs 1984; Tieleman et al. 2004). Yet, broad and stringent predation rates are either similar or greater in these southern loca- tests of temperature influences on interspecific and geographic tions (Skutch 1949; Maclean 1974; Robinson 1990; Major 1991; variation in incubation periods are lacking. Here, we examine Martin et al. 2000, 2006). Thus, the long incubation periods of these alternative hypotheses based on detailed study of 80 passer- many tropical and southern hemisphere species oppose expecta- ine species on three continents (North America, South America, tions from nest predation risk (Martin 1996, 2002), and emphasize and Africa). the importance of understanding alternative causes of geographi- cal variation in embryonic development rates. Indeed we can ask Methods more generally why a species should ever develop slowly given STUDY SYSTEM the accumulating risk of mortality? We studied 80 passerine species (Appendix) at four sites on three One possibility is that slower development may allow greater continents. Study sites were high-elevation (2300 m elevation) development of intrinsic features that enhance performance and mixed forest in Arizona (34◦N latitude), Yungas forest (1000 m) in longevity (Ricklefs 1993; Arendt 1997, 2001, 2003; Billerbeck northwestern Argentina (26◦S), cloud forest (1400–2000 m) in the et al. 2001; Brommer 2003; Shine and Olsson 2003). Increased northern Andes of Venezuela (9◦N), and coastal dwarf shrubland longevity may be more important than increased risk of offspring near Cape Town, South Africa (34◦S) (see Martin 1998; Mar- mortality in long-lived species that depend on iteroparity across tin et al. 2000, 2006; Nalwanga et al. 2004 for descriptions of years to enhance fitness (Williams 1966; Law 1979; Barbraud study sites).

2 EVOLUTION 2007 GEOGRAPHIC VARIATION IN INCUBATION PERIODS

NEST PREDATION, INCUBATION PERIODS, AND failing batteries or poorly sealed holes. We sought to sample six PARENTAL CARE SAMPLING nests per species, but samples varied from two to nine, with 12 and Study sites were searched for nests during entire breeding seasons 15 nests for two species, and an average of 5.6 nests per species. in 1988–2005 in Arizona, 1997–1999 in Argentina, 2000–2004 in South Africa, and 2002–2006 in Venezuela. For species examined EGG TRANSFER EXPERIMENT here, large numbers of nests were monitored for nest predation and To test the ability of egg temperature to influence developmental duration of incubation periods in Arizona (10,157 nests), South rates, we performed an egg transfer experiment in South Africa Africa (4385 nests), Argentina (1716 nests), and Venezuela (2354 during the 2003 and 2004 field seasons. We performed this exper- nests) following long-term protocols (Martin 1998, 2002; Martin iment in South Africa because it includes reasonable variation in et al. 2000). Incubation period was quantified as the difference in incubation periods (Fig. 1) and we found large numbers of nests days between last egg laid and last egg hatched (Nice 1954; Briskie daily, making the experiment feasible (the experiment required and Sealy 1990), but could not be determined at most nests due to synchronous timing of nests during egg laying that required large predation or other events and was not carefully monitored in Ari- numbers of nests on a daily basis). Species were paired based on zona until starting in 1993. Nests were checked every two to four similar egg mass and contrasting egg temperatures and incubation days to determine status and predation events, but were checked periods (species identities of successful egg swaps are identified daily or twice daily near stage-changing events such as hatching in the figure legend); species with cool egg temperatures and long and fledging to accurately measure period durations. Nest preda- incubation periods were paired with species with warm egg tem- tion was assumed when nest contents disappeared during incuba- peratures and short incubation periods of similar egg mass. During tion and was calculated as daily predation rates (Mayfield 1975; egg laying, we swapped one egg, to keep clutch mass constant, Hensler and Nichols 1981) during the incubation period. between nests of these paired species. We chose eggs at random Nest attentiveness was measured as the percent of time that in the laying order for species with low egg temperature (i.e., cool parents were on the nest. Nests were videotaped during incubation incubation nests), whereas we swapped them with the last eggs for the first 6–8 h of the day, beginning within 0.5 h of sunrise in the laying sequence for the paired species with high egg tem- from 1993 to 2005 in Arizona and in all study years in Argentina, perature (i.e., warm incubation nests). Selecting eggs at random Venezuela,and South Africa. This protocol standardized both time from the cool incubation nests randomized potential confounding of day and sampling duration (Martin et al. 2000; Martin 2002). differences in egg composition within a clutch, while placing the In addition, attentiveness was systematically sampled early (two experimental egg into the warm incubation nest on the day the to three days after last egg laid) and late (two to three days prior ultimate egg was laid in warm incubation nests ensured full incu- to average hatching date) in the incubation period in South Africa bation and accurate calculation of incubation period. Some species and Venezuela and in the last six years in Arizona. It was also in South Africa sometimes began incubation before the last egg opportunistically sampled in the middle to incorporate any stage was laid (T. E. Martin, pers. obs.), which may have caused us changes in overall estimates. In Argentina, and earlier years in to underestimate the length of the incubation period for species Arizona, nest attentiveness was randomly sampled across the incu- with warm incubation. Eggs of warm species were swapped into bation period. Preliminary analyses indicate that nest attentiveness cool nests before the last egg was laid in the cool nests and some does not change over the incubation period in Arizona, Argentina, of these may have experienced incubation before the last egg was and South Africa and so timing of sampling is not important. At- laid. Consequently, our measure of incubation based on time start- tentiveness increased across the incubation period among many ing with the last egg laid (see above) then would not include this species in Venezuela, but our sampling incorporated these effects earlier incubation time. This yielded a highly conservative test by averaging behavior early and late in incubation. of whether placement in cooler nests yielded longer incubation. Eggs were marked as they were laid and the experimental eggs EGG TEMPERATURE MEASUREMENTS were swapped within 1 h of being laid to ensure that little or no Egg temperatures were measured by inserting thermisters on the incubation had occurred in the natal nest. This intense experiment first or second day of incubation into the center of eggs through was made even more difficult by the extremely high nest preda- a small hole sealed with glue (Weathers and Sullivan 1989). The tion rates in South Africa (Fig. 1B); the 26 nests with transferred wire was threaded through the nest and connected to a HOBO eggs that escaped predation and allowed us to measure incubation Stowaway XTI datalogger (Onset; Bourne, MA) that recorded period represented just under 20% of the nests in which transfers temperatures every 12–24 sec for five to seven days per nest. We were made. used data from days 2–6 to standardize sampling and minimize Incubation period for experimental nests was calculated us- possible disturbance effects on the first day. We excluded nests ing the same methods as for unmanipulated nests. We calculated in which temperatures systematically changed across days from the difference in incubation periods between the transferred egg

EVOLUTION 2007 3 MARTIN ET AL.

Figure 1. Relationships of mean incubation period, nest predation rate, adult mass, and nest attentiveness (% time spent on the nest) for 80 passerine species studied across four sites on three continents. Duration of the incubation period as a function of: (A) natural logarithm of adult mass; (B) daily nest predation rate during incubation; and (C) nest attentiveness. (D) Nest attentiveness relative to the natural logarithm of adult mass. Lines are based on the pooled slope and intercept from ANCOVA accounting for within and across site variation. relative to the original natal nest, or if the original nest was depre- Results dated, we used the mean for other nests of the natal species that ALLOMETRY AND NEST PREDATION started incubation on the same date. Upon hatching, we returned We first tested the potential influence of the two factors classically the hatchling to its original nest or to another nest of the same thought to influence embryonic development rates: allometry and age of the natal species if the original nest was depredated. All nest predation. Duration of incubation periods varied from 10 to transferred eggs hatched if they survived predation; that is, we did 27 days across species. Incubation periods are expected to increase not impact hatching success with transfers. with mass by about six days over the ten-fold range, from 6 to 77 g, of the species that we studied (Rahn and Ar 1974; Arendt DATA ANALYSIS 1997; Ricklefs and Starck 1998). Incubation periods instead de- We used general linear models to examine variation in log- creased with increased log-transformed body mass (Fig. 1A, Table transformed incubation period, with site as a random factor, and 1A), opposite to allometric expectations. Incubation periods also nest predation, log-transformed body mass, nest attentiveness, decreased with increased nest predation rates (Fig. 1B, Table 1A), and egg temperature as possible covariates. We included inter- as predicted by theory (Case 1978; Bosque and Bosque 1995; actions, but subsequently removed interactions and do not report Arendt 1997; Martin 2002). The significant interaction between them when they were not significant. Where interactions with mass and site (Table 1A) reflected negative slopes in all sites ex- site occurred, they were orderly and, therefore, allowed mean- cept South Africa, where the slope did not differ from 0 (slope = ingful interpretation of main and interaction effects (Ott 1993). 0.46 ± (SE) 0.47, P = 0.35). Given the negative relationships, we We controlled for possible phylogenetic effects using indepen- excluded mass from the remaining analyses of incubation period dent contrasts (IC; Felsenstein 1985) from comparative analysis because it inappropriately corrected for possible allometric effects by independent contrasts (CAIC; Purvis and Rambaut 1995), and in these data (Fig. 1A). a regression approach to analysis of covariance (ANCOVA; Mar- When mass is dropped and nest predation is considered alone, tin 1995, 2002). The phylogenetic hypothesis was based on Sib- it continued to have a clear influence on duration of incubation ley and Ahlquist (1990) but modified by subsequent phylogenetic periods within sites, again as expected by theory. However, incu- studies as summarized previously (Martin et al. 2006). bation periods differed among sites for the same nest predation

4 EVOLUTION 2007 GEOGRAPHIC VARIATION IN INCUBATION PERIODS

Table 1. Results of analyses of covariance to explain interspeciﬁc variation in incubation period and nest attentiveness for Raw data and Independent contrasts. Error df are for Raw Data, and are one less for independent contrasts.

Variables Raw data Independent contrasts

df F-value P-value df F-value P-value (A) Incubation period explained by allometry and nest predation (error df = 66) Ln (Mass) 1 4.9 0.031 1 3.0 0.087 Nest predation rate 1 5.5 0.021 1 9.1 0.004 Site 3 7.8 <0.001 3 5.7 0.002 Site × Mass 3 3.3 0.024 3 3.5 0.021 Site × predation 3 1.3 0.28 3 1.0 0.40 (B) Incubation period explained by nest predation (error df = 70) Nest predation rate 1 13.6 <0.001 1 7.3 0.009 Site 3 7.4 <0.001 3 4.1 0.010 Site × predation 3 1.5 0.22 3 0.9 0.45 (C) Incubation period explained by nest attentiveness and nest predation (error df = 69) Attentiveness 1 18.6 <0.001 1 68.3 <0.001 Nest predation rate 1 8.1 0.006 1 4.8 0.032 Site 3 4.9 0.004 3 5.7 0.002 Site × nest attentiveness 3 4.2 0.009 3 5.2 0.003 (D) Nest attentiveness explained by nest predation, body mass, and when males and females shared incubation duties (error df = 71) Nest predation rate 1 0.2 0.6 1 0.0 0.9 Mass 1 5.8 0.018 1 5.1 0.027 Shared incubation 1 9.9 0.002 1 5.4 0.023 Site 3 10.2 <0.001 3 7.5 <0.001 rate (Table 1B, Fig. 1B). In particular, duration of the incuba- sed nest attentiveness (Fig. 1D), which was associated with shorter tion period in Arizona was less (LSD, P = 0.004) than in South incubation periods (Fig. 1C) and at least partly explained the neg- Africa, which did not differ (P = 0.9) from Argentina, which was ative relationship between incubation period and mass (Fig. 1A). less (P = 0.039) than in Venezuela, when controlled for nest predation. Thus, broad geographic differences in incubation periods PARENTAL INFLUENCES ON EMBRYONIC of passerines did not follow classical theory; geographic differ- TEMPERATURES ences were largely not explained by nest mortality or allometry. Parental time on the nest is potentially important to duration of in- These results indicate that other factors must be critical drivers cubation periods through its effects on temperatures experienced of geographic variation in embryonic development rates, and we by embryos. Average egg temperatures measured over 24 h for explore the possible role of parental behavior in explaining re- the 41 species that we studied increased with nest attentiveness maining variation. of parents (Fig. 2A). Moreover, egg temperatures varied most extensively during the day (Fig. 2B) when nest attentiveness var- NEST PREDATION AND PARENTAL BEHAVIOR ied (Fig. 3); day-time egg temperatures were strongly correlated Parental time on the nest (nest attentiveness) explained a sub- with attentiveness and differed among sites (attentiveness:raw— stantial amount of the variation in incubation periods, even after F1,36 = 28.1, P < 0.001; IC—F1,36 = 27.4, P < 0.001; site:raw— accounting for nest predation, although some of the variation be- F3,36 = 3.3, P = 0.033; IC—F3,36 = 2.9, P = 0.048). Egg tem- tween sites remained unexplained (Table 1C; Fig. 1B, C). The peratures varied much less during the night when attentiveness is interaction between nest attentiveness and site was orderly with constant (Fig. 2B, 3) and largely did not differ among sites (raw— all sites showing negative slopes, but the slope in Arizona was not F3,37 = 2.6, P = 0.067; IC—F3,37 = 2.6, P = 0.068) (Fig. 2B). significant (slope =−0.002 ± 0.002, P = 0.43). These results demonstrate that geographic variation in egg tem- Nest attentiveness was not related to nest predation rates peratures largely result from differences in day-time nest atten- among species, but increased with body mass and when males tiveness and not from geographic differences in ability to keep and females shared incubation (Table 1D, Fig. 1D). Thus, for the eggs warm. However, the extensive variation in day-time egg tem- passerine species that we studied, many larger species had increa- peratures was positively correlated with the limited variation in

EVOLUTION 2007 5 MARTIN ET AL.

night-time egg temperatures within sites (night-time egg tempera-

ture:raw—F1,36 = 14.7, P < 0.001; IC—F1,36 = 15.7, P < 0.001;

site:raw—F3,36 = 4.8, P = 0.007; IC—F1,36 = 5.1, P = 0.005), suggesting that species within sites differ in keeping eggs warm during both day and night. Of course, day-time nest attentiveness did not completely explain egg temperatures; some species that differed in nest attentiveness had similar egg temperatures, and species with similar attentiveness differed to a limited extent in egg temperature (see Fig. 2A). Some species share incubation, where males incubate part of the time, and males of some species do not maintain temperatures as warm as females (Kleindorfer et al. 1995; Reid et al. 2002; Bartlett et al. 2005). In addition, species differ in length of on and off bouts for the same attentiveness and such variation can influence average egg temperature (Hainsworth et al. 1998; Conway and Martin 2000). Nonetheless, day-time attentiveness had a strong influence on average temperatures experienced by embryos (Fig. 2A). The lower day-time attentiveness of species with long incubation periods (Fig. 1C) reflects long off-bouts that caused egg temperatures to fall below 26o C (Fig. 3), where embryo development has been thought to be dramatically slow based on minimal evidence (White and Kinney 1974; Webb 1987; Haftorn 1988). Yet,the temperature provides a convenient benchmark for comparisons. Related species with short incubation periods had off-bouts that were short enough to largely keep egg temperatures above such levels, even in north temperate species where ambient temperatures are much colder (Fig. 3). Indeed, ambient temperature commonly fell below 0oC at night in the north temperate Ari- zona site, while it only fell as low as 15oC in the Venezuela site (T. E. Martin, pers. obs.). Despite the warmer ambient temperatures in Venezuela, eggs were allowed to drop to much colder temperatures in Venezuela than in Arizona (Fig. 3A). Similarly, South African species with shorter incubation periods kept egg temperature consistently warmer than related species with longer incubation periods when measured on the same day and ambient temperature (Fig. 3B). Thus, incubation periods, attentiveness, and egg temperatures differ between closely related species of

Figure 2. Relationships of mean egg temperatures over 24 h with similar mass (Fig. 3), emphasizing the independence of embry- mean nest attentiveness (% time spent on the nest) and incuba- onic development and temperature from size and phylogeny. tion periods of 41 passerine species across four sites on three continents. (A) Average 24-h egg temperature increased with nest at- INCUBATION PERIODS AND EMBRYONIC tentiveness (attentiveness: raw—F 1,36 = 23.6, P < 0.001; IC—F 1,36 TEMPERATURES = 19.4, P < 0.001; site: raw—F 3,36 = 2.9, P = 0.050; IC—F 3,36 = 2.2, Incubation periods were strongly predicted by average 24-h egg P = 0.11); (B) average egg temperatures during night (2200–0400 h) temperatures, with residual variation explained by nest preda- and day (0600–1800 h) on axes of the same scale to allow compar- tion (egg temperature:raw—F , = 124.8, P < 0.001; IC— isons of differences in extent of variation and including the iden- 1 35 F , = P < nest predation F , = tity (1:1) line for comparison; (C) average 24-h egg temperature 1 35 123.0, 0.001; :raw-– 1 35 4.5, = = = and duration of incubation periods. Gray ellipses highlight species P 0.042; IC—F1,35 4.1, P 0.050; Fig. 2C). Site was no with outlier incubation periods for their temperatures: Premno- longer significant (site:raw—F3,35 = 2.6, P = 0.070; IC—F3,35 = plex brunnescens, Mionectes olivaceus, Henicorhina leucophrys. 2.3, P = 0.091), indicating that egg temperatures (warmer

6 EVOLUTION 2007 GEOGRAPHIC VARIATION IN INCUBATION PERIODS

Figure 3. Representative examples of daily temperature ﬂuctuations in comparisons of phylogenetically related passerine species with fast (short incubation periods) versus slow (long incubation periods) development during day 2, 3, or 4 of incubation in: (A) Arizona versus Venezuela, and (B) within South Africa measured on the same day to hold ambient weather constant. The gray line in each cell represents a temperature benchmark in which development has been thought to be dramatically slowed, whereas the two dashed lines at the top of each cell represents the optimum temperature zone for development (White and Kinney 1974; Webb 1987; Haftorn 1988). temperatures were associated with shorter incubation periods— bation periods were shortened (t18 =−9.4, P < 0.001) up to 3.3 Fig. 2C) and nest predation (higher nest predation was associated days in species with normal incubation periods of 14–17 days with shorter incubation periods—Fig. 1B) explained the variation when eggs of species with cool egg temperatures were trans- among, as well as within, sites. ferred into nests of species with warmer incubation temperatures

We experimentally confirmed that parental effects on egg (Fig. 4A). Similarly, incubation periods were lengthened (t6 = 3.2, temperatures influence differences in incubation periods among P = 0.018) when eggs of species with warm egg temperatures were species. We transferred eggs among nests of species that differed transferred into nests of species with cooler incubation tempera- in their nest attentiveness and resulting egg temperatures. Incu- tures (Fig. 4A) even though length of incubation periods may have

EVOLUTION 2007 7 MARTIN ET AL.

Figure 4. (A) Change in incubation period within species relative to the natal nest when one egg of a species with cool egg temperatures was exchanged with one egg of a species with warm egg temperatures and similar egg mass. Black bars reflect cases in which eggs of species with cool incubation temperatures were transferred to nests of species with warmer incubation. Gray bars reflect cases in which eggs of species with warm incubation were transferred to nests of species with cooler incubation. Values below 0 indicate shortened incubation period, and values above 0 indicate lengthened incubation periods, in days. (B) Differences in incubation periods between species for eggs incubated in the same nest based on comparing transferred eggs versus eggs of the host species. Values above 0 indicate longer incubation periods for transferred eggs than for host eggs, and values below 0 indicate shorter incubation periods than hosts, in days. Black bars reflect eggs of species with long incubation periods, which were transferred to nests of species with shorter incubation periods. Gray bars reflect eggs of species with short incubation periods, which were transferred to nests of species with longer incubation periods. Bars are lined up between cells to show the same egg in both comparisons. The species identities (average 24-h egg temperature for the species) for eggs from cool (long incubation) nests transferred to warm (short incubation) nests (black bars) included (N = number of nests with a transferred egg, nest letter identification from x-axis): Prinia maculosa (34.6◦)inSerinus flaviventris (35.6◦)(N = 1, nest a); Prinia in Zosterops pallidus (36.0◦)(N = 4, nests b, f, g, m); Apalis thoracica (34.2◦)inSerinus (N = 6, nests c, d, i, j, n, p); Cisticola subruficapillus (34.9◦)inSerinus (N = 3, nests e, h, s); Apalis in Zosterops (N = 2, nests k, o); Sphenoeacus afer (33.1◦)inPycnonotus capensis (36.1◦)(N = 2, nests l, r); and Cossypha caffra (32.4◦)inPycnonotus (N = 1, nest q). Species identities for eggs from warm (short incubation) nests transferred to cool (long incubation) nests (gray bars) were: Zosterops in Prinia (N = 2, nests t, z); Serinus in Apalis (N = 2, nests u, x); Pycnonotus in Cossypha (N = 1, nest v); Zosterops in Cisticola (N = 1, nest w); Serinus in Cisticola (N = 1, nest y). been underestimated (see Methods) and sample size was smaller in the host nest by as much as 3.5 days (Fig. 4B). Similarly, incu- due to more experimental nests being lost to predation. bation periods for eggs of species with short incubation periods , Although transferred eggs changed duration of incubation which were transferred into nests of species with long incubation periods based on egg temperature (Fig. 4A), they also demon- periods were shorter (t6 =−5.0, P = 0.003) than eggs of the host strated clear intrinsic differences between species. Eggs of species species in the host nest by as much as 3.0 days (Fig. 4B). Note that with long incubation periods, which were transferred into nests differences are intrinsic differences between species in embryonic of species with shorter incubation periods still had longer (t18 = development rate per se because egg mass and egg temperature 6.5, P < 0.001) incubation periods than eggs of the host species were controlled in transfers.

8 EVOLUTION 2007 GEOGRAPHIC VARIATION IN INCUBATION PERIODS

DISCUSSION egg temperature studies of Lill (1979) replicate the causal evidence from our swaps (Fig. 4) that long incubation periods of Incubation periods are thought to be phylogenetically conserved tropical and southern species can be caused by low nest atten- and strongly determined by intrinsic constraints imposed by size tiveness and cool egg temperatures. Third, nest attentiveness is and physiological trade-offs (Lack 1968; Rahn and Ar 1974; broadly and systematically lower in tropical and southern hemi- Ricklefs 1984; Ricklefs and Starck 1998). Slower development in sphere passerines compared with north temperate species, even larger organisms is a reasonable and well-documented expectation among phylogenetic relatives (Martin 2002; Chalfoun and Mar- (Rahn and Ar 1974; Case 1978; Ricklefs and Starck 1998). How- tin 2007). Given the association of nest attentiveness with egg ever, incubation periods decreased, rather than increased, with temperature (Fig. 2, see above), then this geographic pattern of size within our sets of passerines (Fig. 1A). The shorter incu- parental behavior suggests a possible explanation for at least part bation periods were associated with higher attentiveness in larger of the geographic variation in incubation periods. passerine species (Fig. 1D), which may indicate that resource lim- The results raise questions about the extent to which slow itation on incubation attentiveness is reduced in larger birds that development in tropical species represents a trade-off to enhance can rely more on internal resources (e.g., Aldrich and Raveling offspring quality. Slower development can reflect trade-offs to en- 1983; Cresswell et al. 2004). Size constraints on incubation pe- hance offspring quality at a constant temperature (Ricklefs 1993; riod length may be measured most accurately among species that Arendt 1997, 2001, 2003; Billerbeck et al. 2001; Brommer 2003; are ecologically similar (i.e., similar in nest attentiveness, devel- Shine and Olsson 2003), but slower development resulting from opmental mode, and feeding strategies). Nonetheless, our results cooler temperatures may reduce offspring quality. Compromised demonstrate that, within a phylogenetically restricted group, in- quality has been observed within ectotherm species, where cooler trinsic constraints imposed by size can be over-ridden by other incubation temperatures yield slower development and poorer causal factors in determining length of incubation periods. quality offspring (Webb and Cooper-Preston 1989; Shine et al. The strong correlation between nest attentiveness, egg tem- 1997; Qualls and Andrews 1999; Shine and Olsson 2003). The perature, and incubation period across a broad diversity of species exposure of avian embryos to external temperatures makes them and geographic areas (Figs. 1–3), confirmed by our experimen- more similar to ectotherms than viviparous endotherms at this de- tal transfers (Fig. 4A), provides strong evidence for a causal role velopmental stage, and incubation temperature clearly influences of parental care and egg temperature in influencing interspecific bird development (i.e., Fig. 2C, 4), as observed within ectotherm variation in length of incubation periods, as suggested long ago species. Cooler embryonic temperatures caused by low nest at- by Boersma (1982). Indeed, our experiment shifted incubation tentiveness (see Fig. 3) may not only cause slower embryonic periods to values characteristic of other species and suggests a development in birds, but may also compromise hatchling strong parental influence on incubation period that shows system- quality because greater energy may be needed by embryos to de- atic variation among geographic regions. This idea is supported by velop at suboptimal temperatures (Booth 1987). Indeed, a recent a variety of evidence. First, in one of the only careful studies of her- experimental study demonstrated that periodic cooling of embryos itability of incubation periods in passerines, Ricklefs and Smeraski caused slower development and produced smaller hatchlings with (1983) conducted egg swaps among nests of Sturnus vulgaris. reduced yolk reserves compared with warmer embryos (Olson They found that heritability was quite low and concluded that et al. 2006, also see Hepp et al. 2006). Moreover, at least one the incubation period was determined primarily by the incubating measure of immune function was broadly reduced among passer- behavior of the adults and thermal properties of nests. Second, ine species with longer incubation periods in Arizona (Palacios reduced nest attentiveness has been found to yield longer incu- and Martin 2006). Long incubation periods of many tropical and bation periods within species (White and Kinney 1974; Boersma southern hemisphere species (Fig. 1), therefore, may reflect a cost and Wheelwright 1979; Lyon and Montgomerie 1985; Lifjeld et imposed on offspring by parents through egg neglect and resulting al. 1987; Reid et al. 2002), as we also found (Fig. 4A). The ability cool egg temperatures (Fig. 2A, 3) due to fitness trade-offs that of warmer incubation temperatures to yield shorter incubation pe- parents face (Martin 2002). riods has been untested with exception of a single anecdotal test: Such temperature constraints on offspring quality do not Ward (1940) transferred a newly laid egg of Menura superba to a negate the possibility that trade-offs in developmental rate and domestic hen that was bred for constant incubation attentiveness offspring quality also exist. Embryonic temperatures did not ex- and found the normal 50-day incubation period was reduced to 28 plain all variation; indeed, for a given embryonic temperature, days. Menura has an unusually long incubation period associated incubation period still differed among species by several days with 7 h off-bouts that reflect low nest attentiveness (Lill 1979), (Fig. 2C). Moreover, the egg transfer experiment clearly showed as is common for many tropical and southern hemisphere species that interspecific differences in incubation periods remain even (Fig. 1–3, Martin 2002). Thus, the egg swap by Ward (1940) and when two species are incubated at the same temperature (Fig. 4B).

EVOLUTION 2007 9 MARTIN ET AL.

The statistically significant effects of nest predation (see Results) ———. 2003. Reduced burst speed is a cost of rapid growth in anuran tadpoles: suggest an additive role of mortality selection; nest predation did problems of autocorrelation and inferences about growth rates. Funct. not predict attentiveness, but nest predation explained residual Ecol. 17:328–334. Arendt, J. D., S. Wilson, and E. Stark. 2001. Scale strength as a cost of rapid variation in incubation periods after effects of attentiveness or egg growth in sunfish. Oikos 93:95–100. temperature were taken into account (see Results). Nest predation Barbraud, C., and H. Weimerskirch. 2001. Emperor penguins and climate may exert selection on internal mechanisms that then influence change. Nature 411:183–186. development rate (Case 1978; Arendt 1997; Schwabl et al. 2007). Bartlett, T. L., D. W. Mock, and P. L. Schwagmeyer. 2005. Division of labor: incubation and biparental care in house sparrows (Passer domesticus). Variation in development rates caused by such internal mecha- Auk 122:835–842. nisms, rather than extrinsic temperature might then trade-off with Billerbeck, J. M., T. E. Lankford Jr., and D. O. Conover. 2001. Evolution of in- offspring quality, as clearly documented in other taxa (Arendt trinsic growth and energy acquisition rates. I. Trade-offs with swimming 1997, 2001, 2003; Billerbeck et al. 2001; Brommer 2003; Shine performance in Menidia menidia. Evolution 55:1863–1872. Boersma, P. D. 1982. Why some birds take so long to hatch. Am. Nat 120:733– and Olsson 2003). 750. Even within sites, species may vary in evolutionary solu- Boersma, P. D., and N. T. Wheelwright. 1979. Egg neglect in the Procellari- tions and extent of physiological trade-offs in embryonic devel- iforme: reproductive adaptations in the fork-tailed Storm-Petrel. Condor opment. For example, the three outlier species in Venezuela that 81:157–165. had long incubation periods for their egg temperatures (Fig. 2C) Booth, D. T. 1987. Effect of temperature on development of Mallee fowl Leipoa ocellata eggs. Phys. Zool. 64:437–445. may reflect individual species with particularly high adult survival Bosque, C., and M. T. Bosque. 1995. Nest predation as a selective factor in the that develop slow to strongly enhance offspring quality. Of these evolution of developmental rates in altricial birds. Am. Nat. 145:234– three species, only one (Henichorina leucophyrs) had survival es- 260. timated, but survival of this species was estimated to be quite high: Briskie, J. V., and S. G. Sealy. 1990. Evolution of short incubation periods in the parasitic cowbirds, Molothrus spp. Auk 107:789–794. 92.0% annual survival for second year and older birds (Parker et Brommer, J. E. 2003. Immunocompetence and its costs during development: an al. 2006). Although some variation among species clearly reflects experimental study in blue tit nestlings. Proc. R. Soc. Lond. B 271:S110– intrinsic trade-offs, substantial variation in incubation periods was S113. explained by temperature that may have opposing effects on off- Bryan, S., and D. M. Bryant. 1999. Heating nest boxes reveals an energetic spring quality. Indeed, our experimental results and broad correla- constraint on incubation behaviour in great tits, Parus major. Proc. R. Soc. Lond. B 266:157–162. tions, together with evidence from the literature, demonstrate that Case, T. J. 1978. On the evolution and adaptive significance of postnatal growth parental effects on egg temperatures play an important role in in- rates in the terrestrial vertebrates. Q. Rev. Biol. 53:243–282. fluencing incubation periods of passerines throughout the world. Charnov, E. L., and W. M. Schaffer. 1973. Life history consequences of natural These results raise questions about traditional views that devel- selection: Cole’s result revisited. Am. Nat. 107:791–793. Chalfoun, A., and T. E. Martin. 2007. Latitudinal variation in avian incubation opment rates are primarily driven by allometry or other intrinsic attentiveness and a test of the food limitation hypothesis. Anim. Behav. constraints, and place new emphasis on parental care strategies. 73:579–585. Conway, C. J., and T. E. Martin. 2000. Evolution of avian incubation behavior: influence of food, climate and nest predation. Evolution 54:670–685. ACKNOWLEDGMENTS Cresswell, W., S. Holt, J. M. Reid, D. P. Whitfield, R. J. Mellanby, D. Norton, We are grateful to J. Briskie, R. Callaway, K. Dial, D. Emlen, the Tea and S. Waldron. 2004. The energetic costs of egg heating constrain incu- discussion group, and anonymous reviewers for helpful comments on the bation attendance but do not determine daily energy expenditure in the manuscript and to the many people that helped in collecting the data re- pectoral sandpiper. Behav. Ecol. 15:498–507. ported here. C. Bosque and M. DuPlessis provided invaluable help with Feldstein, J. 1985. Phylogenics and the comparative method. Am. Nat. 125:1– logistics. We are especially grateful to E. Kofoed for his help with the egg 15. transfer experiments and his help with general data collection. This work Ghalambor, C. K., and T. E. Martin. 2001. Fecundity-survival trade-offs and was supported by National Science Foundation grants (DEB-9707598, parental risk-taking in birds. Science 292:494–497. INT-9906030, DEB-9981527, DEB-0543178) for studies in Arizona, Haftorn, S. 1988. Incubating female passerines do not let the egg temperature Venezuela, Argentina, and South Africa, and the work in Arizona was fall below the ‘physiological zero temperature’ during their absences also supported by the United States Geological Survey Climate Change from the nest. Ornis Scand. 19:97–110. Research Program, and the National Research Initiative of the USDA Co- Hainsworth, F. R., T. Moonan, M. A. Voss,K. A. Sullivan, and W. W. Weathers. operative State Research, Education and Extension Service, grant number 1998. Time and heat allocations to balance conflicting demands during 2005-02817. intermittent incubation by yellow-eyed Juncos. J. Av. Biol. 29:113–120. Hare, K. M., C. G. Longson, S. Pledger, and C. H. Daugherty. 2004. Size, growth, and survival are reduced at cool incubation temperatures in LITERATURE CITED the temperate lizard Oligosoma suteri (Lacertilia: Scincidae). Copeia Aldrich, T. W., and D. G. Raveling. 1983. Effects of experience and body 2004:383–390. weight on incubation behavior of Canada geese. Auk 100:670–679. Hensler, G. L., and J. D. Nichols. 1981. The Mayfield method of estimating Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. nesting success: a model, estimators and simulation results. Wils. Bull. Q. Rev. Biol. 72:149–177. 93:42–53.

10 EVOLUTION 2007 GEOGRAPHIC VARIATION IN INCUBATION PERIODS

Hepp, G. R., R. A. Kennamer, and M. H. Johnson. 2006. Maternal effects in offspring in the lizard, Sceloporus virgatus. Biol. J. Linn. Soc. 67:353– wood ducks: incubation temperature influences incubation period and 376. neonate phenotype. Funct. Ecol. 20:307–314. Rahn, H., and A. Ar. 1974. The avian egg: incubation time and water loss. Kleindorfer, S., B. Fessl, and H. Hoi. 1995. More is not always better: male Condor 76:147–152. incubation in two Acrocephalus warblers. Behaviour 132:607–625. Reid, J. M., P. Monaghan, and G. D. Ruxton. 2000. Resource allocation be- Lack, D. 1968. Ecological adaptations for breeding in birds. Methuen, Lond. tween reproductive phases: the importance of thermal conditions in de- Law, R. 1979. Optimal life histories under age-specific predation. Am. Nat. termining the cost of incubation. Proc. R. Soc. Lond. B 267:37–41. 114:399–417. ———. 2002. Males matter: the occurrence and consequences of male incuba- Lifjeld, J. T., T. Slagsvold, and G. Stenmark. 1987. Allocation of incubation tion in starlings (Sturnus vulgaris). Behav. Ecol. Sociobiol. 51:255–261. feeding in a polygynous mating system: a study on pied flycatchers Remes,ˇ V., and T. E. Martin. 2002. Environmental influences on the evolution Ficedula hypoleuca. Anim. Behav. 35:1663–1669. of growth and developmental rates in Passerines. Evolution 56:2505– Lill, A. 1979. Nest inattentiveness and its influence on development of the 2518. young in the superb lyrebird. Condor 81:225–231. Ricklefs, R. E. 1984. Prolonged incubation in pelagic seabirds: a comment on Lyon, B. E., and R. D. Montgomerie. 1985. Incubation feeding in snow Boersma’s paper. Am. Nat. 123:710–720. buntings: female manipulation or indirect male parental care? Behav. ———. 1993. Sibling competition, hatching asynchrony, incubation period, Ecol. Sociobiol. 17:279–284. and life span in altricial birds. Curr. Orn. 11:199–275. Maclean, G. L. 1974. Arid-zone adaptations in southern African birds. Cim- Ricklefs, R. E., and C. A. Smeraski. 1983. Variation in incubation period bebasia 2:163–176. within a population of the European starling. Auk 100:926–931. Major, R. E. 1991. Breeding biology of the white-fronted chat Ethianura Ricklefs, R. E., and J. M. Starck. 1998. Embryonic growth and development. albifrons in a saltmarsh near Melbourne. Emu 91:236–249. Pp. 31–58 in J. M. Starck, and R. E. Ricklefs, eds. Avian growth and Martin, T. E. 1995. Avian life history evolution in relation to nest sites, nest development. Oxford Univ. Press, Oxford, U.K. predation and food. Ecol. Monogr. 65:101–127. Robinson, D. 1990. The nesting ecology of sympatric Scarlet Robin Petroica ———. 1996. Life history evolution in tropical and south temperate birds: multicolor and Flame Robin P. phoenicea populations in open Eucalypt What do we really know? J. Av. Biol. 27:263–272. forest. Emu 90:40–52. ———. 1998. Are microhabitat preferences of coexisting species under se- Schwabl, H., M. G. Palacios, and T. E. Martin. 2007. Selection for rapid lection and adaptive? Ecology 79:656–670. development leads to higher embryo exposure to maternal androgens ———. 2002. A new view for avian life history evolution tested on an incu- among passerine species. Am. Nat. 107:196–206. bation paradox. Proc. R. Soc. B 269:309–316. Shine, R., and M. Olsson. 2003. When to be born? Prolonged pregnancy or Martin, T. E., R. D. Bassar, S. K. Bassar, J. J. Fontaine, P.Lloyd, H. Mathewson, incubation enhances locomotor performance in neonatal lizards (Scin- A. Niklison, and A. Chalfoun. 2006. Life history and ecological corre- cidae). J. Evol. Biol. 16:823–832. lates of geographic variation in egg and clutch mass among passerine Shine, R., T. R. L. Madsen, M. J. Elphick, and P.S.Harlow. 1997. The influence species. Evolution 60:390–398. of nest temperatures and maternal brooding on hatchling phenotypes in Martin, T. E., P. R. Martin, C. R. Olson, B. J. Heidinger, and J. J. Fontaine. water pythons. Ecology 78:1713–1721. 2000. Parental care and clutch sizes in North and South American birds. Sibley, C. G., and J. E. Ahlquist. 1990. Phylogeny and classification Science 287:1482–1485. of birds: A study in molecular evolution. Yale Univ. Press, New Mayfield, H. 1975. Suggestions for calculating nest success. Wils. Bull. Haven, CN. 87:456–466. Skutch, A. F. 1949. Do tropical birds rear as many young as they can nourish? Michod, R. E. 1979. Evolution of life histories in response to age-specific Ibis 91:430–455. mortality factors. Am. Nat. 113:531–550. Tieleman, B. I., J. B. Williams, and R. E. Ricklefs. 2004. Nest attentiveness Nalwanga, D., P. Lloyd, M. A. duPlessis, and T. E. Martin. 2004. Nest-site and egg temperature do not explain the variation in incubation periods partitioning in a strandveld shrubland bird community. Ostrich 75:250– in tropical birds. Funct. Ecol. 18:571–577. 258. Visser, M. E., and C. M. Lessells. 2001. The costs of egg production and Nice, M. M. 1954. Problems of incubation periods in North American birds. incubation in great tits (Parus major). Proc. R. Soc. Lond. B 268:1271– Condor 56:173–197. 1277. Olson, C. R., C. M. Vleck, and D. Vleck. 2006. Periodic cooling of bird eggs Ward, J. E. 1940. The passing of the lyre-bird. Bull. N. Y. Zool. Soc. 63:146– reduces embryonic growth efficiency. Phys. Biochem. Zool. 79:927– 152. 936. Weathers, W. W., and K. A. Sullivan. 1989. Nest attentiveness and egg tem- Ott, R. L. 1993. An introduction to statistical methods and data analysis. perature in the yellow-eyed Junco. Condor 91:628–633. Wadsworth, Inc., Belmont, CA. Webb, D. R. 1987. Thermal tolerance of avian embryos: a review. Condor Palacios, M. G., and T. E. Martin. 2006. Incubation period and immune 89:874–898. function: a comparative field study among coexisting birds. Oecologia Webb, G. J. W., and H. Cooper-Preston. 1989. Effects of incubation temper- 146:505–512. ature on crocodiles and the evolution of reptilian oviparity. Am. Zool. Parker, T. H., C. D. Becker, B. K. Sandercock, and A. E. Agreda. 2006. Appar- 29:953–971. ent survival estimates for five species of tropical birds in an endangered White, F. N., and J. L. Kinney. 1974. Avian incubation. Science 186:197–115. habitat in western Ecuador. Biotropica 38:764–769. Williams, G. C. 1966. Natural selection, the cost of reproduction, and a re- Purvis, A., and A. Rambaut. 1995. Comparative analysis by independent con- finement of Lack’s principle. Am. Nat. 100:687–690. trasts (CAIC): an Apple Macintosh application for analysing comparative data. Comp. Appl. Biosci. 11:247–251. Qualls, C. P., and R. M. Andrews. 1999. Cold climates and the evolution of viviparity in reptiles: cold incubation temperatures produce poor-quality Associate Editor: D. Pfennig

EVOLUTION 2007 11 MARTIN ET AL.

Appendix. List of species included in this study. Appendix Continued

Cardellina rubrifrons Arizona Buarremon torquatus Argentina Catharus guttatus Arizona Catharus dryas Argentina Certhia americana Arizona Elaenia parvirostris Argentina Dendroica coronata Arizona Elaenia strepera Argentina Empidonax occidentalis Arizona Geothlypis aequinoctialis Argentina Junco hyemalis Arizona Lathrotriccus euleri Argentina Oporornis tolmiei Arizona Myioborus brunniceps Argentina Pipilo chlorurus Arizona Parula pitiayumi Argentina Piranga ludoviciana Arizona Phylloscartes ventralis Argentina Poecile gambeli Arizona Thraupis sayaca Argentina Sialia mexicana Arizona Todirostrum plumbeiceps Argentina Sitta canadensis Arizona Tolmomyias sulphurescens Argentina Sitta carolinensis Arizona Troglodytes aedon Argentina Sitta pygmaea Arizona Turdus nigriceps Argentina Troglodytes aedon Arizona Turdus rufiventris Argentina Turdus migratorius Arizona Zonotrichia capensis Argentina Vermivora celata Arizona Atlapetes semirufus Venezuela Vermivora virginiae Arizona Basileuterus tristriatus Venezuela Vireo gilvus Arizona Buarremon brunneinucha Venezuela Anthoscopus minutus SouthAfrica Catharus aurantiirostris Venezuela Apalis thoracica SouthAfrica Catharus fuscater Venezuela Cercotrichas coryphaeus SouthAfrica Dysithamnus mentalis Venezuela Cinnyris chalybea SouthAfrica Formicarius analis Venezuela Cisticola subruficapilla SouthAfrica Grallaricula ferrugineipectus Venezuela Cossypha caffra SouthAfrica Henicorhina leucophrys Venezuela Crithagra albogularis SouthAfrica Lathrotriccus euleri Venezuela Crithagra flaviventris SouthAfrica Leptopogon superciliaris Venezuela Emberiza capensis SouthAfrica Mionectes olivaceus Venezuela Euplectes capensis SouthAfrica Myadestes ralloides Venezuela Lanius collaris SouthAfrica Myioborus miniatus Venezuela Parisoma layardii SouthAfrica Myiophobus flavicans Venezuela Parisoma subcaeruleum SouthAfrica Myrmotherula schisticolor Venezuela Prinia maculosa SouthAfrica Platycichla flavipes Venezuela Pycnonotus capensis SouthAfrica Premnoplex brunnescens Venezuela Sphenoeacus afer SouthAfrica Pyrrhomias cinnamomea Venezuela Sylvietta rufescens SouthAfrica Thraupis episcopus Venezuela Zosterops pallidus SouthAfrica Troglodytes aedon Venezuela Arremon flavirostris Argentina Turdus olivater Venezuela Basileuterus bivittatus Argentina Turdus serranus Venezuela Basileuterus signatus Argentina Zimmerius chrysops Venezuela

Continued

12 EVOLUTION 2007