288
Physiological and Behavioral Responses to an Acute-Phase Response in Zebra Finches: Immediate and Short-Term Effects Sandra Sko¨ld-Chiriac1,2,* Andreas Nord1 Jan-A˚ke Nilsson1 Dennis Hasselquist2 1 Section for Evolutionary Ecology, Department of Biology, Ecology Building, Lund University, SE-223 62, Lund, Sweden; 2Section for Molecular Ecology, Microbial Ecology and Evolutionary Genetics, Department of Biology, Ecology Building, Lund University, SE-223 62, Lund, Sweden Accepted 10/25/2013; Electronically Published 2/11/2014
ABSTRACT Activation of the immune system to clear pathogens and mitigate infection is a costly process that might incur fitness costs. When vertebrates are exposed to pathogens, their first line of defense is the acute-phase response (APR), which consists of a suite of physiological and behavioral changes. The dynamics of the APR are relatively well investigated in mammals and domesticated birds but still rather unexplored in passerine birds. In this study, we injected male zebra finches (Taeniopygia guttata) with a bacterial endotoxin (lipopolysaccharide [LPS]) to assess the potential physiological, immunological, and behavioral responses during the time course of an APR and also to record any potential short-term effects by measuring the birds during the days after the expected APR. We found that LPS-injected zebra finches decreased activity and gained less body mass during the APR, compared to control individuals. In addition, LPS-injected birds increased their production of LPS-reactive antibodies and reduced their metabolic rate during the days after the expected APR. Our results show that zebra finches demonstrate sickness behaviors during an APR but also that physiological effects persist after the expected time course of an APR. These delayed effects might be either a natural part of the progression of an APR, which is probably true for the antibody response, or a short-term carryover effect, which is probably true for the metabolic response.
* Corresponding author; e-mail:
[email protected].
Physiological and Biochemical Zoology 87(2):288–298. 2014. 䉷 2014 by The University of Chicago. All rights reserved. 1522-2152/2014/8702-2191$15.00. DOI: 10.1086/674789
Introduction Over the past 2 decades, trade-offs between immune function and other fitness-related traits have become an increasingly important topic in ecology and evolutionary biology (Sheldon and Verhulst 1996; Norris and Evans 2000; Zuk and Stoehr 2002; Martin et al. 2011). Activation of the immune system can have negative consequences on fitness-related traits such as nestling growth (Nilsson 2003; Soler et al. 2003; Tschirren and Richner 2006), parental care (Ra˚berg et al. 2000), breeding success (Ilmonen et al. 2000; Bonneaud et al. 2003; Marzal et al. 2007), and survival (Hanssen et al. 2004; Eraud et al. 2009). This suggests that traits involved in pathogen protection and infection mitigation might demand resources that could otherwise have been invested in other life-history traits (Sheldon and Verhulst 1996; Klasing 2004) or might induce long-term costs such as oxidative stress and immunopathology (Ra˚berg et al. 1998; Hasselquist and Nilsson 2012). Consequently, a better understanding of the progression of physiological responses that might influence resource allocation and induction of costs, and thus regulate immune function, is a central topic in current immunoecology research (Lochmiller and Deerenberg 2000; Owen-Ashley and Wingfield 2007). Activation and operation of the immune system may increase energy requirements, generally measured as a change in metabolic rate (Klasing 2004; Owen-Ashley and Wingfield 2007; Marais et al. 2011; Hegemann et al. 2012; but see Svensson et al. 1998). Metabolic rate is a heritable (Nilsson et al. 2009) yet flexible trait that can vary with season (Piersma et al. 1995; reviewed in McKechnie 2008), age (Broggi et al. 2009; Moe et al. 2009), breeding status (Bech et al. 1999; Nilsson and Ra˚berg 2001), and nutritional access (reviewed in Bozinovic and Sabat 2010). Metabolic rate is also affected by immune system activation, although the magnitude of the effect has not been established and the importance of this energetic cost is debated (Sheldon and Verhulst 1996; Martin et al. 2003; Hasselquist and Nilsson 2012). Activation of the immune system has been documented to cause an increase in metabolic rate in some bird species (e.g., great tits [Parus major] and North American house finches [Carpodacus mexicanus]; Ots et al. 2001; Nilsson et al. 2007; Hawley et al. 2012), whereas in other bird species (e.g., tree sparrows [Passer montanus] and ruffs [Philomachus pugnax]) it has been documented to cause a decrease in metabolic rate (Lee et al. 2005; Mendes et al. 2006). Moreover, even studies done on the same species have yielded conflicting results. Martin et al. (2003) found that immunization with phytohemagglutinin (PHA), which mainly stimulates the cellmediated immune response, increased the metabolic rate in house sparrows (Passer domesticus) by as much as 29%, whereas
Physiological and Behavioral Responses to Acute-Phase Response in Zebra Finches 289 Lee et al. (2005) found no effect on metabolic rate when immunizing house sparrows with PHA. The 29% increase in metabolic rate demonstrated by some house sparrows is equivalent to the energy required to produce half an egg in that species (Martin et al. 2003). Hence, evidence for the importance of the metabolic cost of immune responses is equivocal, although some of the differences between studies might be explained by seasonal differences in immunological responses or differences in infection history (Lee et al. 2005; Owen-Ashley and Wingfield 2007; Pap et al. 2010). In a recent study, Hasselquist and Nilsson (2012) presented a meta-analysis of recently published papers assessing the metabolic cost of an experimental immune system activation, where the results imply that activation of the immune system demands additional energy, although of a relatively small magnitude (5%–15%). The acute-phase response (APR) constitutes the first line of defense against invading pathogens (Hart 1988; Owen-Ashley and Wingfield 2007) and is initiated by circulating immune cells, such as macrophages, dendritic cells, and mast cells, that release proinflammatory cytokines (Tizard 2008), normally within minutes after exposure to the pathogen (Sorci and Faivre 2009). The APR is characterized by a cascade of events, including physiological, metabolic, and hormonal adjustments (Hart 1988; Owen-Ashley and Wingfield 2007). Vertebrates with an activated APR usually demonstrate an initial rapid onset of fever that peaks within a few hours after exposure (Maloney and Gray 1998; Harden et al. 2006). The febrile response is accompanied by other “sickness behaviors,” such as anorexia (i.e., reduced food intake), adipsia (i.e., reduced thirst), lethargy (i.e., reduced activity), and body mass loss (Hart 1988), which are commonly observed during the first 24-h period following the initiation of the APR (e.g., Deak et al. 2005; Adelman et al. 2010a; Burness et al. 2010). Because the APR includes several components that by themselves cause large effects on hosts, such as fever and inflammation, increased production and proliferation of immune cells, and reduced foraging activity (Hart 1988; Owen-Ashley and Wingfield 2007), it could be a particularly costly component of the immune system (Klasing 2004; Hasselquist and Nilsson 2012) that might necessitate trade-offs with other (energetically) costly physiological processes. In experimental studies of mammals and birds, activation of an APR is often triggered by injection with lipopolysaccharide (LPS), which is an immunogenic component of the cell wall of gram-negative bacteria (e.g., Nomoto 1996; Maloney and Gray 1998; Harden et al. 2006; Burness et al. 2010). Because LPS is an element of an infectious agent, hosts injected with LPS initially react as if they were exposed to a live, replicating pathogen (Ashley and Wingfield 2012). Accordingly, precocial birds experimentally immunized with LPS demonstrate the behavioral and physiological responses anticipated during a normal bacterial infection, including fever (Maloney and Gray 1998), decreased activity (Gregory et al. 2009), and increased metabolic rate (Marais et al. 2011). A handful of studies have examined the APR in passerines and found that they also respond to an LPS injection by decreasing food intake (OwenAshley et al. 2006; Coon et al. 2011), reducing activity (Owen-
Ashley et al. 2006; Adelman et al. 2010a; Burness et al. 2010), and increasing metabolic rate (Burness et al. 2010; Hegemann et al. 2012). However, the fever response in passerines is more equivocal; some studies have found the expected fever response (Coon et al. 2011; Nord et al. 2013), whereas other studies have found a short-term hypothermia, that is, a reduced body temperature (Owen-Ashley et al. 2006; Burness et al. 2010). The APR is an important part of the immune system, but it is also very dangerous for the host if not appropriately downregulated (Sorci and Faivre 2009), because, for example, it can induce immunopathology (Ra˚berg et al. 1998). Thus, the APR typically subsides within 24–48 h (Baumann and Gauldie 1994). Consequently, studies examining physiological, metabolic, and behavioral effects of an APR after injection with LPS have mainly focused on the first 48 h following exposure (e.g., Adelman et al. 2010a; Burness et al. 2010; Marais et al. 2011; Nord et al. 2013). However, because the APR can be physiologically costly in terms of body mass loss, decreased activity level, and increased metabolic rate (Bonneaud et al. 2003; Hegemann et al. 2012), it is possible that effects remain during the days subsequent to the expected time frame of an APR (i.e., 148 h after LPS injection). Activation of the immune system has previously been shown to induce carryover effects over relatively long time periods, for example, by reducing survival to the next breeding season (Hanssen et al. 2004). In this study, we wanted to investigate whether activation of the immune system also induced effects that were manifested during the days immediately after the expected time frame of the APR. We use the term “short-term carryover effect” for such delayed effects of an APR on physiology and behavior. Such short-term carryover effects are interesting to study, because they would mean that the cost of an APR might have been underestimated in previous studies and because these delayed physiological effects might entail fitness costs. In this study, we experimentally induced an APR in captive adult male zebra finches (Taeniopygia guttata) by intramuscular LPS injection and then measured metabolic rate, body mass, core body temperature, and activity levels during the 4 days following injection. In addition, because antibodies constitute an essential part of the immune defense of vertebrates (Abbas and Lichtman 2011), we wanted to investigate whether the concentrations of LPS-reactive antibodies and the total level of antibodies (immunoglobulin Y [IgY]) in the circulation change as a consequence of an LPS injection mimicking a bacterial infection. Ecological studies examining sickness behaviors during an APR have generally not measured antibody responses (e.g., Adelman et al. 2010a, 2010b; Burness et al. 2010; Coon et al. 2011), and antibody analyses would thus significantly add to our understanding of the dynamics of physiological responses connected to an APR. Our aims were to examine the effects of an APR on these physiological and behavioral parameters to see whether it induces any immediate (0–48 h after injection) or short-term carryover effects (148 h after injection) in the birds. To the best of our knowledge, this is the first attempt to assess both the immediate and short-term carryover effects of an APR on physiology and behavior in passerines.
290 S. Sko¨ld-Chiriac, A. Nord, J.-A˚. Nilsson, and D. Hasselquist Material and Methods Animal Husbandry A total of 27 adult male zebra finches from a captive population held in indoor facilities were used in this study, which was conducted during April and May 2009. Birds were divided into four batches, and all individuals within the same batch received the same immunological challenge (2 LPS, 2 control; see below). We chose this experimental batch setup to avoid any social interaction effects between individuals of different treatment groups held together in the communal cages. The batches were sequentially tested; that is, we conducted all measurements of one batch before the measurements of the next batch started (except for day 12 measurements, which were obtained after the next batch had been initiated). The test order of the batches was control, LPS, LPS, control. Different batches in the same treatment category did not differ in their response to the experimental treatment (t-tests between batches within a treatment; LPS-reactive antibodies: LPS: P p 0.43, control: P p 0.61; metabolic rate: LPS: P p 0.072, control: P p 0.082; body mass: LPS: P p 0.57, control: P p 0.10; activity: LPS: P p 0.23, control: P p 0.22), implying that batch and batch order did not affect the measurements. The close-to-significant differences between batches for metabolic rate seem to depend on a trend toward lower preinjection metabolic rate over the season. However, as we tested one control group and one LPS group in both the early and late periods of the study, any such seasonal effects should not have induced systematic effects that could have influenced our results and affected the conclusions of our study. Birds were housed together by treatment (n p 6–7) in cages (75 # 50 # 50 cm) under constant light conditions (14L : 10D) and temperature (15⬚ Ⳳ 2⬚C). This type of small-group housing is the normal situation for these zebra finches, and we wanted to keep the conditions as normal as possible. We provided commercial seed mixture, water, and cuttlefish bone ad lib. Birds in different treatments were visually but not acoustically separated. The experimental setup is illustrated in figure 1.
Metabolic Rate, Experimental Injections, and Core Body Temperature We measured the resting metabolic rate (RMR) of birds for a total of six nights by means of flow-through respirometry, according to the procedures described in Nord and Nilsson (2011). Briefly, birds were placed individually in 0.6-L hermetic metabolic chambers, which were positioned in a dark, temperature-controlled cabinet set at 32⬚C (i.e., within the thermoneutral zone of zebra finches; Calder 1964). The respirometer was equipped with eight identical channels, one of which was left empty for measurements of baseline air, and hence, seven birds could be measured simultaneously. Each recording cycle started with 13-min measurements of baseline oxygen and carbon dioxide levels (i.e., the blank channel), after which each metabolic chamber was measured serially for 13 min (separated by 2 min of air flushing), and hence, each measurement cycle lasted for 2 h. To ensure that measurements were fully comparable, individuals were placed in the same metabolic chamber during all nights. A metabolic measurement started between 9:00 and 10:00 p.m. and ended between 7:00 and 8:00 a.m., and hence, four cycles were possible each night. Oxygen concentration in ingoing air was calibrated against external air to 20.95%, and the flow rate within the system was set at 200 mL min⫺1. The first measurement was performed during the night before the experiment started. Immediately after RMR measurements were finished the following morning, birds were injected intramuscularly in the pectoral muscle with either 50 mL of LPS derived from Escherichia coli (Sigma-Aldrich, catalog no. L-2880) diluted in phosphate-buffered saline (PBS) to a concentration of 0.1 mg kg⫺1 (based on the mean body mass of the birds at the start of the experiment; n p 14) or 50 mL of PBS (n p 13). The concentration of LPS was similar to, or lower than, concentrations used in other studies on passerines (Grindstaff et al. 2006; Burness et al. 2010; Nord et al. 2013). The strength of the response to LPS exposure might depend on route of injection (Ashley and Wingfield 2012) and dose
Figure 1. Schematic representation of the experimental setup, illustrating the timing of measurements of resting metabolic rate (RMR), experimental injections with either bacterial endotoxin (lipopolysaccharide) or phosphate-buffered saline, blood sampling, and behavioral recordings. We recorded birds’ body mass before and after each RMR measurement and measured their body temperature immediately after removing them from the metabolic chambers in the mornings.
Physiological and Behavioral Responses to Acute-Phase Response in Zebra Finches 291 (Maloney and Gray 1998; Koutsos and Klasing 2001). Hence, we performed a test study before the experiment, in which we measured the response to intramuscular injection with different LPS doses (0.001–1.0 mg LPS kg⫺1) in male zebra finches, and found that 0.1 mg LPS kg⫺1 triggered the physiological responses expected to occur during an APR, that is, changed body temperature, decreased food intake, and decreased daily body mass gain (S. Sko¨ld-Chiriac, A. Nord, M. Tobler, J.-A˚. Nilsson, and D. Hasselquist, unpublished data). We then continued to measure RMR for an additional four consecutive nights. The final RMR measurements were performed 12 nights after immunization (except one batch of control birds that for logistical reasons could not be measured at this occasion). We weighed birds to the nearest 0.1 g before and after each respirometer session. In addition, we measured core body temperature of the birds within 1 min after removing them from the metabolic chambers in the morning, using a Testo 925 digital thermometer (Testo, Lenzkirch, Germany) with a standard type K thermocouple (diameter, 0.9 mm; ELFA, Ja¨rfa¨lla, Sweden) inserted 12 mm into the cloaca. Further insertion did not alter the temperature reading (A. Nord, unpublished data). We obtained three body temperature readings within 30 s (intersample repeatability: r p 0.96, F152, 306 p 81.92, P ! 0.001; Lessells and Boag 1987) and used the average value in all subsequent analyses. Following the first (i.e., before experimental injections) and fifth (i.e., 4 d after immunization) metabolic measurements, we collected a blood sample from the jugular vein (80 mL) for analyses of antibody titers (see below). Blood was kept on ice until it was centrifuged at 3,000 rpm for 10 min. Plasma was then separated from red blood cells with a Hamilton syringe and kept at ⫺20⬚C until further analyses. Behavioral Analyses Behavior of the birds was recorded two times for 47 min each, starting 24 and 72 h after immunization, with a digital camera (Canon Ixus 65). Video recordings were subsequently transferred to a computer, and behavior was analyzed manually. All birds in each batch were filmed simultaneously in their cage. When the films were analyzed for movements, each bird was followed individually throughout the whole filmed sequence. Hence, each video recording was watched the same number of times as the number of birds in the cage, that is, 6 or 7 times per film. This was necessary to obtain reliable movement scorings for each bird. We excluded 1 min at the beginning of each recording from the analyses, to avoid any potential disturbance effects from setting up the camera device. We choose to exclude only 1 min because the birds started to move around in the cage immediately after the person setting up the camera left the room. When analyzing the films, we divided the cage into 10 sections (the bottom area divided into two halves, each of the tree perches divided into two halves, and the two food holders) and noted the location (i.e., which section) of the individual bird every 30 s. Birds started with a score of 0 and were given a score of 1 for each movement between sections.
Hence, birds obtained a total score that could range between 0 (no movement) and 92 (changed section during all time periods). Because of equipment problems, video recording for one batch of LPS-immunized birds is available for only 10.5 min during the second recording session. This video recording was included when performing the activity analysis of day 4 because excluding this batch from the analysis did not alter the results (P p 0.78).
Quantification of Humoral Immunity Humoral immunity of birds was quantified by means of enzyme-linked immunosorbent assay (ELISA), according to procedures in Grindstaff et al. (2006). We assessed both the total level of antibodies (IgY) and the level of LPS-reactive antibodies in plasma. Briefly, 96-well ELISA plates for analyses of total IgY were coated with 100 mL of goat-anti-bird immunoglobulin G (IgG; Novus Biologicals, catalog no. NB7226) diluted in carbonate buffer (0.15 M, pH 9.6) to a concentration of 0.4 mg mL⫺1. Plates for analyses of LPS-reactive antibodies were coated with 100 mL of LPS diluted in carbonate buffer to a concentration of 5 mg mL⫺1. Following overnight incubation at 4⬚C, plates were blocked at room temperature for at least 2 h (to prevent nonspecific binding of antibodies to the plate) using 3% milk powder solution diluted in 0.01 M PBS and Tween 20 (PBS/Tween 20). Then, 100-mL plasma samples diluted in diluent (1% powdered milk in PBS/Tween 20) were added to each well in duplicates. Samples for analyses of total IgY were diluted 1 : 12,000, whereas samples for analyses of LPS-reactive antibodies were diluted 1 : 600 (optimal dilutions were tested before analyses). All plates also included two blanks and a series-diluted standard from a pool of plasma from LPSinjected zebra finches that covered the range of antibody concentrations in the unknown samples. Plates were incubated overnight at 4⬚C. The following day, 100 mL of rabbit-anti–redwinged blackbird IgG diluted 1 : 1,000 in diluent was added to all wells. This secondary antibody has been confirmed to detect antibodies in zebra finches (Tobler et al. 2010). Plates were then incubated for 1 h at 37⬚C. Afterward, 100 mL of peroxidaselabeled goat-anti-rabbit serum antibody (Sigma-Aldrich, catalog no. A6154) diluted 1 : 2,000 in diluent was added, and plates were incubated for another 30 min at 37⬚C. Between all steps, washing was conducted for 3 # 10 s during shaking in PBS/Tween 20 in an ELISA washer (BioTek ELx50). Then, 100 mL of peroxidase substrate (2,2-azino-bis-3-ethylbenzothiazoline-6-sulfonic acid [ABTS]; Sigma-Aldrich, catalog no. A1888) and peroxidase diluted in citrate buffer (pH 4.0) was added to all wells. Plates were then immediately read in a kinetic ELISA reader (BioTek EL 808) with a 405-nm wavelength filter at 30s intervals for 14 min. Antibody concentrations are presented as the slopes of color change of the substrate over time (measured in 10⫺3 optical density per minute [mOD min⫺1]), with a steeper slope representing a higher concentration of antibodies in plasma. Antibody levels are calculated as the mean of the duplicates minus the mean value of the blanks and
292 S. Sko¨ld-Chiriac, A. Nord, J.-A˚. Nilsson, and D. Hasselquist transformed according to the standard series to correct for variation between plates. Data Analysis and Statistical Methods Oxygen consumption (mL O2 min⫺1) was defined as the difference between oxygen concentrations in effluent sample air and the reference air in the baseline channel and was calculated according to equation C in Hill (1972) with ExpeData 1.1.9 for Windows. Data were manually inspected for drift or inconsistencies before analyses to confirm that oxygen consumption was consistent during the entire measurement period. The value of oxygen expenditure used in the analyses was defined as the single lowest value from 7-min running averages during a measurement session. Data for oxygen consumption were subsequently transformed to metabolic rates (kJ d⫺1), assuming an energy equivalent of 20 J (mL O2)⫺1. Two individuals in the control group had very high initial levels of LPS-reactive antibodies (16.4 and 21.5, respectively, both 12 SD). They were excluded from further analyses because of a potential progressing infection that could interfere with experimental effects. All statistical tests were performed with JMP, edition 9.0 for Microsoft. If necessary, variables were log transformed before analysis for the model to meet the criteria of normality (LPS and IgY plasma titer). The level of LPS-reactive antibodies and the total level of IgY antibodies in the different treatments were compared by means of a linear mixed model, with antibody titer as the dependent variable, treatment and day (and their interaction) as explanatory factors, and bird identity as a random factor. RMR before experimental manipulation and on the twelfth night after manipulation was analyzed with an ANCOVA, with RMR as the dependent variable, treatment as an explanatory factor, and morning body mass as a covariate (because morning body mass was closest in time to RMR readings; see Nord and Nilsson 2011). RMR during the four nights following immunization was analyzed in a linear mixed model using nighttime differences between RMR before and after immunization as the dependent variable, treatment, day (and their interaction), and respirometer channel as explanatory factors, morning body mass as a covariate, and bird identity as a random factor. Body mass changes were assessed in two ways, namely, body mass gain during days (days 1–4) and body mass loss during nights (nights 0–4). Body mass gain was analyzed with a linear mixed model, with evening body mass as the dependent variable, treatment and day (and their interaction) as explanatory factors, morning body mass as a covariate, and bird identity as a random factor. A similar analysis was performed to assess nocturnal body mass loss, but in that analysis, morning body mass was used as the dependent variable and evening body mass as a covariate. Core body temperature changes were assessed with a linear mixed model, with body temperature as the dependent variable, treatment, day (and their interaction), and measurement order (i.e., the order in which they were removed from the respirometer; this could be important because of stress effects on body temperature) as
explanatory factors, and bird identity as a random factor. Final models were derived by stepwise backward elimination of nonsignificant variables (P 1 0.05) until only significant terms remained. When the interaction between the experimental treatment and day was significant, we compared differences between treatments, using independent t-tests with P values adjusted for multiple comparisons with a sequential Bonferroni correction (Holm 1979). Differences in bird activity between treatments were tested separately for each day via independent ttests with Bonferroni-corrected P values. The activity analysis includes the two control individuals with high initial levels of LPS-reactive antibodies, because it was not possible to determine bird identity in the video, as they were not marked with color rings. However, if anything, the inclusion of these two control individuals should have reduced the difference between treatments and thus strengthened our results. All results are presented as means with standard errors (mean Ⳳ SE), and all significances are two-tailed. Results Humoral Immune Responses At the time of immunization, there was no difference between treatments in either LPS-reactive antibody titers (table 1; fig. 2) or the total level of IgY (LPS: 55.51 Ⳳ 1.49; control: 54.85 Ⳳ 1.55; P p 0.68). On day 5 (96–110 h after injection), the LPSimmunized birds had increased their level of LPS-reactive antibodies, resulting in LPS-injected birds having 49% higher level of LPS-reactive antibodies than control individuals (table 1; fig. 2). However, there was no corresponding effect on the level of total IgY on day 5 after immunization (LPS: 54.81 Ⳳ 1.54; control: 53.67 Ⳳ 1.11; P p 0.76). Resting Metabolic Rate There was no difference in RMR between treatments before experimental manipulation (LPS: 24.63 Ⳳ 1.31; control: 23.39 Ⳳ 1.07; P p 0.21). However, the change in RMR differed significantly over time between treatment groups following injection (table 1; fig. 3). Comparison between treatments for each day separately revealed that RMR differed significantly between treatments on night 4 (86–96 h after injection), when LPSinjected individuals had an RMR 7% lower than that before immunization, whereas RMR in control individuals was unchanged relative to that before experimental injection (P p 0.0040). There was no difference in RMR between treatments 12 nights after experimental manipulation (LPS: 24.60 Ⳳ 1.17; control: 22.72 Ⳳ 1.58; P p 0.27). Body Mass Changes Body mass did not differ between treatments when the experiment started (LPS: 17.17 Ⳳ 0.90 g; control: 17.00 Ⳳ 0.66 g; P p 0.88). However, body mass gain during the days differed significantly over time between treatment groups (table 1; fig. 4). Specifically, when treatments were compared for each day
Physiological and Behavioral Responses to Acute-Phase Response in Zebra Finches 293 Table 1: Final models of statistical tests, presented with their F values, degrees of freedom, and resultant P values Parameter LPS-specific antibodies: Treatment Day Treatment # day Resting metabolic rate (nights 1–4): Treatment Day Treatment # day Body mass gain (days 1–4): Morning body mass Treatment Day Treatment # day Mass loss during nights (nights 0–4): Night Evening body mass Core body temperature (days 1–5): Day Activity (day 2): Treatment
F
df
P
1.25 .91 6.09
1, 23 1, 23 1, 23
.28 .35 .021
1.93 .70 2.83
1, 22.5 3, 67.6 3, 67.6
.18 .55 .045
3,110 .0024 1.93 3.70
1, 1, 3, 3,
21.6 20.9 67.1 67.1
!.001
11.08 9,050
4, 94.0 1, 23.5
!.001 !.001
6.07
4, 94.0
!.001
10.1
1, 25
.96 .13 .016
.0039
Note. Models for total immunoglobulin Y, activity on day 4, and resting metabolic rate on day 12 are not presented because none of the variables in our original models could explain any significant variation in the dependent variable. The number of days varies between models because data were collected during five consecutive nights but only four days. Boldface indicates significant P values (P ! 0.05). LPS p lipopolysaccharide.
separately, individuals injected with LPS gained 27% less body mass than controls on the day after the experimental manipulation, that is, during the APR (day 2; 24–38 h after injection; P p 0.012). Moreover, LPS-injected birds may have shown a tendency to compensate for the attenuated body mass gain, because on the subsequent day (day 3; 48–62 h after injection), they gained 20% more in body mass than the control birds, but this difference was not significant (P p 0.12). Although body mass loss over the nights varied significantly over time, this variation was not related to the experimental manipulation (table 1). Core Body Temperature Body temperature in the mornings increased over time (table 1) but was not affected by experimental manipulation (P p 0.41). Activity The activity pattern differed markedly between treatments on day 2 (24–38 h after injection), that is, during the APR, when individuals injected with LPS were 40% less active than control birds (table 1; fig. 5). However, there was no difference in activity patterns between treatments on day 4 (72–86 h after injection; P p 0.28).
Discussion By experimentally injecting male zebra finches with LPS to mimic a bacterial infection, we demonstrated that zebra finches displayed some of the general sickness behaviors that are typical of a bacterial infection within the expected time frame of an APR (within 48 h; Baumann and Gauldie 1994; Owen-Ashley and Wingfield 2007). Hence, within 48 h after LPS injection, the birds showed reduced activity and lower body mass gain, compared to controls, but no effect on RMR. These responses could facilitate pathogen elimination during a natural bacterial infection by minimizing energy spent on foraging while simultaneously reducing micronutrient availability for the pathogens (Hart 1988). While behavioral and physiological responses were expressed within the first 48 h following experimental manipulation, the zebra finches injected with LPS also demonstrated a decrease in RMR on night 4 (86–96 h after injection). Thus, our results imply that besides the immediate effects following LPS immunization in passerines, there also appear to be short-term delayed effects on RMR that act several days after the expected time frame of an APR. In addition, we found that there was an increase in the level of circulating LPSreactive antibodies on day 5 (96–110 h after injection) but no effect on total level of circulating antibodies (total IgY). These physiological short-term changes might be a natural part of the progression of the APR, which is probably true for the LPS-
294 S. Sko¨ld-Chiriac, A. Nord, J.-A˚. Nilsson, and D. Hasselquist relatively unexplored area in passerines, but our study corroborates a recent study suggesting that birds can show an immunological response, producing more LPS-reactive antibodies, in response to an LPS challenge (Toomey et al. 2010). However, whether this is a compensatory production of nonspecific but LPS-reactive antibodies or whether it involves activation of the presumably very limited number of LPS-specific B-cell clones that could induce a more normal primary (and secondary) antibody response remains to be studied. That it was a secondary antibody response is, however, unlikely because the magnitude of the response was too small to be a “true” secondary antibody response, which normally involves a 35– 100-fold or even greater increase above preimmunization antibody levels in passerine birds (Svensson et al. 1998; Hasselquist et al. 1999).
RMR and Body Temperature Figure 2. Mean (Ⳳ SE) levels of lipopolysaccharide (LPS)-reactive antibodies in the plasma of zebra finches immediately before (day 1) and 4 d after (day 5) immunization with either bacterial endotoxin (LPS; n p 14) or phosphate-buffered saline (control; n p 11). The asterisk denotes a significant difference between treatments over sampling occasions (P ! 0.05). mOD min⫺1 p 10⫺3 optical density per minute.
reactive antibody response, in which case data suggest that an APR proceeds for a longer time period than 48 h in passerines. Alternatively, our data might indicate that there are short-term carryover effects expressed during the phase of recovery from the APR, and in our view, this is the most likely explanation for the higher RMR in LPS-injected birds on night 4 after the injection. In either case, our study shows that in passerines, the physiology might be affected for at least 4–5 days after an LPS immunization.
Activation of the immune system is often associated with an increase in metabolic rate, as has been demonstrated during an APR (Burness et al. 2010; Marais et al. 2011; Hegemann et al. 2012) as well as during humoral (Ots et al. 2001) and cellmediated immune responses (Martin et al. 2003; Nilsson et al. 2007). A recent meta-analysis conducted on birds found a pattern of increased metabolic rate during immune system activation, but of the rather limited magnitude of 5%–15% (Hasselquist and Nilsson 2012). In our study, by contrast, we did not find any effects on RMR during the expected time course of the APR. A possible explanation for the absence of a metabolic cost during the APR is that we did not detect any fever response in the birds in our study. Fever is generally considered an integral part of the APR (Hart 1988; Owen-Ashley and Wingfield 2007) in both mammals (Harden et al. 2006) and birds (Maloney and Gray 1998; Gray et al. 2013) and has been associated with an increase in metabolic cost (Marais et al. 2011). However, the febrile response to LPS immunization is
Immunologic Response Besides activating the innate immune system, LPS also activates a broad range of B- and T-cell types in a nonspecific way by attaching to general LPS-binding receptors (Klasing 2004; Sorci and Faivre 2009). This LPS activation of B-cells is different from responses to most other antigens, because it does not involve immunological memory and hence does not induce a true secondary antibody response (Abbas and Lichtman 2011). Nevertheless, we found significantly higher levels of LPSreactive antibodies in the experimental group on day 5 (96– 110 h after injection). Here it is important to note that we measured antibody levels only before injection and on day 5, so we do not know the time frame of the upregulation of these LPS-reactive antibodies. However, LPS immunization did not affect the level of total IgY, suggesting that it was increased production of LPS-reactive antibodies, rather than a general increase in the total number of antibodies in the circulation, that had occurred. To the best of our knowledge, the development of specific and general antibodies during an APR is a
Figure 3. Mean (Ⳳ SE) change in resting metabolic rate (RMR) relative to preexperimental RMR in zebra finches during four days following immunization with either bacterial endotoxin (lipopolysaccharide [LPS]; n p 14) or phosphate-buffered saline (control; n p 11). Experimental injections were carried out in the morning of day 1. An asterisk denotes a significant difference between treatments within a night (P ! 0.05), and “NS” represents nonsignificant differences between treatments within a night (P 1 0.05).
Physiological and Behavioral Responses to Acute-Phase Response in Zebra Finches 295 system amplifies the production of reactive oxygen species (ROSs) and reactive nitrogen species (RNSs; Sorci and Faivre 2009) as well as increasing the level of stress hormones (Ra˚berg et al. 1998). ROSs and RNSs cannot distinguish between host and pathogen molecules and can thus be harmful to the host if present in large amounts for extended time periods (Sorci and Faivre 2009). A decrease in metabolically costly functions during an immunologic challenge, which will cause a decline in RMR and thereby lower ROS and RNS production, can thus be an adaptive modification to reduce the total amount of potentially harmful molecules (von Schantz et al. 1999) and can consequently reduce the risk of immunopathology (Ra˚berg et al. 1998). Figure 4. Mean (Ⳳ SE) body mass gain during the days in zebra finches immunized with either bacterial endotoxin (lipopolysaccharide [LPS]; n p 14) or phosphate-buffered saline (control; n p 11). Experimental injections were carried out in the morning of day 1. An asterisk denotes a significant difference between treatments within a day (P ! 0.05), and “NS” represents nonsignificant differences between treatments within a day (P 1 0.05).
typically of short duration and often occurs within hours after immunization (Nomoto 1996; Maloney and Gray 1998; Harden et al. 2006; Nord et al. 2013; but see Hegemann et al. 2012). Thus, because our first body temperature measurement was conducted 24 h after immunization, we might have failed to detect any early change in body temperature. Nevertheless, some recent studies examining the APR in passerines have also failed to demonstrate the induction of fever and instead found a reduction in body temperature a few hours after LPS immunization (Owen-Ashley et al. 2006; Burness et al. 2010). Furthermore, body temperature response to LPS immunization might depend on LPS dose (Maloney and Gray 1998). However, some of the studies that found a body temperature decrease in passerines (Owen-Ashley et al. 2006; Burness et al. 2010) used a higher dose to trigger the APR (1 mg LPS kg⫺1), whereas we, in a previous study, found a pronounced fever response in great tits (Parus major) exposed to the same LPS dose as the zebra finches in this study (Nord et al. 2013). These findings, together with our test study evaluating different LPS doses before this experiment, indicate that the absence of a fever response in zebra finches in this study is unlikely to be explained by administration of a too-low LPS dose. The LPS-injected zebra finches in our study did not change RMR during the expected time frame of the APR, but they decreased RMR by 7% during night 4 (86–96 h after LPS injection). There are two possible explanations for why the metabolic rate did not increase during the APR and was even lower than normal during the immunologic response. First, a downregulated metabolic rate in response to an immune system challenge has been explained as a potential energy-saving mechanism (Gutierrez et al. 2011). For example, Mendes et al. (2006) argued that ruffs (Philomachus pugnax) experiencing a 15% decline in metabolic rate following a cell-mediated immune response might have reallocated energy from costly physiological functions to the immune system to such an extent that RMR actually decreased. Second, activation of the immune
Body Mass Changes and Activity LPS-injected birds were 40% less active than control birds 24 h after immunization and gained 27% less body mass during the day, compared to control birds, on the day after immunization (day 2, 24–38 h after injection), which is in line with findings from other studies of passerine birds (Adelman et al. 2010a, 2010b; Burness et al. 2010). Our results indicate that the anorexia symptoms had ceased before day 3 (i.e., within 48 h after LPS injection) and that reduced activity in the LPSimmunized birds was no longer evident on day 4 (72–86 h after LPS injection). However, note that we did not measure activity on day 3, and hence we cannot exclude the possibility that the effects on activity had already subsided by this time. In other studies of passerines, lethargy and anorexia have been observed within 48 h after triggering an APR, but these studies did not investigate when these behaviors subsided (Owen-Ashley et al. 2006; Adelman et al. 2010a; Burness et al. 2010; Coon et al. 2011). Anorexia and reduced activity during pathogen infection are host strategies assumed to diminish access of micronutrients essential for pathogen replication and survival while simultaneously minimizing the energy spent on other
Figure 5. Mean (Ⳳ SE) activity score in male zebra finches 24 h (day 2) and 74 h (day 4) after immunization with either bacterial endotoxin (lipopolysaccharide [LPS]; n p 14) or phosphate-buffered saline (control; n p 13). Data are presented as movements per minute between different sections in the cage, as described in the text. An asterisk denotes a significant difference between treatments within a day (P ! 0.05), and “NS” represents nonsignificant differences between treatments within a day (P 1 0.05).
296 S. Sko¨ld-Chiriac, A. Nord, J.-A˚. Nilsson, and D. Hasselquist physiologically demanding behaviors, such as foraging and movement (Hart 1988). The demonstrated reduction in activity and body mass gain (which presumably is a result of reduced food intake, i.e., anorexia) might reflect adaptive behavioral and physiological changes to minimize negative effects on hosts caused by a real pathogen infection (Hart 1988). However, this inactive state can probably be maintained for only a short period of time without seriously threatening the host’s health and survival, which might explain why we did not detect any effects on activity on day 4 (72–86 h after LPS injection).
Conclusions We have demonstrated that male zebra finches respond to LPS injection with lethargy and reduced body mass gain (i.e., anorexia) within the first 48 h after immunization and with increased production of LPS-reactive antibodies and reduced RMR 4–5 d (86–110 h) after the initiation of the APR. Although there is convincing evidence that immune system activation can have substantial fitness costs (Bonneaud et al. 2003; Hanssen et al. 2004; Eraud et al. 2009; reviewed in Hasselquist and Nilsson 2012), there is still much debate over the proximate mechanisms that may mediate such costs (Sheldon and Verhulst 1996; Ra˚berg et al. 1998, 2003; Svensson et al. 1998; Martin et al. 2003; Hasselquist and Nilsson 2012). Our study supports the generality of sickness behaviors as an immediate (!48 h) response to a bacterial challenge and hence a part of the APR. However, our study is the first to demonstrate that the physiological effects of an APR might proceed for several days longer than expected and that besides the immediate expression of sickness behaviors, there are also other physiological effects occurring more than 48 h after the triggering of an APR. At least in the case of the reduction in RMR, this seems to represent a short-term carryover effect of the APR, because it only occurred 86–96 h after LPS injection, which is several days after the expected time frame of an APR. To what extent these proximate physiological modifications might be translated into fitness costs remain to be investigated. Acknowledgments This study was supported by grants from the Swedish Research Council (to D.H. and J.-A˚.N.), the Lund Animal Protection Foundation (to A.N.), and the Research Excellence Center CAnMove at Lund University (funded by a Linnaeus grant from the Swedish Research Council and Lund University). The experimental design followed Swedish legislation and was approved by the Malmo¨/Lund Ethical Committee on Animal Research. Literature Cited Abbas A.K. and A.H. Lichtman. 2011. Basic immunology: functions and disorders of the immune system. Updated ed. Saunders-Elsevier, Philadelphia.
Adelman J.S., G.E. Bentley, J.C. Wingfield, L.B. Martin, and M. Hau. 2010a. Population differences in fever and sickness behaviors in a wild passerine: a role for cytokines. J Exp Biol 213:4099–4109. Adelman J.S., S. Co´rdoba-Co´rdoba, K. Spoelstra, M. Wikelski, and M. Hau. 2010b. Radiotelemetry reveals variation in fever and sickness behaviours with latitude in a free-living passerine. Funct Ecol 24:813–823. Ashley N.T. and J.C. Wingfield. 2012. Sickness behavior in vertebrates. Pp. 45–91 in G.E. Demas and R.J. Nelson, eds. Ecoimmunology. Oxford University Press, New York. Baumann H. and J. Gauldie. 1994. The acute phase response. Immunol Today 15:74–80. Bech C., I. Langseth, and G.W. Gabrielsen. 1999. Repeatability of basal metabolism in breeding female kittiwakes Rissa tridactyla. Proc R Soc B 266:2161–2167. Bonneaud C., J. Mazuc, G. Gonzalez, C. Haussy, O. Chastel, B. Faivre, and G. Sorci. 2003. Assessing the cost of mounting an immune response. Am Nat 161:367–379. Bozinovic F. and P. Sabat. 2010. On the intraspecific variability in basal metabolism and the food habits hypothesis in birds. Curr Zool 56:759–766. Broggi J., E. Hohtola, K. Koivula, M. Orell, and J.-A˚. Nilsson. 2009. Long-term repeatability of winter basal metabolic rate and mass in a wild passerine. Funct Ecol 23:768–773. Burness G., C. Armstrong, T. Fee, and E. Tilman-Schindel. 2010. Is there an energetic-based trade-off between thermoregulation and the acute phase response in zebra finches? J Exp Biol 213:1386–1394. Calder W.A. 1964. Gaseous metabolism and water relations of the zebra finch, Taeniopygia castanotis. Physiol Zool 37:400– 413. Coon C.A.C., R.W. Warne, and L.B. Martin. 2011. Acute-phase responses vary with pathogen identity in house sparrows (Passer domesticus). Am J Physiol Regul Integr Comp Physiol 300:R1418–R1425. Deak T., C. Bellamy, and K.A. Bordner. 2005. Protracted increases in core body temperature and interleukin-1 following acute administration of lipopolysaccharide: implications for the stress response. Physiol Behav 85:296–307. Eraud C., A. Jacquet, and B. Faivre. 2009. Survival cost of an early immune soliciting in nature. Evolution 63:1036–1043. Gray D.A., M. Marais, and S.K. Maloney. 2013. A review of the physiology of fever in birds. J Comp Physiol B 183:297– 312. Gregory N.G., S.R. Payne, C.D. Devine, and C.J. Cook. 2009. Effect of lipopolysaccharide on sickness behaviour in hens kept in cage and free range environments. Res Vet Sci 87: 167–170. Grindstaff J.L., D. Hasselquist, J.-A˚. Nilsson, M. Sandell, H.G. Smith, and M. Stjernman. 2006. Transgenerational priming of immunity: maternal exposure to a bacterial antigen enhances offspring humoral immunity. Proc R Soc B 273:2551– 2557. Gutierrez J.S., J.A. Masero, J.M. Abad-Gomez, A. Villegas, and J.M. Sanchez-Guzman. 2011. Metabolic consequences of
Physiological and Behavioral Responses to Acute-Phase Response in Zebra Finches 297 overlapping food restriction and cell-mediated immune response in a long-distance migratory shorebird, the little ringed plover Charadrius dubius. J Avian Biol 42:259–265. Hanssen S.A., D. Hasselquist, I. Folstad, and K.E. Erikstad. 2004. Costs of immunity: immune responsiveness reduces survival in a vertebrate. Proc R Soc B 271:925–930. Harden L.M., I. du Plessis, S. Poole, and H.P. Laburn. 2006. Interleukin-6 and leptin mediate lipopolysaccharide-induced fever and sickness behavior. Physiol Behav 89:146–155. Hart B.L. 1988. Biological basis of the behavior of sick animals. Neurosci Biobehav Rev 12:123–138. Hasselquist D., J.A. Marsh, P.W. Sherman, and J.C. Wingfield. 1999. Is avian humoral immunocompetence suppressed by testosterone? Behav Ecol Sociobiol 45:167–175. Hasselquist, D. and J.-A˚. Nilsson. 2012. Physiological mechanisms mediating costs of immune responses: what can we learn from studies of birds? Anim Behav 83:1303–1312. Hawley D.M., S.E. DuRant, A.F. Wilson, J.S. Adelman, and W.A. Hopkins. 2012. Additive metabolic costs of thermoregulation and pathogen infection. Funct Ecol 26:701–710. Hegemann, A., K.D. Matson, M.A. Versteegh, and B.I. Tieleman. 2012. Wild skylarks seasonally modulate energy budgets but maintain energetically costly inflammatory immune responses throughout the annual cycle. PLoS ONE 7:e36358. Hill R.W. 1972. Determination of oxygen consumption by use of the paramagnetic oxygen analyzer. J Appl Physiol 33:261– 263. Holm S. 1979. A simple sequentially rejective multiple test procedure. Scand J Stat 6:65–70. Ilmonen P., T. Taarna, and D. Hasselquist. 2000. Experimentally activated immune defense in female pied flycatchers results in reduced breeding success. Proc R Soc B 267:665–670. Klasing K.C. 2004. The costs of immunity. Acta Zool Sin 50: 961–969. Koutsos E.A. and K.C. Klasing. 2001. The acute phase response in Japanese quail (Coturnix coturnix japonica). Comp Biochem Physiol C 128:255–263. Lee K.A., L.B. Martin II, and M.C. Wikelski. 2005. Responding to inflammatory challenges is less costly for a successful avian invader, the house sparrow (Passer domesticus), than its lessinvasive congener. Oecologia 145:244–251. Lessells C.M. and P.T. Boag. 1987. Unrepeatable repeatabilities: a common mistake. Auk 104:116–121. Lochmiller R.L. and C. Deerenberg. 2000. Trade-offs in evolutionary immunology: just what is the cost of immunity? Oikos 88:87–98. Maloney S.K. and D.A. Gray. 1998. Characteristics of the febrile response in Pekin ducks. J Comp Physiol B 168:177–182. Marais M., S.K. Maloney, and D.A. Gray. 2011. The metabolic cost of fever in Pekin ducks. J Therm Biol 36:116–120. Martin L.B., D.M. Hawley, and D.R. Ardia. 2011. An introduction to ecological immunology. Funct Ecol 25:1–4. Martin L.B., A. Scheuerlein, and M. Wikelski. 2003. Immune activity elevates energy expenditure of house sparrows: a link between direct and indirect costs? Proc R Soc B 270:153– 158.
Marzal A., M. Reviriego, F. de Lope, and A.P. Møller. 2007. Fitness costs of an immune response in the house martin (Delichon urbica). Behav Ecol Sociobiol 61:1573–1580. McKechnie A.E. 2008. Phenotypic flexibility in basal metabolic rate and the changing view of avian physiological diversity: a review. J Comp Physiol B 178:235–247. Mendes L., T. Piersma, D. Piersma, and D. Hasselquist. 2006. Two estimates of the metabolic costs of antibody production in migratory shorebirds: low costs, internal reallocation, or both? J Ornithol 147:274–280. Moe B., B. Rønning, S. Verhulst, and C. Bech. 2009. Metabolic ageing in individual zebra finches. Biol Lett 5:86–89. Nilsson J.-A˚. 2003. Ectoparasitism in marsh tits: costs and functional explanations. Behav Ecol 14:175–181. Nilsson J.-A˚., M. A˚kesson, and J.F. Nilsson. 2009. Heritability of resting metabolic rate in a wild population of blue tits. J Evol Biol 22:1867–1874. Nilsson J.-A˚., M. Granbom, and L. Ra˚berg. 2007. Does the strength of an immune response reflect its energetic cost? J Avian Biol 38:488–494. Nilsson J.-A˚. and L. Ra˚berg. 2001. The resting metabolic cost of egg laying and nestling feeding in great tits. Oecologia 128:187–192. Nomoto S. 1996. Diurnal variations in fever induced by intravenous LPS injection in pigeons. Pfluegers Arch Eur J Physiol 431:987–989. Nord A., S. Chiriac, D. Hasselquist, and J.-A˚. Nilsson. 2013. Endotoxin injection attenuates rest-phase hypothermia in wintering great tits through the onset of fever. Funct Ecol 27:236–244. Nord A. and J.-A˚. Nilsson. 2011. Incubation temperature affects growth and energy metabolism in blue tit nestlings. Am Nat 178:639–651. Norris K. and M.R. Evans. 2000. Ecological immunology: life history trade-offs and immune defense in birds. Behav Ecol 11:19–26. Ots I., A.B. Kerimov, E.V. Ivankina, T.A. Ilyina, and P. Ho˜rak. 2001. Immune challenge affects basal metabolic activity in wintering great tits. Proc R Soc B 268:1175–1181. Owen-Ashley N.T., M. Turner, T.P. Hahn, and J.C. Wingfield. 2006. Hormonal, behavioral, and thermoregulatory responses to bacterial lipopolysaccharide in captive and freeliving white-crowned sparrows (Zonotrichia leucophrys gambelii). Horm Behav 49:15–29. Owen-Ashley N.T. and J.C. Wingfield. 2007. Acute phase responses of passerine birds: characterization and seasonal variation. J Ornithol 148:583–591. Pap P.L., G.A. Czirjak, C.I. Vagasi, Z. Barta, and D. Hasselquist. 2010. Sexual dimorphism in immune function changes during the annual cycle in house sparrows. Naturwissenschaften 97:891–901. Piersma T., N. Cadee, and S. Daan. 1995. Seasonality in basal metabolic rate and thermal conductance in a long-distance migrant shorebird, the knot (Calidris canutus). J Comp Physiol B 165:37–45. Ra˚berg L., M. Grahn, D. Hasselquist, and E. Svensson. 1998.
298 S. Sko¨ld-Chiriac, A. Nord, J.-A˚. Nilsson, and D. Hasselquist On the adaptive significance of stress-induced immunosuppression. Proc R Soc B 265:1637–1641. Ra˚berg L., J.-A˚. Nilsson, P. Ilmonen, M. Stjernman, and D. Hasselquist. 2000. The cost of an immune response: vaccination reduces parental effort. Ecol Lett 3:382–386. Ra˚berg L., M. Stjernman, and D. Hasselquist. 2003. Immune responsiveness in adult blue tits: heritability and effects of nutritional status during ontogeny. Oecologia 136:360–364. Sheldon B.C. and S. Verhulst. 1996. Ecological immunology: costly parasite defences and trade-offs in evolutionary ecology. Trends Ecol Evol 11:317–321. Soler J.J., L. De Neve, T. Perez-Contreras, M. Soler, and G. Sorci. 2003. Trade-off between immunocompetence and growth in magpies: an experimental study. Proc R Soc B 270: 241–248. Sorci G. and B. Faivre. 2009. Inflammation and oxidative stress in vertebrate host-parasite systems. Philos Trans R Soc B 364: 71–83. Svensson E., L. Ra˚berg, C. Koch, and D. Hasselquist. 1998.
Energetic stress, immunosuppression and the costs of an antibody response. Funct Ecol 12:912–919. Tizard I. 2008. Sickness behavior, its mechanisms and significance. Anim Health Res Rev 9:87–99. Tobler M., D. Hasselquist, H.G. Smith, and M.I. Sandell. 2010. Short- and long-term consequences of prenatal testosterone for immune function: an experimental study in the zebra finch. Behav Ecol Sociobiol 64:717–727. Toomey M.B., M.W. Butler, and K.J. McGraw. 2010. Immunesystem activation depletes retinal carotenoids in house finches (Carpodacus mexicanus). J Exp Biol 213:1709–1716. Tschirren B. and H. Richner. 2006. Parasites shape the optimal investment in immunity. Proc R Soc B 273:1773–1777. von Schantz T., S. Bensch, M. Grahn, D. Hasselquist, and H. Wittzell. 1999. Good genes, oxidative stress and conditiondependent sexual signals. Proc R Soc B 266:1–12. Zuk M. and A.M. Stoehr. 2002. Immune defense and host life history. Am Nat 160:9–22.
