Context-dependent costs and constraints of begging and non-begging activity by common grackle nestlings at the scale of the nanoclimate

Authors


Summary

  1. Environmental factors such as heat and solar radiation directly affect the open-cup nest environment and can impact nestling body temperature.

  2. The aim of our research was to understand how open-cup passerine nestlings use behavioural thermoregulation to mitigate solar heat gain and to measure the metabolic cost of begging and non-begging activity under three light treatments: room light (<2 Wm−2), simulated shade (500 Wm−2) and simulated sun (1000 Wm−2). Our study coupled behavioural (field) and physiological (laboratory) experiments to explore the adaptive behavioural response of passerine nestlings to nest cup nanoclimates (sun and shade), using the common grackle (Quiscalus quiscula) as a model species.

  3. Non-begging behaviour was evaluated in the field by randomly assigning nestlings to a homogeneous (all sun) or heterogeneous (sun and shade) nest environment; measuring body temperature (Tb); and identifying behaviours from videotaped footage. The Tb of nestlings without access to shade was significantly higher, and nestlings spent more time moving and panting, suggesting an increased metabolic expenditure. To test this, we duplicated the insolation nanoclimate under laboratory conditions using a 1·0 KW Sciencetech Illumination System and measured the energetic cost of begging and non-begging activity using open-system respirometry on non-irradiated and irradiated nestlings. Our results suggest that, in grackle nestlings, the relationship between activity (begging and non-begging) and energy expenditure is context-dependent.

  4. Our results identified: (i) a reduction in energy expenditure associated with accessing the shade; (ii) expenditure on non-begging activity equalled or exceeded expenditure on begging activity; and (iii) a potential behavioural constraint on begging imposed by exposure to a homogeneous sun environment (sun without access to shade). The combination of context-dependent costs and constraints suggests the potential for strategic costs associated with the heterogeneous nest environment.

Introduction

Altricial passerine nestlings are especially vulnerable to overheating due to their small size and low thermal capacity (Visser 1998). Conditions that affect their immediate nest environment, such as ambient temperature and solar radiation (Webb & King 1983; Wolf & Walsberg 1996), directly influence nestling body temperature, as well as their begging and non-begging activity. Short-wave radiation and convective cooling processes (resulting from shade or wind) have been identified as important routes of heat transfer for open-cup nestlings (Webb & King 1983). At the scale of the ‘nanoclimate’ – environmental conditions on the spatial scale of the organism (sensu Tracy 1977; Bakken 1992) – young birds can experience high heat loads when solar radiation exacerbates the effect of ambient temperature.

Under natural conditions, direct exposure to solar radiation quickly elicits heat stress behaviour, including panting and non-begging movements such as jostling and shade seeking (including neck stretching and whole-body movements) (Morton & Carey 1971; Gotie & Kroll 1973; Lustick, Battersby & Kelty 1979; Rauter, Reyer & Bollmann 2002; Glassey & Amos 2009). In red-winged blackbird (Agelaius phoeniceus) nestlings, optimal temperature-dependent performance at the tissue (i.e. isolated muscle) and whole-animal scale is restricted to a narrow ‘temperature tolerance zone’ (sensu Choi & Bakken 1990). Ambient temperatures that exceed this tolerance zone can overwhelm nestling thermolytic mechanisms, resulting in impaired begging behaviour, restricted locomotory activity and delayed response time (Choi & Bakken 1990; review in Hohtola & Visser 1998). Further, begging behaviour may be constrained as more time is spent panting (Choi & Bakken 1990). Ultimately, prolonged exposure to solar radiation may negatively affect nestling growth, development and survival (Bartholomew & Dawson 1954; Morton & Carey 1971; Lustick et al. 1979; Murphy 1985; Lloyd & Martin 2004).

Begging and non-begging behaviours contribute to activity costs which, together with growth, thermoregulation and maintenance (measured as resting metabolic rate, RMR), comprise the activity budget of nestlings (Weathers 1992; review in Chappell & Bachman 2002). To be considered honest from an evolutionary perspective the energetic cost of begging must reduce individual fitness by impacting one of the other contributing factors (e.g. growth, Kilner 2001; review in Moreno-Rueda & Redondo 2011; non-begging activities, review in Chappell & Bachman 2002; RMR, review in Burton et al. 2011).

In passerine nestlings, measurements of individual energy expenditure associated with begging have focused primarily on oxygen consumption under artificial light conditions (Leech & Leonard 1996; McCarty 1996; Bachman & Chappell 1998; Chappell & Bachman 1998). These laboratory studies identified the energetic cost of begging by individuals as being relatively low, contrary to the predictions of theoretical signalling models (reviews in Chappell & Bachman 2002; Johnstone & Godfray 2002; Moreno-Rueda 2007). However, Leech & Leonard (1996) alluded to the potential for weather to increase the energetic cost of begging signals, and, more recently, Burton et al. (2011) linked RMR to growth and survival by postulating that environmental variation contributes to context-dependent, individual variation in RMR.

We view the thermal environment of the open-cup nest as a product of two decisions made at different scales: (i) a parental decision at the scale of the nest site involving a trade-off between predation and microclimate, a spatial scale at the level of the nest and surrounding vegetation (e.g. Martin 2001; Lloyd & Martin 2004); and (ii) a nestling decision about movement, which may involve a trade-off between activities involved in resource acquisition: begging vs. accessing the preferred nanoclimate. Although the energetic cost of begging as a signal by passerine nestlings has been well explored, empirical testing of the impact of environmental variation in the cost of activity metabolism is lacking.

The broad aim of our study is to assess the effect of solar radiation on nestling activity (begging and non-begging) and energy expenditure in the heterogeneous open-cup nest environment to determine whether these costs are context-dependent – in this case, associated with specific environmental conditions, including the risk of overheating (Piersma 2011) at the nanoclimate scale.

Context dependence associated with environmental heterogeneity at coarser scales is suggested by studies of small adult passerine and juvenile non-passerine birds. Specifically, whereas exposure to simulated high-intensity short-wave radiation under moderate-to-high ambient temperatures contributes to increased body temperature and metabolic expenditure in adult white-crowned sparrows (Zonotrichia leucophrys; De Jong 1976) and verdins (Auriparus flaviceps, Wolf & Walsberg 1996), both the rate of heat gain and evaporative water loss can be substantially reduced by moving a few centimetres, from the sun to the shade (Wolf & Walsberg 1996). Juvenile gulls (Larus argentatus) in a heterogeneous environment are similarly capable of manipulating heat exchange and reducing energetic expenditure by moving to the shade (Dunn 1976a; Lustick et al. 1979), and among altricial seabirds, the energy expenditure of double-crested cormorant (Phalacrocorax auritus) chicks is higher in the sun than in shade (Dunn 1976b).

Our research couples field and laboratory experiments to explore the effect of solar radiation on the behaviour (begging and non-begging) and oxygen consumption using the common grackle (Quiscalus quiscula) as a model species for open-cup passerine nestlings. Common grackle nestlings are capable of detecting differences in nanoclimate, and in a thermally heterogeneous environment respond behaviourally by moving to cooler, more favourable environmental conditions within the nest cup (Glassey & Amos 2009).

We predict that if begging cost varies with nanoclimate (sun vs. shade), exposure to simulated sunlight will increase the energetic cost of begging over simulated shade and of simulated shade over room light. Furthermore, nestlings exposed to simulated sun without access to shade are predicted to expend more time and energy on non-begging behaviour – specifically, on movements associated with shade seeking – than under exposure to simulated shade or room light, reducing the time nestlings spend begging.

Materials and methods

Field methods

Field experiments were conducted in the interval between May 26 and 20 June 2007 between 10.00 and 15.00 h under clear, sunny, warm (T= 18·4 ± 0·6 °C) conditions at four local wetland sites near Sydney, Nova Scotia, Canada. Thirteen nestling pairs, matched for age and size, were transported to an experimental set-up immediately adjacent to the natal wetland (Glassey & Amos 2009). Following Glassey & Amos (2009), each pair was from a different brood in the ‘transitional stage’ (sensu Morton & Carey 1971) of development (d6–d10) between ectothermic and endothermic temperature regulation, in which the behavioural response of grackles to heat stress may include escape movements, postural adjustments and panting (Gotie & Kroll 1973).

Nest cup temperatures were recorded in the sunny (Tsun) and shaded (Tshade) regions of the heterogeneous nest and from corresponding positions (Tsun1, Tsun2) in the homogeneous nest; ambient air temperature was measured above the nest in the shade of the researcher's body (Glassey & Amos 2009).

Cloacal body temperature (± 0·1 °C) of each nestling was measured at the beginning (Tb initial) and at the end (Tb final) of the trial using a hand-held thermocouple thermometer (YSI Barnant, Barrington, IL, USA).

Behaviour involved in thermoregulation was evaluated in 15-min field trials. Nestlings from each pair were randomly assigned to either a homogeneous (all sun) or heterogeneous (sun and shade) nest environment. Nestlings were oriented in the same direction at the start of each trial and handled in the same manner.

Nestling activity during the trial was recorded continuously, using a SONY Digital8 video camera (Kitashinagawa, Shinagawa-ku, Tokyo, Japan) mounted approximately one metre above the experimental apparatus. Videotape observations were recorded from 11/13 field trials. Two field trials were excluded owing to equipment malfunction. In addition, a series of still photos was taken every 5 min (times 0, 5, 10, 15) with a Pentax Optio 43WR digital camera (Golden, CO, USA).

We subsequently documented the latency (time to start) and duration of the following non-begging behavioural responses: panting, horizontal neck extension, escape behaviour and mouth open – hereafter gaping – for the duration of the 15-min videotape footage, using Capture Wizard 3·0 (ADS Technologies, Cerritos, CA, USA). Nestling orientation at each 5-min interval was recorded from the still photos using a compass overlay, measuring the angle between 0° (orientation at time 0) and the central body axis (from tail to bill tip). Following Glassey & Amos (2009), change in orientation, measured in degree distance, was used as an estimate of whole-body movement. Latency of whole-body movement was measured as described for the other four non-begging behavioural responses above.

Laboratory methods

Laboratory experiments were conducted on twelve randomly selected nestlings (aged d8–d10), each from eleven different broods, between May and June 2008. Mass and cloacal body temperatures were recorded at the natal nest, and then, nestlings were transported temporarily from local wetlands near Sydney, NS, Canada, to the Avian Thermal Ecology Lab at CBU. Upon arrival, each nestling was placed in an empty nest where it was acclimated under infrared light until its cloacal body temperature (Tb) was within 1 °C of field body temperature. Two nestlings that exhibited escape behaviour when introduced into the chamber were immediately excluded from the trial due to animal care concerns.

Experimental design

We duplicated the radiative environment of the nest cup under laboratory conditions using a 1·0 KW Sciencetech Illumination System and used open-system respirometry to measure the energetic cost of begging and non-begging activity of individual nestlings under three simulated light treatments. Treatment 1, an absence of solar radiation, was simulated by room light (<2 Wm−2). Treatments 2 and 3, designed to mimic levels of solar radiation similar to those experienced by nestlings in the nest cup during daylight hours under natural, clear conditions (Glassey & Amos 2009), corresponded to irradiance conditions of shade (500 Wm−2) and sun (1000 Wm−2), respectively, following Wolf & Walsberg (1996). We maintained consistency of irradiance values by daily system calibrations using a Gentec Power Detector.

Prior to treatment 1 (absence of solar radiation), nestlings were equilibrated to the respiratory chamber (see below) until the differential O2 measurement stabilized (Chappell & Bachman 1998), a period lasting on average 25·83 ± 1·49 min. Then, the resting metabolic rate (RMR, VO2 mL min−1) of each subject was measured for 10 min without stimulation. Nestlings were then stimulated to beg by casting a hand across the chamber and randomly tapping the chamber throughout the 10-min period to simulate the arrival of a parent (McCarty 1996) (AMRRoomLight; J∙g−1∙s−1). VO2 (mL min−1) measurements associated with begging and non-begging activities were measured during this period.

Following treatment 1, nestlings were consecutively assigned to the simulated shade (treatment 2) or simulated sun (treatment 3) in random sequence (room light–sun–shade: n = 7; room light–shade–sun: n = 4). Following a 10-min equilibration period, oxygen consumption (VO2 mL min−1) was measured following the same protocol as outlined for treatment 1.

Each nestling acted as its own control and remained in the chamber between consecutive treatments for a 10-min recovery session under room light without stimulation. Spontaneous begging by nestlings during the treatments was rare. Whole-body movements were common, although treatment specific. Since whole-body movements occurred between begging bouts, we categorized them as non-begging activity, consistent with the field observations.

Begging activity typically involved gaping and vertical neck stretching, typical of passerine nestlings (Glassey & Forbes 2002; Kilner 2002). The location of the solar beam and the walls of the nest cup restricted the positioning of the camera, precluding detailed measurements of begging intensity (e.g. wing flapping, leg stretching), so following McCarty (1996), we report the total duration of begging. We conducted a Mann–Whitney U-test to test for the possibility of an order effect. The analysis revealed no significant difference in beg duration (s) with treatment order under exposure to simulated solar radiation (shade: Z = 14·0, = 0·571; sun: = 12·0, = 0·705), suggesting that increasing hunger was likely not a significant factor in the duration of observed begging behaviours.

Observations of nestling behaviour were documented in real-time (Qubit Systems C950 acquisition software) and were also recorded using a SONY Digital8 video camera (Kitashinagawa) for later analysis (Ulead VideoStudio 7 DVD media software, Ulead Systems, Torrance, CA, USA). Temporary equipment malfunction prevented data collection on the duration of whole-body movements for one sample under simulated solar radiation, reducing the sample size to n = 11.

Following the final experimental treatment, the cloacal body temperature and mass of each nestling were recorded, each was fed to satiation from a 1·0-cm3 syringe using canned cat food (Clotfelter et al. 2003) and transported back to the nest of origin the same day. Nestlings lost an average of 3·58 ± 0·76 g (n = 12) (faecal sac included), a significant reduction (t11 = 4·793, = 0·001) from field mass. Nestlings did not defecate during the experiment, but the majority (n = 9) of nestlings released a faecal sac (averaging 3·28 ± 0·26 g) when they were weighed following treatment. Nestlings consumed 3·44 ± 0·83 g (n = 8) of cat food. Consumption data were unavailable for n = 4 nestlings. Nestlings were typically fed until they either stopped begging or (with one exception) ingested 4·00 g.

Respirometry measurement

We measured the energetic cost of begging and non-begging activity using open-system respirometry (Lighton 2008) following the methods of Bachman & Chappell (1998). We used an 8·5 cm ID Qubit Systems (Kingston, ON, Canada) environmental chamber fitted with a flint glass window to protect nestlings from potentially damaging ultraviolet radiation (Wolf & Walsberg 1996). The chamber was equipped with two Qubit Systems Peltier Temperature Controllers powered by BK Precision power supply units, to maintain thermoneutral temperatures between 32 °C and 34 °C. Chamber temperature measurements were downloaded at 1-s intervals from a Qubit Systems thermistor. A nest cup was placed in the chamber to ensure consistency between nestling movements in laboratory and field experiments, and light was directed into the nest from above the chamber.

We measured O2 (%) of the sample (excurrent) air using dry air pumped through Tygon® tubing under standard temperature and pressure conditions (STP). Baseline oxygen concentration and consumption was measured at 1-s intervals, while simultaneously directing flow rates (FR) of incurrent air (850 mL min−1) and excurrent air (1700 mL min−1) through the S104 Differential O2 Analyzer (DOX) using a G276 Dual Pump. Baseline concentrations were adjusted for analyser drift prior to analysis.

Reference air was sampled automatically, generating on average 300 baseline data points every 600 s. Oxygen sensors were calibrated daily by directing 150 mL min−1 of magnesium perchlorate dried, CO2-free (sodium lime) 21% air (adjusted for atmospheric pressure) through the Qubits S104 Differential O2 Analyzer using a Qubits G276 Dual Pump.

Carbon dioxide exchange was measured for FeCO2 by pumping dried air through the S157 CO2 Analyzer using the P251 pump at a flow rate of 80 mL min−1. The CO2 sensor was zeroed daily using dry, CO2-free 21% balanced air and spanned against dry, 495 ppm CO2 pumped through an S157 CO2 Analyzer by a P651 pump at a rate of 200 mL min−1 (G152).

Metabolic measurement

The rate of oxygen consumption (VO2) was derived by converting %CO2 using the equation specified for respirometry systems where the flow meter is located upstream and CO2 is not removed (Mode2):

display math

where FR, flow rate (mL min−1); Fi, input fractional concentration (%); and Fe, excurrent fractional concentration (%) (http://warthog.ucr.edu/WartHogPage/LAX%20website/respirometry.html).

Following Leech & Leonard (1996), VO2 (mL min−1) measurements associated with begging and non-begging activities were expressed as energy turnover/unit time (e.g. J·g−1·s−1) using a constant of 20·8 ( kJ L−1 O2) (Williams & Prints 1986). We calculated energetic cost as the metabolic (aerobic) scope of activity (AMR:RMR), using the ratio of activity metabolic rate (specified as either AMRbeg or AMRnonbeg) to resting metabolic rate (RMR) for each light treatment (e.g. Bachman & Chappell 1998). We then incorporated measurements of begging duration to calculate the incremental (I) and instantaneous (CI) energy costs of begging (Leech & Leonard 1996; McCarty 1996).

Statistical analyses

We used the Kolmogorov–Smirnov test to test our data for normality, and applied nonparametric tests when the assumption of normality was not met. Oriana 2.0 software (Kovach Computing Services) was used to analyse circular data (Watson–Williams F-test, single-sample binomial test; Zar 1999), and ibm spss Statistics 20 software was used to analyse linear data. A Friedman test was used to compare metabolic output, energy expenditure and beg duration across laboratory treatments, followed by a Wilcoxon signed-rank test on significant outcomes to identify differences between treatment pairs. A Bonferroni adjustment (P value/#tests) was applied to the Wilcoxon signed-rank test results. Analysis of data from field experiments followed the statistical protocol used in Glassey & Amos (2009). All tests were two-tailed and used an alpha value of 0·05 to determine significance.

Results

Field experiments

Air temperature on the irradiated side of the heterogeneous nest was significantly warmer than on the shaded side (Tsun = 38·37 ± 1·00 °C; Tshade = 27·61 ± 1·12 °C; paired t-test: t9 = 14·44, < 0·001), but as expected temperatures associated with the corresponding regions in the homogeneous nest did not differ significantly (Tsun1 = 37·49 ± 1·08 °C; Tsun2 = 36·70 ± 1·31 °C; paired t-test: t9 = 1·179, = 0·261).

Initial body temperature (Tb initial) was not significantly different between nestling pairs (Tb initial difference = 0·34 ± 0·26 °C; paired t-test: t12 = 1·303, = 0·217). Following exposure, Tb increased significantly in both nests, reaching 41·03 ± 0·20 °C in the heterogeneous nest (paired t-test: t12 = −2·670, = 0·020) and 42·39 ± 0·34 °C in the homogeneous nest (paired t-test: t12 = 5·397, < 0·001). Heat gain by nestlings in the heterogeneous nest was significantly lower than nestlings without access to shade (paired difference: 1·35 ± 0·25 °C; t12 = −5·398, < 0·001; Fig. 1).

Figure 1.

Heat gain (∆°C) by nestlings in heterogeneous and homogeneous nest cup environments. *Paired samples test < 0·05. n = 12.

Five key thermal behaviours, including neck extension, whole-body movements, gaping, panting and escape behaviour, were identified from videotape observations of 11/13 field trials. Two field trials were excluded owing to equipment malfunction.

Escape attempts were only observed in the homogeneous nest, where two nestlings spent an average of 1·80 ± 1·34 min attempting to mount the nest rim (Table 1). Conversely, nestlings that initiated gaping did so within the first 4 min of the experiment (paired difference: 0·74 ± 1·80 min; independent t-test: t19 = 0·419, = 0·680) and continued to gape for the duration of the experiment. There was no detectable difference in gape duration between treatments (Table 1). In contrast, panting was observed in the majority of nestlings in the homogeneous nest (75%), but in only three nestlings in the heterogeneous nest (23%) so that on average, nestlings in the homogeneous nest spent significantly more time panting compared to the heterogeneous treatment (Table 1). Once initiated (heterogeneous = 11·34 ± 6·62 min, homogeneous = 14·60 ± 4·12 min; independent t-test: t10 = −0·400, = 0·698), panting typically continued for the duration of the experiment.

Table 1. The duration (min) of behavioural response by nestlings exposed to heterogeneous and homogeneous nest cup environments. n = 11
BehaviourHeterogeneous nestHomogeneous nestPaired t-testP (two-tailed)
Gaping9·70 ± 1·6611·17 ± 1·19t10 = −1·772= 0·107
Panting1·51 ± 0·816·87 ± 1·54t10 = −3·508= 0·006
Neck extended4·86 ± 1·6411·92 ± 0·81t10 = −3·932= 0·003
Escape0·00 ± 0·001·80 ± 1·34  

All nestlings extended their necks beyond the nest rim after the first minute in the homogeneous nest earlier, albeit non-significantly, than the seven nestlings that exhibited the behaviour in the heterogeneous nest (paired difference: 3·85 ± 1·76 min, t6·5 = 2·19, = 0·068). Overall, nestlings in the homogeneous nest spent significantly more time neck stretching compared to the heterogeneous treatment (Table 1).

Nestlings typically initiated whole-body movement within 60 s of the start of the experiment (heterogeneous = 0·74 ± 0·42 min, homogeneous = 0·46 ± 0·18 min; paired t-test: t20 = 0·610, = 0·549). However, whereas nestlings responded behaviourally to nanoclimate heterogeneity by orienting to the shade (Binomial test: 12/13, < 0·003; Fig. 2a), proportionally fewer nestlings oriented in the corresponding direction under homogenous conditions (Binomial test: 8/13, = 0·581; Fig. 2b). Rather, circumnavigation of the homogeneous nest cup contributed to a uniform nestling distribution (Rayleigh test: = 0·06, = 0·94, Fig. 2b). In the homogenous nest, nestlings covered four times the degree distance than nestlings in the heterogeneous nest, where directional movement oriented to shade was evident (Fig. 2a)

Figure 2.

The final orientation (μ) of nestling pairs exposed to (a) heterogeneous and (b) homogeneous nest cup conditions. Shade is depicted as grey and sun as white. The start angle (0°/360°) at time 0 is 0°, and triangles depict the directionality of individual nestling response. The direction of the arrow represents the mean angular direction of final nestling orientation, and the length of the arrow is a measure of nestling dispersion (r). Results of a one-sample Watson–Williams F-test comparing the initial vs. final orientation are presented for each vector. n = 11.

Laboratory results

Metabolic expenditure

Essentially, it was metabolically more costly for a nestling to be in the sun, regardless of the activity, than it was to be in the shade. Activity metabolism increased by 10% over resting in the sun (AMRnon-begging/RMR), which represented a significant increase over the scope of both non-begging (AMRnon-begging:RMR = 1·05; Z = −2·982, P = 0·003) and begging activity (AMRbegging:RMR = 1·03; = 3·059, = 0·002) under shade exposure (Fig. 3).

Figure 3.

The proportional increase in activity metabolic rate above RMR for begging and non-begging activity in the absence of solar radiation (room light; <2 Wm−2) and under two intensities of simulated solar radiation: shade (500 Wm−2) and sun (1000 Wm−2). *Wilcoxon signed-rank test < 0·05 for room light vs. shade and sun vs. shade. < 0·05. n = 12.

There was an inverse relationship between begging duration and intensity of solar radiation (Wm−2). Time spent begging was longest in treatment 1 (room light), and shortest in treatment 3 (sun) (sun vs. room light: Wilcoxon signed-rank test Z = 3·059, = 0·002; Table 2).

Table 2. Metabolic output and energy expenditure associated with begging and non-begging activity in laboratory treatments using three categories of exposure, that is, the absence of solar radiation (room light; <2 Wm−2), simulated shade (500 Wm−2) and simulated sun (1000 Wm−2). n = 12
MeasurementTreatment 1 Room light (<2 Wm−2)Treatment 2 Shade (500 Wm−2)Treatment 3 Sun (1000 Wm−2)Friedman test
  1. a

    n = 11.

Begging activity
Vo2Beg (mL min−1)3·790 ± 0·1153·636 ± 0·1053·864 ± 0·127math formula = 15·17, = 0·001
AMRNonBeg (J·g−1·s−1)0·019 ± 0·0010·019 ± 0·0010·020 ± 0·001math formula = 15·17, = 0·001
I (J g−1 s−1)0·002 ± 0·0000·001 ± 0·0010·006 ± 0·001math formula = 10·67, = 0·005
CI1·103 ± 0·0171·061 ± 0·0261·359 ± 0·081math formula = 10·67, = 0·005
Beg duration(s)237·25 ± 11·65156·58 ± 21·30115·83 ± 16·10math formula = 15·17, = 0·001
Beg (percentage)63·7142·5431·63 
WBM duration (s)0·00 ± 0·001·42 ± 1·1710·66 ± 3·05 
WBM degrees moved (°)15·00 ± 15·0054·17 ± 31·63455·00 ± 126·02 
Non-begging activity
Vo2NonBeg (mL min−1)3·938 ± 0·1223·710 ± 0·1093·893 ± 0·133math formula = 12·67, = 0·002
AMRNonBeg (J·g−1·s−1)0·020 ± 0·0010·019 ± 0·0010·020 ± 0·001math formula = 12·67, = 0·002
WBM duration (s)0·57 ± 0·575·59 ± 2·5145·56 ± 10·65a 
WBM degrees moved (°)3·33 ± 3·33126·67 ± 64·511020·83 ± 203·81 

There was also significant variation in the incremental cost (I) of begging that nestlings expended across all three treatments; room light was intermediate between shade and sun (Friedman test: x2 = 10·67, = 0·005; Table 2).

Metabolic expenditure: sun vs. shade

Nestlings consumed oxygen (AMRnon-begging; J·g−1·s−1) at a significantly higher rate and expended significantly more energy on non-begging activity in the sun, than those in the shade (Wilcoxon signed-rank test Z = −2·981, = 0·003; Table 2).

Laboratory results also confirmed our prediction that nestlings would consume significantly more oxygen (mL O2 per min) when stimulated to beg in the sun relative to the shade. Oxygen consumption, expressed as activity metabolic rate (AMRbeg; J∙g−1∙s−1), was significantly higher in sun than in shade (Wilcoxon signed-rank test Z = −3·059, = 0·002; Table 2).

Application of the Bonferroni adjustment to the results of the Wilcoxon signed-rank test generated an adjusted probability value of 0·017, meaning that the reduction in beg duration by nestlings in the sun relative to the shade approached, but did not meet, significance (Wilcoxon signed-rank test: Z = −2·118, = 0·034; Table 2), likely due to the small sample size contributing to low statistical power. The biological relevance of the difference in beg duration was, however, clearly evident in the results generated for the cost of begging. Specifically, the incremental cost of begging (I) was significantly higher under sun exposure (0·006 J·g−1·s−1) relative to shade exposure (0·001 J·g−1·s−1) (Wilcoxon signed-rank test Z = −3·059, = 0·002). This means that nestlings consumed significantly more energy during 1 s of begging under sun exposure, despite begging for about 10% less time. As expected, the instantaneous cost of begging (CI), a ratio derived from I, also showed the same pattern.

Predictably, the highest metabolic scope (AMR: RMR) was associated with begging in the sun where the begging metabolic rate exceeded RMR (3·55 ± 0·10 mL/O2 per min) by an average of 9%, significantly higher than the 3% factorial increase associated with exposure to the shade (Wilcoxon signed-rank test sun vs. shade: Z = −3·059, = 0·002; Fig. 2).

Non-begging Activity: whole-body movements

When not stimulated to beg, nestlings under sun exposure spent significantly more time moving than under shade (Wilcoxon signed-rank test Z = −2·547, = 0·011) or room light conditions (Wilcoxon signed-rank test Z = −2·521, = 0·012; Table 2). The degree distance moved by nestlings during exposure to the sun was equivalent to circumnavigating the nest cup 2·8 times (Table 2). In comparison, nestlings moved roughly one-third the circumference of the nest under shade exposure and less than a body width under room light. Lack of movement by nestlings under exposure to sun and shade generated a high number of zeros, precluding statistical analysis of duration data across treatments.

Nestlings also moved between begging bouts when exposed to the sun and shade (duration sun vs. shade: Wilcoxon signed-rank test Z = −2·100, = 0·036; Table 2), but did not move when not begging under room light conditions.

Discussion

Our results highlight the importance of nanoclimate selection (accessing shade) by nestling grackles in successfully regulating their body temperature in a heterogeneous nest environment, supporting our previous suggestion (Glassey & Amos 2009) that thermally directed movements are an adaptive response to the variable thermal and radiative environment of the open-cup nest. The field and laboratory experiments reported here suggest that nestlings use shade seeking as a strategy to mitigate the effect of solar radiation on heat gain and energy expenditure.

In a heterogeneous environment, grackle nestlings regulated body temperature behaviourally by initiating whole-body movements when exposed to sun and stopping movement when their heads were oriented to the shade, consistent with the interpretation of head-shading as a thermoregulatory strategy previously described by Glassey & Amos (2009). Whereas nestlings exposed to a homogeneous shade environment do not initiate movement or adjust their orientation (Glassey & Amos 2009), nestlings without access to shade (homogeneous sun environment) moved throughout the experiment, in agreement with earlier anecdotal observations of apparently heat-stressed nestlings in full sun (Morton & Carey 1971; Murphy 1985). Laboratory results confirmed an energetic cost associated with unsuccessful shade seeking movements. Whether nestlings exposed to sunlight with access to shade similarly exhibit reduced begging behaviour and a higher energetic cost warrants future testing.

In contrast to whole-body movements, nestling behaviours such as gaping and panting, once initiated, generally continued for the duration of the experiment, regardless of exposure. Gaping is temperature-insensitive (Choi & Bakken 1990) and considered an accessory behavioural component of the nestling begging repertoire (Mugaas & King 1981). Gaping behaviour, combined with mouth colour, may signal to parent birds that their nestlings need shade at high ambient temperatures (e.g. Ta > 39 °C, Clotfelter et al. 2003). Gaping in the shade likely benefits nestlings by cooling the vascularized surface of their mouths and reducing evaporative water loss (e.g. Lustick et al. 1979).

Nestlings in full sun without access to shade experienced higher body temperatures than those in a variable radiative environment (see also Bartholomew & Dawson 1954; Lustick et al. 1979; Glassey & Amos 2009), emphasizing the importance of the heterogeneous nest environment to mitigating short-term heat loads behaviourally. Their body temperatures approached the upper threshold of the ‘temperature tolerance zone’ (Tb 43–44 °C) within which the begging activity of nestling red-winged blackbirds (open-cup nesting icterids) is constrained by panting and a loss of coordination (Choi & Bakken 1990).

The laboratory experiments show that nestlings can save energy by accessing the shade, and identify the mechanism as a reduction in activity energy expenditure, an interpretation that highlights the contribution of the environment (nanoclimate) to nestling behavioural response and energy expenditure. Nestlings consumed significantly more oxygen under exposure to simulated sunlight (1000 W m−2) than under simulated shade (500 W m−2) consistent with the findings of Wolf & Walsberg (1996) and Lustick et al. (1979). We attribute the increase in non-begging activity primarily to panting (evaporative cooling), a likely increase in the heat increment of activity (H. I A.; Bicudo et al. 2010) associated with whole-body movement, and the absorption of solar energy.

Metabolic expenditure on begging fell within the lower range of previously published values of metabolic scope measured by respirometry for passerine begging activities under artificial light (begging: 0·90–1·28, McCarty 1996; Leech & Leonard 1996; Bachman & Chappell 1998), but below the reported values for begging calls under artificial light (1·2–3·1, Jurisevic et al. 1999), forced exercise (1·45–1·70; Chappell & Bachman 1998) and thermogenesis (1·91, Chappell & Bachman 1998). However, metabolic expenditure by grackle nestlings on non-begging activity equalled or exceeded that expended in begging, consistent with observations reported for house wrens (Troglodytes aedon; Bachman & Chappell 1998). Direct measurement of daily energy expenditure (DEE) (great spotted cuckoo, Clamator glandarius, magpie, Pica pica; Soler et al. 1999) and extrapolation of house wren metabolic results to daily energy budgets (Bachman & Chappell 1998; Chappell & Bachman 1998) demonstrated the same pattern: more time and energy is allocated to the non-begging component of the nestling activity budget than to the begging component.

Our study identified both relatively higher energetic costs as well as behavioural constraints as potentially context-dependent consequences of nanoclimate heterogeneity. This novel finding indicates that ecological factors can affect begging signals. Begging in the sun is energetically more costly than begging in the shade, suggesting that the small-scale nest cup environment may add a cost to signalling for nestlings by imposing an ecological constraint on their activity metabolism and begging behaviour. Our results show that nestlings beg at higher cost in the sun compared to the shade, despite begging for less time. The activity cost of begging in the sun exceeds the cost of begging in the shade by 6% (AMRbeg(shade):RMR = 1·03 vs. AMRbeg(sun):RMR = 1·09). Given that correcting for beg duration produces an instantaneous cost of begging (CI) in the sun of 36%, more than six times the instantaneous cost of begging in the shade (CIshade = 1·06), the length of time that nestlings beg appears to be constrained by their exposure to solar radiation.

According to a review by Számadó (2011), begging can be a reliable signal of need under environmentally imposed constraints, stipulating an increase in AMR relative to RMR as a necessary precondition. Our findings show that when metabolic scope <1·05, AMR does not differ significantly from RMR (see also McCarty 1996). Nestling AMR significantly exceeds RMR under both room light and simulated solar radiation, thereby meeting the initial precondition (Számadó's 2011). However, relative to room light, the radiative environment imposes a constraint on nestling begging behaviour in addition to their relatively higher activity energy costs. Whereas the latter (energy cost) can be assessed under artificial light, potential environmental constraints to begging that would inform the evaluation of strategic costs may be overlooked.

Contrary to our expectations, the activity cost of begging to grackle nestlings tested under room light was intermediate between simulated shade and sunlight, a result we attribute to prolonged begging by non-irradiated nestlings. The treatment order may have contributed to longer begging by nestlings when exposed to room light relative to simulated shade and sun; since we consistently measured room light first to facilitate direct comparison with other studies nestlings may have been less tired and able to beg for longer. Regardless of treatment, the initial begging response typically involved gaping with vertical neck stretching (BG, pers. obs.) indicative of intense begging (e.g. Cotton, Wright & Kacelnik 1999). We speculate that nestlings were capable of begging harder under room light because the environmental context required them to expend less energy on thermal activity – specifically moving – between begging bouts, so that they benefitted from a ‘recovery’ period.

The data presented here show that the energy costs associated with reduced non-begging activity were not reallocated to begging activity; rather they appear to have contributed to a reduction in overall activity costs. This suggests that context-dependent costs may be additive. Számadó (2011) suggested there is a strategic cost to signals through the imposition of an ecological constraint on activity metabolism and begging behaviour such that cheaters and honest signallers are both impacted. We predict that nest cup heterogeneity provides a mechanism for adding a strategic cost.

Our results suggest that the thermal environment can impose differential costs and constraints on offspring within the same brood. That siblings could experience dramatically different conditions, potentially living side by side in different worlds, has deep implications for the study of sibling rivalry (Forbes 2013). It suggests that one of the context-dependent consequences of nest cup heterogeneity may be conflict for a thermal resource in which sibling competition mediates differences in survival rates through the thermal environment. The implications of younger offspring being unable to access the preferred environment may extend upwards from asymmetric sibling rivalry (sensu Forbes & Glassey 2001) to parent–offspring conflict (review in Mock & Parker 1997; Forbes 2007). Since patches of sun and shade are created by parents through nest cup architecture and microsite selection (e.g. With & Webb 1993; Lloyd & Martin 2004; Glassey & Amos 2009), they play a role in determining the thermal outcome of offspring competition.

Although it is beyond the scope of this study to determine whether individual nestlings can strategically exploit the heterogeneous environment, or manipulate their total daily energy budget or a life-history component of it (e.g. growth, maintenance, RMR; Olson 1992; Burton et al. 2011) through ‘behaviourally based energy reallocation’ (sensu Williams & Ternan 1999; reviewed in Vézina, Speakman & Williams 2006), our results do suggest the potential for context-dependent, behaviourally based energy allocation (reviewed in Vézina et al. 2006) with its implication for survival (review in Burton et al. 2011) and intrabrood competition for shade as a thermal resource.

Acknowledgements

We are grateful to P. Hall, M. Leonard and S. Forbes for reviewing an earlier draft of the manuscript, and to T. Ayers for editorial comments. We appreciated the helpful comments and criticisms by two anonymous reviewers that greatly contributed to improvements in the manuscript. We would like to thank CBU undergraduate students P. Bonnar, M. DiPinto, E. Mizier-Barre, B. Furlong and E. Roach for assistance in the field and laboratory. Our sincere thanks to landowners C. Ash, A., J. and L. Maclean, and D. MacKinnon for access to nest sites. Funding for the project was provided by NSERC, NSRIT, a CFI New Opportunities Award and CBU RP Grants to B.G. Experiments were conducted under Cape Breton University's Animal Care Committee Licenses # 2007-02 and 2008-01 and complied with Canada Council of Animal Care guidelines. No birds sustained injuries or died during our experiments. The authors have no conflict of interests to declare.

Ancillary