- Top of page
- Materials and methods
- Supporting Information
Phenotypic flexibility allows animals to adjust their physiology to diverse environmental conditions encountered over the year. Examining how these varying traits covary gives insights into potential constraints or freedoms that may shape evolutionary trajectories. In this study, we examined relationships among haematocrit, baseline corticosterone concentration, constitutive immune function and basal metabolic rate in red knot Calidris canutus islandica individuals subjected to experimentally manipulated temperature treatments over an entire annual cycle. If covariation among traits is constrained, we predict consistent covariation within and among individuals. We further predict consistent correlations between physiological and metabolic traits if constraints underlie species-level patterns found along the slow-fast pace-of-life continuum. We found no consistent correlations among haematocrit, baseline corticosterone concentration, immune function and basal metabolic rate either within or among individuals. This provides no evidence for constraints limiting relationships among these measures of the cardiovascular, endocrine, immune and metabolic systems in individual red knots. Rather, our data suggest that knots are free to adjust individual parts of their physiology independently. This makes good sense if one places the animal within its ecological context where different aspects of the environment might put different pressures on different aspects of physiology.
- Top of page
- Materials and methods
- Supporting Information
Organismal traits can show amazing flexibility, with physiological characteristics from gut size, to metabolic rate, to immune function showing as much variation as behaviour (Piersma & van Gils, 2011). This phenotypic flexibility allows animals to adjust their physiology to diverse environmental conditions and can potentially influence evolution through selection operating on suites of flexible but linked traits with downstream consequences for the genome (West-Eberhard, 2003), and/or, through the accumulation and release of cryptic genetic variation (reviewed by Pfennig et al., 2010). Therefore, studying how flexible traits relate to each other – how they covary – adds a layer of complexity needed to gain insight into the constraints or freedoms that may shape evolutionary trajectories (Lande & Arnold, 1983).
Trait covariances, for example those among morphological traits, can have profound evolutionary implications (Lande & Arnold, 1983) because they affect how populations respond to selection by constraining responses, generating trade-offs, or otherwise shaping evolutionary trajectories (Roff & Fairbairn, 2007). Therefore, the way that physiological traits covary at the individual level may be relevant for testing hypotheses about how these same traits covary at the species level. For example, the pace-of-life hypothesis (Ricklefs & Wikelski, 2002) argues that observed correlations between demographic and metabolic traits along a slow-fast continuum (Promislow & Harvey, 1990) might indicate that individual responses to different environments are limited by physiological mechanisms. In other words, different physiological systems are intricately linked and thus constrained (i.e. the endocrine system, the immune system and the metabolic system; Dhabhar et al., 1995; Lochmiller & Deerenberg, 2000; Speakman, 2005; Landys et al., 2006; Ardia et al., 2011). If such constraints underlie the slow–fast continuum seen at the species level, then physiological traits should covary in the same way at the species and individual levels (Lande, 1979). This line of thinking also predicts consistent correlations among physiological traits, and between these traits and metabolic traits (i.e. basal metabolic rate, BMR) at the individual level.
Thus, at the individual level, consistent correlations among physiological traits are predicted if functional interactions among suites of traits limit trait evolution – functional constraint (Schwenk & Wagner, 2001). However, traits may also be negatively correlated in some environments or parts of the year due to trade-offs necessitated by the allocation of limited resources (Ardia et al., 2011). Finally, phenotypic flexibility may allow organisms to adjust individual aspects of their physiology independently. This may reflect the fact that an organism’s overall physiology must serve multiple functions, with each physiological system responding to environmental conditions particular to its function (Piersma & van Gils, 2011). This freedom to adjust individual aspects of physiology might be important in environments with, or during times of the year when, incompatible demands result in physiological conflicts (Ramenofsky & Wingfield, 2006; Vézina et al., 2010, 2012).
Though seemingly clear-cut, these hypotheses may be difficult to tease apart. Consider a simplified and hypothetical example of two physiological traits (A and B) over the annual cycle (Fig. 1a). The scenarios detailed in the figure show that relationships between the traits in different periods of the year (columns represent Jan, May and Aug) could indicate: (a1) true functional constraint; (a2) negative relationships during periods of scarcity (trade-offs) and positive relationships during periods of abundance even though traits are independent; (a3) negative or positive relationships due to parallel or opposite responses to the environment (Ricklefs, 2000) even thought traits are independent; or (a4) a consistent lack of correlations indicating independent traits responding to independent aspects of the environment. Importantly, the negative correlations in column one (Jan) in a1–a3 could be interpreted as constraint if studied in isolation, but only a1, where the traits are correlated in the same way in all months, represents true functional constraint. This complexity highlights the importance of sampling traits in different environments and over time.
Figure 1. Hypothetical scenarios illustrating how physiological traits A and B might covary both over time (a) and among individuals (b). Data clouds are represented by ellipses, where the tilt, narrowness and colour of the ellipse represent the direction, strength and significance of the correlations. In A points in the data cloud are individuals, and each graph in a row represents a month in the annual cycle (Jan, May or Aug). In B points in the data cloud represent repeated samplings of the same individual, and each graph in a row represents a different individual.
Download figure to PowerPoint
At the individual level, physiological syndromes (akin to behavioural syndromes reviewed by Bell, 2007) may exist if physiological traits are consistently correlated within an individual over time and across environments. The scenarios detailed in Fig. 1b show that relationships between traits measured repeatedly in different individuals (columns represent birds 1–3) could indicate: (b1) physiological syndromes caused by functional constraint; (b2) consistent syndromes within an individual over time, but because at least two relationships are possible in the population, no evidence for constraints that operate consistently in all individuals (i.e. if they exist, they are not universal); (b3 and b4) no syndromes or functional constraint. The pattern in b2 may also be caused by ecological constraints (e.g. individuals using different habitats). In b3, the strategy of ‘no relationship’ is not a syndrome because the relationship is not consistent over time (dots in the data could represent multiple samplings within the individual, and all possible combinations are present).
Red knots Calidris canutus (Linnaeus 1758) of the northerly wintering subspecies C. c. islandica (hereafter knots) are medium-sized (100–200 g) long-distance migrant shorebirds. Because they live well in captivity and have well-studied annual cycles and physiology (Piersma, 2007; Buehler & Piersma, 2008; Piersma & van Gils, 2011), we used them as a model to investigate covariation among different physiological traits over time and among individuals. Although many aspects of knot physiology are flexible, in this study, we focus on four traits: (i) haematocrit to represent the cardiovascular system, (ii) baseline corticosterone to represent the endocrine system, (iii) BMR to represent metabolism and (iv) aspects of constitutive (noninduced) immune function to represent the immune system. Haematocrit is the proportion of red blood cells per total blood volume and, along with haemoglobin concentration and oxygen affinity, determines blood oxygen-carrying capacity. Increased haematocrit is associated with increased oxygen-carrying capacity during increased workload in migrating birds (Bairlein & Totzke, 1992; Piersma et al., 1996; Prats et al., 1996), including associations between haematocrit and premigratory mass gain in knots (Piersma et al., 2000a; D.M. Buehler, unpublished); and with cardiovascular responses accompanying changes in body temperature in lizards (Snyder, 1977) and frogs (Withers et al., 1991). Thus, haematocrit represents an easily obtainable and widely measured aspect of the caridiovascular system’s ability to deliver oxygen. Corticosterone is a widely measured glucocorticoid hormone involved in the onset and regulation of migratory movements and is elevated in association with migration and arrival at the breeding grounds in knots (Piersma et al., 2000b; Reneerkens et al., 2002; Landys et al., 2004). Corticosterone may therefore be correlated with other physiological processes known to vary in association with migration including haematocrit (Piersma et al., 1996, 2000a; Landys-Ciannelli et al., 2002), immune function (Buehler et al., 2008b) and BMR (Vézina et al., 2011).
In this study, we describe annual variation in haematocrit and baseline corticosterone and combine these data with previously published datasets of immune function (Buehler et al., 2008b) and BMR (Vézina et al., 2011) measured in the same individuals. These birds were part of a yearlong experiment in which we exposed them to cold, warm (thermoneutral) or variable (tracking seasonal conditions) temperatures to manipulate thermoregulatory costs and to uncouple seasonal changes from physiological adjustments to ambient temperature (see Buehler et al., 2008b; Vézina et al., 2011 for details). Indices of immune function were chosen to cover a range of protective functions including the functional capacity of blood to limit microbial infection (Tieleman et al., 2005; Millet et al., 2007), concentrations of circulating immune cells and levels of complement and natural antibodies. These immune indices are seasonally variable but also repeatable characteristics of individual birds (Buehler et al., 2008b); furthermore, they assay an evolutionarily important branch of the immune system that provides broad spectrum and immediate protection against invaders. We used these combined data on haematocrit, corticosterone, immune function and BMR to look at covariation among traits over time and at the individual level. If functional constraints give rise to physiological syndromes, then we predict consistent covariation within and among individuals. Furthermore, if functional constraints underlie species-level patterns found along the slow-fast pace-of-life continuum, then we predict consistent correlations between physiological and metabolic traits.