Skip to main content
NCBI home page
Ecology and Evolution logoLink to Ecology and Evolution
. 2015 Aug 21;5(17):3808–3817. doi: 10.1002/ece3.1638

Large-scale spatial variation in feather corticosterone in invasive house sparrows (Passer domesticus) in Mexico is related to climate

Gillian D Treen 1, Keith A Hobson 1,2, Tracy A Marchant 1, Gary R Bortolotti 1,†
PMCID: PMC4567882  PMID: 26380707

Abstract

Ecologists frequently use physiological tools to understand how organisms cope with their surroundings but rarely at macroecological scales. This study describes spatial variation in corticosterone (CORT) levels in feathers of invasive house sparrows (Passer domesticus) across their range in Mexico and evaluates CORT–climate relationships with a focus on temperature and precipitation. Samples were collected from 49 sites across Mexico. Feather CORT (CORTf) was measured using methanol-based extraction and radioimmunoassay. Relationships between CORTf and spatial and climate variables were examined using simple linear regressions. Ordination was used on climate data, CORTf was plotted against the resulting axes, and univariate regression trees were used to identify important predictors of CORTf. Universal kriging interpolation was used to illustrate spatial variation in CORTf across Mexico. Correlations with ordination axes showed that high CORTf was associated with low precipitation during the rainy season and low dry season temperatures. Specifically, CORTf was negatively related to May precipitation and January and July minimum temperatures, and positively related to April deuterium excess and June minimum temperatures. CORTf was higher in second-year birds compared to after-hatch years and after-second years. House sparrows had higher CORTf levels in the hot, dry, north-central region of Mexico, and CORTf was negatively related to temperature and precipitation. House sparrows molt primarily from August–September but climate conditions throughout the year were important predictors of CORTf, suggesting that conditions outside of molt can carry over to influence energetics during feather growth. These data suggest that dry conditions are challenging for house sparrows in Mexico, supporting previous work showing that precipitation is an important predictor of broad-scale CORT variation. This work highlights the utility of CORTf for evaluating the influence of physiology on current avian range limits; furthermore, these data may allow us to predict future changes in species distributions.

Keywords: Climate, corticosterone, ecophysiology, feather CORT, invasive species, macrophysiology, Mexico, Passer domesticus, precipitation, temperature

Introduction

Combining ecology and physiology has led to considerable advancement in these fields, and the advantages of using ecophysiological data collected at broad scales to answer biological questions have long been recognized (Chown et al. 2004). Macrophysiology emphasizes variation in physiological traits at large scales, information which can be used to generate novel, holistic answers to basic questions in ecophysiology. For example, assessing current patterns can help us understand species distribution limits and predict future responses to climate or land-use changes (Chown and Gaston 2008). This approach is especially powerful for invasive species, as understanding what factors limit their range expansion is key to planning where to target management and what types of interventions will be most effective. It may also help determine how current species' ranges could expand or shift in response to predicted climate variation (Zuckerberg et al. 2011).

By linking broad-scale variation in physiological variables to climate, we can infer how well organisms are coping with their environments. The glucocorticoid (GC) hormone axis in particular is likely to be a physiological process that reflects the effects of climate. GCs are involved in energy regulation, especially processes that provide energy for routine daily tasks such as glucose synthesis and fat breakdown (Dallman et al. 1993; Toates 1995). Corticosterone (CORT) is the main avian GC and is often incorporated into ecological studies of birds because it is secreted in higher amounts when an animal experiences an unpredictable environmental perturbation or stressor (Wingfield et al. 1998; Romero 2004). Previous work has shown that plasma CORT (CORTp) levels are higher following storms or periods of extreme precipitation (Rogers et al. 1993; Smith et al. 1994; Astheimer et al. 1995; Boyle et al. 2010), and CORTp and fecal CORT increases have also been associated with cool temperatures (Frigerio et al. 2004; Jenni-Eiermann et al. 2008; Lobato et al. 2008), cool, rainy conditions (Bize et al. 2010), and prolonged bouts of precipitation (Pereyra and Wingfield 2003). Furthermore, song wrens (Cyphorhinus phaeocephalus) had higher CORTp levels in dry areas near their range limits on the isthmus of Panama, suggesting that they are unable to cope easily with these conditions and that this limited further expansion (Busch et al. 2011).

CORT can be measured in plasma, feces, or feathers, but there are several advantages to using feather CORT (CORTf). Blood measures provide an instantaneous picture of individual hormone levels, whereas CORTf offers a longer-term perspective, integrating both baseline levels and any elevations occurring during the period of feather growth (Bortolotti et al. 2008). CORTf also allows investigators to avoid the potential negative effects of blood sampling on survival (Brown and Brown 2009) and the difficulty of obtaining baseline blood samples in the field (Romero and Romero 2002; Romero and Reed 2005). CORT–climate relationships can vary with species, life-history stage, sex, age class, and social status. For example, Wingfield (1985a,b) found that an early spring storm was associated with increased CORTp in female but not male song sparrows (Melospiza melodia), while during a late spring storm, males but not females showed elevated CORTp levels. Rubenstein (2007) found that rainfall during the pre-breeding period was negatively related to CORTp in subordinate superb starlings (Lamprotornis superbus), while dominant individuals showed no relationship (for other examples, see Schwabl et al. 1985; Rohwer and Wingfield 1981; Romero et al. 2000). Furthermore, CORTf levels of common eiders (Somateria mollissima) in Nunavut, Canada, were positively related to August and September temperatures, and these variables explained the majority of the interannual variation in CORTf (Legagneux et al. 2013). CORTf has also been correlated with clutch size, social signals, and habitat conditions (Bortolotti et al. 2008; Harms et al. 2010; Fairhurst 2011) and associated with future survival probability of wild house sparrows (Passer domesticus) on an island off the coast of England, suggesting that it may be a useful biomarker for conservation (Koren et al. 2012).

Here, we present the first exploration of CORTf variation in a single species at a broad spatial scale. Our primary aim was to look for patterns in house sparrow CORTf across their range in Mexico and identify potential drivers of this variation. Specifically, we assessed relationships between CORTf and spatial and local climate variables operating over various time scales. We predicted that CORTf levels would be highest in birds living in low precipitation areas, similar to the findings of Busch et al. (2011). Also, we expected to see negative relationships between CORTf and both elevation and latitude. We predicted that CORTf would be negatively related to temperatures and precipitation levels during the molt period, defined as August–September (Casto 1974; Lowther and Cink 2006). This relationship with temperature could reflect direct effects of cool temperatures on energetic demands or indirect effects of temperature on food availability. Associations between rainfall and CORTf are likely indirect, mediated through the effects of rainfall on food supply, as previously hypothesized in CORTp studies (e.g., Bize et al. 2010; Busch et al. 2011).

Methods

Study species and study area

House sparrows are native to Europe and Asia but were introduced to the United States in the 1850s, and their New World range now extends from northern Saskatchewan and Manitoba, Canada, to Panama (Lowther and Cink 2006). House sparrows spread across the majority of Mexico between 1910 and the 1970s (Robbins 1973; Schrey et al. 2011). Their success as an invasive species has been partly attributed to the fact that they are generalist feeders and human commensals (Lowther and Cink 2006). The human population of Mexico has increased rapidly since the 1940s, and the percentage of the population living in urban areas increased from 35.1% in 1940 to 78% in 2010 (United Nations 2012). Although most house sparrows in Mexico likely reside in urban areas or near rural dwellings, climate and physical geography vary considerably across the country. The coasts and the Yucatan peninsula are characterized by warmer temperatures (mean annual temperature >22°C), while the majority of the interior is cooler (mean annual temperatures ranging from 12 to 22°C; in some high-altitude regions, mean annual temperatures range from <5 to 12°C; Alemán and García 1974; Rudolph 1985). The Sonoran and Chihuahuan deserts are located in north-central Mexico, and along with the Baja peninsula, these regions are arid and subject to extreme high temperatures (mean July temperatures can range from 25 to 30°C) and considerable annual temperature variability (range of 16–20°C; Alemán and García 1974). Mean annual rainfall is highest on the Yucatan peninsula (range of 116–131 cm) and in the southern and central regions of Mexico (range of approx. 38–115 cm), while the northern plateau region and the Baja peninsula are more arid (mean annual levels 11–27 and 1–10 cm, respectively; Alemán and García 1974; Rudolph 1985).

Field methods and climate data

Feathers were collected from December 2006 to March 2007 as part of a study that developed a feather δ2H isoscape for Mexico (Hobson et al. 2009). Sampling sites (n = 49) were chosen based on obtaining adequate coverage of the country and on accessibility from roadways. Birds were captured using mist nets, and individuals were sexed and assigned an age class (hatch year, HY; after-hatch year, AHY; second year, SY; after-second year, ASY; or unknown, U). Unflattened wing length was measured to the nearest mm. The number of individuals sampled per site ranged from 1 to 20, with a mean and mode of 9. Feathers were stored in paper envelopes at room temperature until they were used for hormone analysis in 2011.

Monthly precipitation, mean annual precipitation, and average monthly minimum and maximum temperatures for each sampling site were obtained from high resolution (1-km spatial resolution) interpolated global climate surfaces developed by Hijmans et al. (2005). Monthly deuterium excess in groundwater (d-excess = δ2H – 8*δ18O; Clark and Fritz 1997) was calculated for each site based on water samples as described in Wassenaar et al. (2009). Deuterium excess can be used as a proxy for evaporative conditions (Clark and Fritz 1997); we included this variable in addition to precipitation data because d-excess may be more representative of how much water is actually available in the local food web. It is likely that water availability rather than precipitation levels strongly affect factors influencing CORT secretion, such as dehydration or food availability.

Corticosterone analysis

For each individual, the first secondary (S1) was typically sampled (n = 438), but the outer rectrix was used in 10 cases. CORT levels did not differ significantly based on which feather was used (Welch two-sample t-test, P = 0.22). Prior to analysis, the calamus was removed and each feather was measured to the nearest mm using a ruler. CORTf values are reported as pg/mm, based on evidence that CORT is deposited into feathers in a time-dependent rather than a mass-dependent fashion (Bortolotti et al. 2008; Bortolotti 2010). After measurement, feathers were cut into small pieces (<5 mm2) using scissors. CORT was recovered from feathers using a methanol-based extraction technique described and validated in Bortolotti et al. (2008). To evaluate the efficiency of the recovery procedure, three feather samples spiked with ∼5000 CPM of 3H-corticosterone (Perkin Elmer, Woodbridge, ON) were included in each recovery. The 448 feather samples were recovered in 5 batches; for all batches, >90% of the radioactivity was recoverable from the reconstituted samples (mean recovery efficiency 95.8%, SE ±1.28%). Final CORTf values were adjusted to account for recovery efficiency. Samples were stored in a −20°C freezer prior to radioimmunoassay.

Reconstituted samples were analyzed via standard radioimmunoassay procedures (Wayland et al. 2002), with each sample analyzed in duplicate. To avoid bias, samples were placed in random order prior to analysis and the investigator was blind to the site at which each sample was collected. Serial dilutions of feather extracts were shown to parallel the CORT standard curve (see Appendix S2 in Supporting Information), indicating that there were no substances in the extracts that compromised the assay (Buchanan and Goldsmith 2004; Bortolotti et al. 2008). To evaluate assay variability, three internal standards of known hormone concentration were included in each assay. Samples were processed in 10 assays, with a mean intra-assay coefficient of variation (CV) of 6.57% (range 4.50–10.6%) and an interassay CV of 4.63%. The mean detection limit (80% bound) was 10.0 pg CORT per 100 μL of sample, and all data values exceeded this limit (mean = 43.0 pg/100 μL, range 17.2–237).

Statistical analyses

Data exploration was conducted following Zuur et al. (2010), including outlier analysis and evaluation of heteroscedasticity and collinearity of explanatory variables. The distribution of CORTf in the sample was plotted; other preliminary analyses included using t-tests and ANOVAs to explore relationships between CORTf and sex and age class, and using simple linear regressions, controlling for sex, to evaluate relationships between CORTf and wing length.

To explore spatial variation in CORTf, correlations between CORTf and latitude, longitude, and elevation were evaluated using simple linear regressions. T-tests were used to evaluate variation in CORTf between the Atlantic and Pacific drainage basins and between the interior and exterior regions of Mexico (see Hobson et al. 2009). We tested for differences in CORTf related to evapotranspiration using ANOVA; simple linear regressions were used to test for one-way relationships between CORTf and deuterium excess or monthly climate variables. We applied a Bonferroni correction to account for multiple testing, with P < 0.0008 considered significant. Universal kriging interpolation was used to illustrate spatial variation in CORTf across Mexico.

We used ordination to reduce the dimensionality of the climate data. Climate data were non-normal and nonlinear so we used non-metric multidimensional scaling (NMDS; McCune and Grace 2002). The Euclidean distance metric was used as it minimized the loss function compared to other metrics considered, and a final solution with two dimensions (axes) was chosen based on the same criterion (minimal loss function). CORTf was plotted against the ordination axes, and correlations between original variables and axis scores (similar to factor loadings produced by PCA) were used to interpret the climate conditions represented by each axis.

Univariate regression trees were used to identify the most important predictors of CORTf. Trees were “pruned” by plotting tree size and then selecting the tree with the lowest error value. Each tree was run 10 times, and each run included 1000 cross-validations. Cross validation error values estimate prediction error of the model, and error averaged over all runs is reported for each tree. All statistical analyses were performed in R v. 2.14.0 and the “mvpart” library was used for regression trees (R Development Core Team 2012); kriging interpolation was carried out in ArcMap v. 10.1.

Results

CORTf levels did not differ significantly between the sexes (males Inline graphic = 5.8 ± 2.9 pg/mm, n = 257; females Inline graphic = 5.6 ± 2.1 pg/mm, n = 191; P = 0.42) or age classes (ANOVA, df = 4, P = 0.74). CORTf levels were not strongly related to wing length for either sex (males R2 = 0.007705, F = 1.833, df = 236, P = 0.1771; females R2 = 0.0003422, F = 0.06229, df = 182, P = 0.803).

The majority of the individuals sampled had CORTf levels ranging from 2.5 to 10 pg/mm. Individuals with CORTf levels above the 95th percentile (n = 23) were considered “outliers,” and their CORTf levels ranged from 9.6 to 32.4 pg/mm. These individuals did not consistently belong to one sex or age class. However, they were all sampled at one of seven sites, with 12 captured at a single site. Additionally, 22 of the 23 individuals with high CORTf values were sampled in the interior of Mexico and in the Atlantic drainage basin. CORTf values were highest in north-central Mexico and the Baja peninsula and slightly lower in central Mexico compared to the rest of the country (Fig.1).

Figure 1.

Figure 1

Open in a new tab

Kriged surface of feather corticosterone (CORTf) values of house sparrows (n = 425) sampled from 49 sites in Mexico. Universal kriging was used and data are shown with a cylindrical equal-area map projection. Birds with CORTf levels above the 95th percentile (ranging from 9.6 – 32.4 pg/mm; n = 23) were considered “outliers” and were removed from the data set prior to analysis. CORTf values were averaged for each site (sample size varied between sites; minimum n = 1, maximum n = 19, median n = 8).

One-way associations between CORTf and latitude, longitude, and elevation were weak (R2 values all <0.045) although the regression with latitude was significant (P < 0.0001). CORTf values differed significantly between east and west drainage basins (Atlantic Inline graphic = 6.4 ± 3.3 pg/mm, n = 213; Pacific Inline graphic = 5.1 ± 1.4 pg/mm, n = 235; P < 0.0001) and between the interior and exterior drainage basins (as described in Wassenaar et al. 2009; interior Inline graphic = 6.4 ± 3.5 pg/mm, n = 192; exterior Inline graphic = 5.2 ± 1.3 pg/mm, n = 256; P < 0.0001). After removing 23 outliers from the data set, the difference was not significant for either comparison (Atlantic versus Pacific P = 0.003; interior versus exterior P = 0.071). CORTf also differed significantly based on evapotranspiration level (ANOVA, df = 11, F = 4.46, P < 0.0001), with higher mean CORTf values associated with low evapotranspiration values (<100 and 200–300 mm water loss due to evapotranspiration, Fig.2). Regressions modeling relationships between CORTf and monthly precipitation, minimum and maximum temperature, and d-excess values did not reveal strong one-way associations (all R2 values <0.08); however, several models had significant P-values at the α = 0.0008 level (see Appendix S1).

Figure 2.

Figure 2

Open in a new tab

The relationship between feather corticosterone (CORTf) and evapotranspiration level in house sparrows (n = 425) sampled form 49 sites across Mexico; CORTf levels differed significantly based on evapotranspiration level (ANOVA, df = 11, F = 4.46, P < 0.0001).

The first ordination axis (NMDS1) was positively associated with precipitation during the molt period (in house sparrows molt primarily occurs in August and September, although more conservative estimates define the molt period as June–November; see Casto 1974; Mathew and Naik 1986; Lowther and Cink 2006; Romero et al. 2006) which overlaps with the rainy season in Mexico (July–September). NMDS1 was also positively associated with minimum and maximum temperatures of months outside of the molt period, in the dry season. The second axis (NMDS2) was negatively associated with precipitation during the molt period/rainy season and positively associated with minimum and maximum monthly temperatures during the molt period/rainy season. High CORTf values were associated with low values of NMDS1, which corresponds to dry conditions during the molt period/rainy season and cool temperatures during the dry season (see Appendix S1). Mid-range values of NMDS2, which correspond to intermediate temperatures and levels of precipitation during the molt period/rainy season, were also associated with high CORTf. Further examination showed that some high CORTf values were also associated with high values of NMDS2, which correspond to low levels of precipitation during molt, and mid-range values of NMDS1, which correspond to intermediate levels of precipitation during the molt period.

Regression trees were run with and without site groundwater d-excess data. In all cases, the most important variable predicting CORTf was site. To determine whether the dominance of site was driven by seven sites from which outliers were sampled, a stepwise elimination of these sites was conducted with regression trees run after each removal. After removing these sites, site was still the top predictor in all trees. Age, mean annual precipitation, and d-excess also emerged as important predictors of CORTf. For the data set excluding d-excess, the modal best tree size from the 10 runs considered was three nodes, and both splits were based on site. Regression trees run without site identified minimum temperatures in January, June, and July, and age class as important predictors of CORTf (Fig.3). Rerunning regression tree analyses of the data set including d-excess data after removing site identified d-excess in April and May precipitation as important predictors of CORTf (Fig.4). As some of the climate variables were collinear, we also looked at surrogate variables for each split for the trees developed after removing site. Generally, the surrogate variables were quite similar to the variable on which the split was based. However, in some cases, precipitation variables or spatial data (latitude or longitude) appeared as surrogates for temperature variables, or vice versa.

Figure 3.

Figure 3

Open in a new tab

Regression tree showing the division of 448 house sparrows (Passer domesticus) sampled across Mexico into groups based on feather corticosterone (CORTf), with group divisions determined based on one of 45 predictor variables, each describing spatial location or climatic conditions of sampling sites. For each terminal node, the group size (n) and the mean CORTf for that group (in pg/mm) are listed. Threshold values for each split are also listed; for temperature variables, units are °C.

Figure 4.

Figure 4

Open in a new tab

Results of rerunning the regression tree analysis shown in Figure3 with monthly deuterium excess included as predictor variables.

Discussion

House sparrows in Mexico had higher CORTf levels in areas characterized by low precipitation and increased temperature variability. Overall CORTf was negatively related to both temperature and precipitation, and weather conditions throughout the annual cycle were important predictors of CORTf. As CORTf can be interpreted as a measure of energetic demand, our results indicate that weather may be an important factor limiting the distribution of this invasive species. More broadly, this work shows that linking CORTf to environmental variables can identify conditions that are difficult for a species to cope with, which is likely an important determinant of the boundaries of the species' current range. Expanding upon this work, if CORTf can be linked to survival, it may prove useful for predicting the extent of future range shifts or expansions.

To our knowledge, this study is the first to establish how intraspecific CORT levels vary across a broad geographical range and adds to a growing body of evidence confirming that CORTp–climate relationships can be replicated using CORTf (see Fairhurst et al. 2012, 2013; Legagneux et al. 2013). While variation in house sparrow CORTf across their range in Mexico was generally subtle, small differences in CORT levels can have fitness consequences (Romero and Wikelski 2001; Fairhurst 2011; Koren et al. 2012). Individuals with extremely low or extremely high CORTp values are compromised in their ability to cope with their environment and often do not survive (Romero et al. 2009), reducing the probability of observing greater ranges of variability in CORTf. Alternately, because house sparrows tend to live in close associations with humans (see below), they may be “buffered” from environmental variation and as a result show fairly consistent CORTf levels, rather than CORTf levels seeming consistent due to reduced survival of individuals at either extreme.

Many studies have shown that sex, age, and body condition can influence plasma and fecal CORT (e.g., Bonier et al. 2007; Cabezas et al. 2007; Wilcoxen et al. 2011), but CORTf did not differ significantly between the sexes or age groups and was not significantly associated with wing length in our study. One explanation for the uniformity of CORTf levels of birds sampled across the country could be that house sparrows are human commensals. As such, their habitats may be fairly consistent in terms of food supply, shelter, predation pressure, and other factors affecting energetic requirements. The north-central region of the country and to an extent the Baja peninsula appear to be exceptions to this hypothesis, perhaps suggesting that in these areas, birds are less buffered by associations with humans. Studies have found that other urban bird species show CORTp differences based on sex (Bonier et al. 2007), age class (De Neve et al. 2010), and body condition (Fokidis et al. 2011) despite their close association with humans. A primary aim of this study was to create a large-scale “feather CORTscape,” which required sampling a species with a broad distribution. However, it would be interesting to conduct a similar study on a species that is less associated with humans, to determine whether their CORTf levels are more strongly linked to environmental variables.

The dominant spatial pattern in CORTf was the clustering of the high CORT birds in the north-central region of Mexico, where temperatures vary considerably over the course of the year relative to other parts of the country (mean annual temperatures ranging from 12 to 22°C) and conditions are dry (mean annual rainfall ranging from 11 to 27 cm; Alemán and García 1974). Birds with high CORTf levels were also sampled near the tip of the Baja peninsula, where conditions are similar, dry with variable temperatures. The tendency for CORTf levels to be higher in the driest parts of a species' range is similar to previous findings using CORTp of song wrens (Busch et al. 2011), and a positive CORTf–temperature relationship has been shown in common eiders (Legagneux et al. 2013). The pattern we observed in the present study could result from the combined challenges of hot and dry conditions in this area. Alternatively, high temperatures alone could be perceived as a stressor or increase energetic demands due to thermoregulation. While our results suggest a potential physiological mechanism linking weather conditions to house sparrow range limits in Mexico, specific data on weather conditions at the species' range limits are necessary to further investigate this connection.

Contrary to our a priori predictions, CORTf showed a weak, positive relationship with latitude. Other work has shown no effect of latitude on CORTp (Lynn et al. 2003), CORTp–latitude relationships that varied between years (Lindström et al. 2005) and positive relationships between latitude and CORTp (Wingfield et al. 1995; Martin et al. 2005), suggesting that the latitudinal effect on CORT is variable. As birds sampled in desert-like northern Mexico had higher CORTf levels, the positive CORTf–latitude relationship may reflect a negative association between latitude and rainfall in the study area. Similarly, higher latitudes could be associated with increased temperature variation and less stable temperatures with increased CORTf. Our study found an inconsistent association between elevation and CORTf in house sparrows, with the intermediate elevations associated with the highest CORTf levels. Previous studies have found a negative relationship between CORTp and elevation (Bears et al. 2003; Pereyra and Wingfield 2003; Li et al. 2008). It is difficult to conceive of a biological explanation for our results, and it is possible that there are site specific variables that we did not capture in the present analysis. Alternatively, house sparrows may be somewhat “buffered” from altitude effects on the GC axis because they often live in close association with humans, and therefore, the availability of food, protection from adverse weather conditions, etc. may be similar across their altitudinal range.

CORTf levels were significantly higher in the Atlantic versus the Pacific drainage basin and in the interior versus the coastal region. The majority of the high CORTf outliers were sampled in the interior/Atlantic drainage basin; after removing these outliers from the data set, the Atlantic versus Pacific comparison still yielded a significant P-value although the interior versus exterior comparison did not. This suggests that unmeasured difference(s) between these broad regions of the country are also associated with CORTf variation.

CORTf was negatively associated with temperatures and precipitation levels, but one-way associations were weak. However, the clustering of the high CORTf birds in the dry north-central region of the country suggests that precipitation is an important ecological variable explaining variation in CORTf. Previous studies have found negative associations between CORTp and precipitation levels in white-crowned sparrows (Zonotrichia leucophrys pugetensis) in Washington (Wingfield et al. 1983), in dark-eyed juncos (Junco hyemalis) wintering across the United States (Rogers et al. 1993), and in Alpine swifts (Apus melba) in Switzerland (Bize et al. 2010). House sparrows eat grains primarily but also consume insects (Lowther and Cink 2006). As rainfall is important for plant growth and climate can influence the availability of insect prey (Murphy 1987), the CORTf–precipitation associations in the Mexican population could be driven by effects of precipitation on food supply. These sparrows could also be at a risk of dehydration in extremely arid regions. In this case, rainfall could be directly affecting CORTf by increasing the energetic costs of obtaining enough water to maintain homeostasis.

In our study, CORTf showed a negative relationship with temperature and birds with the highest CORTf levels were sampled in a region where annual temperature variation is relatively high. Low ambient temperatures have been associated with higher plasma and fecal CORT levels in multiple species and contexts, including diving petrels (Pelecanoides urinatrix) off the coast of South Georgia Island (Smith et al. 1994), Greylag geese in Austria (Frigerio et al. 2004), and nestling blue tits (Cyanistes caeruleus) and pied flycatchers (Ficedula hypoleuca) in central Spain (Lobato et al. 2008). Additionally, tree swallow (Tachycineta bicolor) nestlings that experienced greater temperature variability within nest boxes had higher CORTf levels (Fairhurst et al. 2012), suggesting that CORTf reflects differences in the range of temperatures experienced, in addition to differences in absolute temperatures. These associations likely reflect effects of ambient temperature on energetic requirements or food availability. During the molt period, extreme minimum temperatures at our sampling sites ranged from 8.2°C for June to 1.4°C for November and the lower bound of the thermoneutral zone for house sparrows has been designated as 20–22°C (Hudson and Kimzey 1966).

Interestingly, CORTf was not most strongly related to climate variables during the molt period although CORTf reflects energetic condition during feather growth (Fairhurst 2011). This suggests that conditions experienced during the wintering or breeding stages could have effects that carry over to influence energetic state during molt. For example, cool temperatures during the pre-breeding or breeding stages could cause birds to expend more energy to maintain their body temperature, resulting in reduced fat storage. Unless these negative effects can be countered by increased fat storage during breeding, the individual would be in relatively poor condition during the feather growth period, which could result in higher CORTf. Alternately, extreme temperatures or precipitation levels earlier in the year could determine the quantity or quality of food available during the breeding and molt periods.

Conclusions

We found broad spatial patterns in house sparrow CORTf levels and linked variation in CORTf to climatic variables, indicating that CORTf can enhance our understanding of physiological limitations to avian distributions. Our results suggests that assessing CORTf–climate relationships can inform how invasive and non-invasive species will respond to climate change, critical knowledge required to minimize their impacts on native species and the habitats on which they depend. Additionally, this work shows that previously described relationships between climate variables and plasma or fecal CORT can be detected using CORTf, adding to a growing body of evidence highlighting the utility of CORTf for assessing how birds cope with environmental challenges.

Acknowledgments

We thank Jill Johnstone, Karen Machin, and Graham Fairhurst for their thoughtful comments on the thesis chapter that formed the basis for this manuscript. Sonia Cabezas and Vicky Fachal assisted with corticosterone analysis and Michel Laforge assisted with kriging analysis. We are grateful to Keith Larson and all those who assisted with field sampling in Mexico. GDT was supported by an Alexander Graham Bell Canada Graduate Scholarship (CGS-M) from the National Science and Engineering Research Council of Canada (NSERC), a University of Saskatchewan Dean's Scholarship, and the Gary Bortolotti Graduate Award in Ecology or Animal Behaviour. KAH was supported by an operating grant from Environment Canada.

Conflict of Interest

None declared.

Supporting Information

Figure S1. Feather CORT values of house sparrows (Passer domesticus) sampled in Mexico plotted as contour lines on the reduced space created by two axes produced by ordination of 37 climate variables.

ece30005-3808-sd1.pdf (88.2KB, pdf)

Figure S2. Serial dilutions (1:1, 1:2, and 1:4) of four house sparrow feather extracts plotted against a 4-parameter standard curve fitted to nine corticosterone (CORT) standards ranging in concentration from 3.125 to 800 pg/100 µL.

ece30005-3808-sd2.pdf (82.1KB, pdf)

Table S1. Results of linear regressions between feather corticosterone (CORTf) of house sparrows sampled from 49 sites across Mexico and measures of monthly temperature, precipitation, and deuterium excess (d-excess).

ece30005-3808-sd3.docx (15.1KB, docx)

Table S2. Recovery efficiencies of five methanol extractions used to extract corticosterone from house sparrow feathers (n = 448).

ece30005-3808-sd4.docx (12.9KB, docx)

Table S3. Inter-assay comparisons of ten radioimmunoassays used to measure corticosterone (CORT) levels of house sparrow feather extracts (n = 448 samples).

ece30005-3808-sd5.docx (14.4KB, docx)

Appendix S1-S2. Complete captions of Figures S1, S2 and Tables S1–S3.

ece30005-3808-sd6.doc (28KB, doc)

References

  1. Alemán PAM. García E. The climate of Mexico. In: Bryson RA, Hare FK, editors; World survey of climatology volume 11: climates of North America. Amsterdam: Elsevier Scientific Publishing Company; 1974. pp. 345–404. , eds., and. [Google Scholar]
  2. Astheimer LB, Buttemer WA. Wingfield JC. Seasonal and acute changes in adrenocortical responsiveness in an arctic-breeding bird. Horm. Behav. 1995;29:442–457. doi: 10.1006/hbeh.1995.1276. [DOI] [PubMed] [Google Scholar]
  3. Bears H, Smith JNM. Wingfield JC. Adrenocortical sensitivity to stress in Dark-eyed Juncos (Junco hyemalis oregonus) breeding in low and high elevation habitat. Ecoscience. 2003;10:127–133. [Google Scholar]
  4. Bize P, Stocker A, Jenni-Eiermann S, Gasparini J. Roulin A. Sudden weather deterioration but not brood size affects baseline corticosterone levels in nestling Alpine swifts. Horm. Behav. 2010;58:591–598. doi: 10.1016/j.yhbeh.2010.06.020. [DOI] [PubMed] [Google Scholar]
  5. Bonier F, Martin PR, Sheldon KS, Jensen JP, Foltz SL. Wingfield JC. Sex-specific consequences of life in the city. Behav. Ecol. 2007;18:121–129. [Google Scholar]
  6. Bortolotti GR. Flaws and pitfalls in the chemical analysis of feathers: bad news–good news for avian chemoecology and toxicology. Ecol. Appl. 2010;20:1766–1774. doi: 10.1890/09-1473.1. [DOI] [PubMed] [Google Scholar]
  7. Bortolotti GR, Marchant TA, Blas J. German T. Corticosterone in feathers is a long-term, integrated measure of avian stress physiology. Funct. Ecol. 2008;22:494–500. [Google Scholar]
  8. Boyle WA, Norris DR. Guglielmo CG. Storms drive altitudinal migration in a tropical bird. Proc. Biol. Sci. 2010;277:2511–2519. doi: 10.1098/rspb.2010.0344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brown MB. Brown CR. Blood sampling reduces annual survival in Cliff Swallows (Petrochelidon Pyrrhonota. Auk. 2009;126:853–861. [Google Scholar]
  10. Buchanan KL. Goldsmith AR. Noninvasive endocrine data for behavioural studies: the importance of validation. Anim. Behav. 2004;67:183–185. [Google Scholar]
  11. Busch DS, Robinson WD, Robinson TR. Wingfield JC. Influence of proximity to a geographical range limit on the physiology of a tropical bird. J. Anim. Ecol. 2011;80:640–649. doi: 10.1111/j.1365-2656.2010.01791.x. [DOI] [PubMed] [Google Scholar]
  12. Cabezas S, Blas J, Marchant TA. Moreno S. Physiological stress levels predict survival probabilities in wild rabbits. Horm. Behav. 2007;51:313–320. doi: 10.1016/j.yhbeh.2006.11.004. [DOI] [PubMed] [Google Scholar]
  13. Casto SD. Molt schedule of house sparrows in northwestern Texas. Wilson Bull. 1974;86:176–177. [Google Scholar]
  14. Chown SL. Gaston KJ. Macrophysiology for a changing world. Proc. Biol. Sci. 2008;275:1469–1478. doi: 10.1098/rspb.2008.0137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Chown SL, Gaston KJ. Robinson D. Macrophysiology: large-scale patterns in physiological traits and their ecological implications. Funct. Ecol. 2004;18:159–167. [Google Scholar]
  16. Clark I. Fritz P. Environmental Isotopes in hydrogeology. Boca Raton, FL: CRC Press; 1997. [Google Scholar]
  17. Dallman MF, Strack AM, Akana SF, Bradbury MJ, Hanson ES, Scribner KA, et al. Feast and Famine: critical role of glucocorticoids with insulin in daily energy flow. Front. Neuroendocrinol. 1993;14:303–347. doi: 10.1006/frne.1993.1010. [DOI] [PubMed] [Google Scholar]
  18. De Neve L, Ibañez-Alamo JD. Soler M. Age- and sex-related morphological and physiological differences influence escape capacity in house sparrows (Passer domesticus. Can. J. Zool. 2010;88:1021–1031. [Google Scholar]
  19. Fairhurst GD. 2011. Does an integrated measure of corticosterone from feathers improve our understanding of avian ecophysiology? Doctoral thesis, University of Saskatchewan.
  20. Fairhurst GD, Treen GD, Clark RG. Bortolotti GR. Nestling corticosterone response to microclimate in an altricial bird. Can. J. Zool. 2012;90:1422–1430. [Google Scholar]
  21. Fairhurst GD, Vögeli M, Serrano D, Delgado A, Tella JL. Bortolotti GR. Can synchronizing feather-based measures of corticosterone and stable isotopes help us better understand habitat-physiology relationships? Oecologia. 2013;173:731–743. doi: 10.1007/s00442-013-2678-8. [DOI] [PubMed] [Google Scholar]
  22. Fokidis HB, Hurley L, Rogowski C, Sweazea K. Deviche P. Effects of captivity and body condition on plasma corticosterone, locomotor behavior, and plasma metabolites in curve-billed thrashers. Physiol. Biochem. Zool. 2011;84:595–606. doi: 10.1086/662068. [DOI] [PubMed] [Google Scholar]
  23. Frigerio D, Dittami J, Möstl E. Kotrschal K. Excreted corticosterone metabolites co-vary with ambient temperature and air pressure in male Greylag geese (Anser anser. Gen. Comp. Endocrinol. 2004;137:29–36. doi: 10.1016/j.ygcen.2004.02.013. [DOI] [PubMed] [Google Scholar]
  24. Harms NJ, Fairhurst GD, Bortolotti GR. Smits JEG. Variation in immune function, body condition, and feather corticosterone in nestling Tree Swallows (Tachycineta bicolor) on reclaimed wetlands in the Athabasca oil sands, Alberta, Canada. Environ. Pollut. 2010;158:841–848. doi: 10.1016/j.envpol.2009.09.025. [DOI] [PubMed] [Google Scholar]
  25. Hijmans RJ, Cameron SE, Parra JL, Jones PG. Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 2005;25:1965–1978. [Google Scholar]
  26. Hobson KA, Van Wilgenburg SL, Larson K. Wassenaar LI. A feather hydrogen isoscape for Mexico. J. Geochem. Explor. 2009;102:63–70. [Google Scholar]
  27. Hudson JW. Kimzey SL. Temperature regulation and metabolic rhythms in populations of the house sparrow, Passer domesticus. Comp. Biochem. Physiol. 1966;17:203–217. doi: 10.1016/0010-406x(66)90021-1. [DOI] [PubMed] [Google Scholar]
  28. Jenni-Eiermann S, Glaus E, Grüebler M, Schwabl H. Jenni L. Glucocorticoid response to food availability in breeding barn swallows (Hirundo rustica. Gen. Comp. Endocrinol. 2008;155:558–565. doi: 10.1016/j.ygcen.2007.08.011. [DOI] [PubMed] [Google Scholar]
  29. Koren L, Nakagawa S, Burke T, Soma KK, Wynne-Edwards KE. Geffen E. Non-breeding feather concentrations of testosterone, corticosterone and cortisol are associated with subsequent survival in wild house sparrows. Proc. Biol. Sci. 2012;279:1560–1566. doi: 10.1098/rspb.2011.2062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Legagneux P, Harms NJ, Gauthier G, Chastel O, Gilchrist HG, Bortolotti G, et al. Does feather corticosterone reflect individual quality or external stress in arctic-nesting migratory birds? PLoS ONE. 2013;8:e82644. doi: 10.1371/journal.pone.0082644. . doi: 10.1371/journal.pone.0082644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Li D, Wang G, Wingfield JC, Zhang Z, Ding C. Lei F. Seasonal changes in adrenocortical responses to acute stress in Eurasian tree sparrow (Passer montanus) on the Tibetan Plateau: comparison with house sparrow (P. domesticus) in North America and with the migratory P. domesticus in Qinghai Province. Gen. Comp. Endocrinol. 2008;158:47–53. doi: 10.1016/j.ygcen.2008.06.002. [DOI] [PubMed] [Google Scholar]
  32. Lindström KM, Hawley DM, Davis AK. Wikelski M. Stress responses and disease in three wintering house finch (Carpodacus mexicanus) populations along a latitudinal gradient. Gen. Comp. Endocrinol. 2005;143:231–239. doi: 10.1016/j.ygcen.2005.04.005. [DOI] [PubMed] [Google Scholar]
  33. Lobato E, Merino S, Moreno J, Morales J, Tomás G, Martínez-de la Puente J, et al. Corticosterone metabolites in blue tit and pied flycatcher droppings: effects of brood size, ectoparasites and temperature. Horm. Behav. 2008;53:295–305. doi: 10.1016/j.yhbeh.2007.10.010. [DOI] [PubMed] [Google Scholar]
  34. Lowther PE. Cink CL. House Sparrow (Passer domesticus. In: Poole A, editor; The birds of north America online. Ithaca, NY: Cornell Lab of Ornithology; 2006. ,ed., and. Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/ species/012 doi: 10.2173/bna.12. [Google Scholar]
  35. Lynn SE, Hunt KE. Wingfield JC. Ecological factors affecting the adrenocortical response to stress in Chestnut-Collared and McCown's Longspurs (Calcarius ornatus Calcarius mccownii. Physiol. Biochem. Zool. 2003;76:566–576. doi: 10.1086/375435. [DOI] [PubMed] [Google Scholar]
  36. Martin LB, Gilliam J, Han P, Lee K. Wikelski M. Corticosterone suppresses cutaneous immune function in temperate but not tropical House Sparrows, Passer domesticus. Gen. Comp. Endocrinol. 2005;140:126–135. doi: 10.1016/j.ygcen.2004.10.010. [DOI] [PubMed] [Google Scholar]
  37. Mathew KL. Naik RM. Interrelation between moulting and breeding in a tropical population of the House Sparrow Passer domesticus. The Ibis. 1986;128:260–265. [Google Scholar]
  38. McCune B. Grace JB. Analysis of ecological communities. Gleneden Beach, OR: MjM Software Design; 2002. [Google Scholar]
  39. Murphy MT. The impact of weather on kingbird foraging behaviour. Condor. 1987;89:721–730. [Google Scholar]
  40. Pereyra ME. Wingfield JC. Changes in plasma corticosterone and adrenocortical response to stress during the breeding cycle in high altitude flycatchers. Gen. Comp. Endocrinol. 2003;130:222–231. doi: 10.1016/s0016-6480(02)00568-3. [DOI] [PubMed] [Google Scholar]
  41. R Development Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2012. [Google Scholar]
  42. Robbins CS. Introduction, spread, and present abundance of the house sparrow in North America. Ornithol. Monogr. 1973;14:3–9. [Google Scholar]
  43. Rogers CM, Ramenofsky M, Ketterson ED, Nolan V., Jr Wingfield JC. Plasma corticosterone, adrenal mass, winter weather, and season in nonbreeding populations of Dark-Eyed Juncos (Junco hyemalis hyemalis. Auk. 1993;110:279–285. [Google Scholar]
  44. Rohwer S. Wingfield JC. A field study of social dominance, plasma levels of luteinizing hormone, and steroid hormones in wintering Harris' Sparrows. Z. Tierpsychol. 1981;57:173–183. [Google Scholar]
  45. Romero LM. Physiological stress in ecology: lessons from biomedical research. Trends Ecol. Evol. 2004;19:249–255. doi: 10.1016/j.tree.2004.03.008. [DOI] [PubMed] [Google Scholar]
  46. Romero LM. Reed JM. Collecting baseline corticosterone samples in the field: is under 3 min good enough? Comp. Biochem. Physiol., Part A Mol. Integr. Physiol. 2005;140:73–79. doi: 10.1016/j.cbpb.2004.11.004. [DOI] [PubMed] [Google Scholar]
  47. Romero LM. Romero RC. Corticosterone responses in wild birds: the importance of rapid initial sampling. Condor. 2002;104:129–135. [Google Scholar]
  48. Romero LM. Wikelski M. Corticosterone levels predict survival probabilities of Galápagos marine iguanas during El Niño events. Proc. Natl Acad. Sci. USA. 2001;98:7366–7370. doi: 10.1073/pnas.131091498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Romero LM, Reed JM. Wingfield JC. Effects of weather on corticosterone responses in wild free-living passerine birds. Gen. Comp. Endocrinol. 2000;118:113–122. doi: 10.1006/gcen.1999.7446. [DOI] [PubMed] [Google Scholar]
  50. Romero LM, Cyr NE. Romero RC. Corticosterone responses change seasonally in free-living house sparrows (Passer domesticus. Gen. Comp. Endocrinol. 2006;149:58–65. doi: 10.1016/j.ygcen.2006.05.004. [DOI] [PubMed] [Google Scholar]
  51. Romero LM, Dickens MJ. Cyr NE. The reactive scope model – A new model integrating homeostasis, allostasis, and stress. Horm. Behav. 2009;55:375–389. doi: 10.1016/j.yhbeh.2008.12.009. [DOI] [PubMed] [Google Scholar]
  52. Rubenstein DR. Stress hormones and sociality: integrating social and environmental stressors. Proc. Biol. Sci. 2007;274:967–975. doi: 10.1098/rspb.2006.0051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Rudolph JD. Mexico: a country study. Washington, DC: United States Government; 1985. [Google Scholar]
  54. Schrey AW, Grispo M, Awad M, Cook MB, McCoy ED, Mushinsky HR, et al. Broad-scale latitudinal patterns of genetic diversity among native European and introduced house sparrow (Passer domesticus) populations. Mol. Ecol. 2011;20:1133–1143. doi: 10.1111/j.1365-294X.2011.05001.x. [DOI] [PubMed] [Google Scholar]
  55. Schwabl H, Wingfield JC. Farner DS. Influence of weather on endocrine state and behavior in European Blackbirds (Turdus merula. Z. Tierpsychol. 1985;68:244–252. [Google Scholar]
  56. Smith GT, Wingfield JC. Veit RR. Adrenocortical response to stress in the common diving petrel, Pelecanoides urinatrix. Physiol. Zool. 1994;67:526–537. [Google Scholar]
  57. Toates F. Stress: conceptual and biological aspects. Chichester, UK: John Wiley & Sons; 1995. [Google Scholar]
  58. United Nations Demographic Workbook. 2012. United Nations Statistics Division. copyright United Nations. http://unstats.un.org/unsd/demographic/products/dyb/dyb2.htm (accessed 10 August 2012)
  59. Wassenaar LI, Van Wilgenburg SL, Larson K. Hobson KA. A groundwater isoscape (δD, δ18O) for Mexico. J. Geochem. Explor. 2009;102:123–136. [Google Scholar]
  60. Wayland M, Gilchrist HG, Marchant TA, Keating J. Smits JE. Immune function, stress response, and body condition in arctic-breeding common eiders in relation to cadmium, mercury and selenium concentrations. Environ. Res. 2002;90:47–60. doi: 10.1006/enrs.2002.4384. [DOI] [PubMed] [Google Scholar]
  61. Wilcoxen TE, Boughton RK, Bridge ES, Rensel MA. Schoech SJ. Age-related differences in baseline and stress-induced corticosterone in Florida scrub-jays. Gen. Comp. Endocrinol. 2011;173:461–466. doi: 10.1016/j.ygcen.2011.07.007. [DOI] [PubMed] [Google Scholar]
  62. Wingfield JC. Influences of weather on reproductive function in male Song sparrows, Melospiza melodia. J. Zool. 1985a;205:525–544. [Google Scholar]
  63. Wingfield JC. Influences of weather on reproductive function in female Song sparrows, Melospiza melodia. J. Zool. 1985b;205:545–558. [Google Scholar]
  64. Wingfield JC, Moore MC. Farner DS. Endocrine responses to inclement weather in naturally breeding populations of White-Crowned Sparrows (Zonotrichia leucophrys pugetensis. Auk. 1983;100:56–62. [Google Scholar]
  65. Wingfield JC, Kubokawa K, Ishida K, Ishii S. Wada M. The adrenocortical response to stress in male bush warblers, Cettia diphone: a comparison of breeding populations in Honshu and Hokkaido, Japan. Zoolog. Sci. 1995;12:615–621. [Google Scholar]
  66. Wingfield JC, Maney DL, Breuner CW, Jacobs JD, Lynn S, Ramenofsky M, et al. Ecological bases of hormone-behavior interactions: the “emergency life history stage”. Am. Zool. 1998;38:191–206. [Google Scholar]
  67. Zuckerberg B, Bonter DN, Hochachka WM, Koenig WD, DeGaetano AT. Dickinson JL. Climatic constraints on wintering bird distributions are modified by urbanization and weather. J. Anim. Ecol. 2011;80:403–413. doi: 10.1111/j.1365-2656.2010.01780.x. [DOI] [PubMed] [Google Scholar]
  68. Zuur AF, Ieno EN. Elphick CS. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 2010;1:3–14. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1. Feather CORT values of house sparrows (Passer domesticus) sampled in Mexico plotted as contour lines on the reduced space created by two axes produced by ordination of 37 climate variables.

ece30005-3808-sd1.pdf (88.2KB, pdf)

Figure S2. Serial dilutions (1:1, 1:2, and 1:4) of four house sparrow feather extracts plotted against a 4-parameter standard curve fitted to nine corticosterone (CORT) standards ranging in concentration from 3.125 to 800 pg/100 µL.

ece30005-3808-sd2.pdf (82.1KB, pdf)

Table S1. Results of linear regressions between feather corticosterone (CORTf) of house sparrows sampled from 49 sites across Mexico and measures of monthly temperature, precipitation, and deuterium excess (d-excess).

ece30005-3808-sd3.docx (15.1KB, docx)

Table S2. Recovery efficiencies of five methanol extractions used to extract corticosterone from house sparrow feathers (n = 448).

ece30005-3808-sd4.docx (12.9KB, docx)

Table S3. Inter-assay comparisons of ten radioimmunoassays used to measure corticosterone (CORT) levels of house sparrow feather extracts (n = 448 samples).

ece30005-3808-sd5.docx (14.4KB, docx)

Appendix S1-S2. Complete captions of Figures S1, S2 and Tables S1–S3.

ece30005-3808-sd6.doc (28KB, doc)

Articles from Ecology and Evolution are provided here courtesy of Wiley

RESOURCES