Avian Conservation and Ecology
Home | Archives | About | Login | Submissions | Subscribe | Contact | Search
 ACE Home > Vol. 5, No. 2 > Art. 5

Copyright © 2010 by the author(s). Published here under license by The Resilience Alliance.
Go to the pdf version of this article

The following is the established format for referencing this article:
DesGranges, J., and F. Morneau. 2010. Potential sensitivity of Quebec's breeding birds to climate change. Avian Conservation and Ecology 5 (2): 5. [online] URL: http://www.ace-eco.org/vol5/iss2/art5/
http://dx.doi.org/10.5751/ACE-00410-050205


Research Papers

Potential Sensitivity of Québec's Breeding Birds to Climate Change
 
Sensibilité potentielle des oiseaux nicheurs du Québec aux changements climatiques

Jean-Luc DesGranges 1 and François Morneau 2


1Environment Canada, 2Biological consultant



ABSTRACT


We examined the relationship between climatic factors and the distribution of breeding birds in southern Québec, Canada to identify the species whose distribution renders them potentially sensitive to climate change in the study area. We determined the degree of association between the distribution of 65 breeding bird species (601 presence-absence squares of the Atlas of the Breeding Birds of Québec) and climate variables (212 climatological stations in operation for at least 20 years over the period 1953–1984) by statistically correcting for the effects of several factors that are correlated with bird distribution. Factors considered were the nature and scale of land cover patterns that included vegetation types and landscape characterization, geographical coordinates, and elevation. Canonical Correspondence Analysis (CCA) was used to investigate the effect of climatic variables on breeding bird distribution. Independent variables accounted for a total of 29.1% of the variation in the species matrix. A very large portion of the variance explained by climate variables was shared with spatial variables, reflecting the relationships among latitude, longitude, elevation, and climate. After correcting for the effect of land cover variables, climatic variables still explained 11.4% of the variation in the species matrix, with temperature, i.e., warmer summers and milder winters, having a greater influence than precipitation, i.e., wetter summers. Of the 65 species, 14 appeared to be particularly climate-sensitive. Eight are insectivorous neotropical migrants and six species are at the northern limit of their range in the study area. The opposite is largely true for the eight others; they are practically absent from the southern part of the study area, except for the Dark-eyed Junco (Junco hyemalis), which is widespread there. The White-breasted Nuthatch (Sitta carolinensis) is the only resident species that seemed responsive to climatic variables, i.e., milder winters. Climate warming is thus likely to induce northward shifts for several neotropical migrant species. Many species that currently breed in the northern portion of eastern United States are likely to move northward into Canada. It is thus crucial that sufficient habitats be preserved in Canada to accommodate these future “climate refugees.” Forests in the study area are under management for lumber and therefore, their conservation should receive particular attention.

RÉSUMÉ

La présente étude vise à examiner les liens entre les conditions climatiques et la distribution des oiseaux nicheurs du Québec (Canada) et à dégager les espèces qui paraissent les plus sensibles au climat, de manière à identifier des indicateurs potentiels des incidences du changement climatique sur les écosystèmes. L'approche méthodologique a consisté à déterminer le degré d'association entre la répartition de 65 espèces d'oiseaux nicheurs (601 parcelles de présence-absence de l'Atlas des oiseaux nicheurs du Québec) et des variables climatiques (212 stations climatologiques en opération au moins 20 ans sur la période 1953-1984) en supprimant statistiquement l'effet du maximum de facteurs qui peuvent l'obscurcir. Les facteurs qui ont été considérés sont la nature de l'affectation du sol et son importance, la description du paysage, les coordonnées géographiques et l'altitude. L'analyse canonique des correspondances (CCA) a été utilisée pour estimer l'effet des variables climatiques sur la répartition des espèces d'oiseaux nicheurs. L'ensemble des variables indépendantes expliquait 29,1 % de la variation de la matrice des espèces. Une très grande partie de la variation expliquée par les variables climatiques était partagée avec les variables spatiales traduisant de ce fait l'association entre latitude, longitude, altitude et climat. En supprimant l'effet des variables d'affectation du sol, les variables climatiques expliquaient encore une importante partie de la variation de la matrice des espèces (11,4 %). Une fois supprimé l'effet de l'affectation du sol, les variables décrivant la température (étés plus chauds et hivers moins froids) étaient prédominantes sur celles décrivant les précipitations (étés pluvieux). Lorsqu'on corrigeait pour l’effet des variables d’affectation du sol, la température avait plus d'effet sur la distribution des espèces étudiées que les précipitations. Quatorze (14 des 65) espèces paraissaient plus sensibles que d'autres au climat. La plupart (8) sont des migrateurs néo-tropicaux insectivores. Six de ces espèces atteignent la limite nord de leur aire de reproduction dans la zone d'étude. L'inverse est presque observé pour les huit autres espèces; elles sont pratiquement absentes au sud de la zone d'étude, sauf le Junco ardoisé (Junco hyemalis) qui y est répandu. Seule la répartition de la Sittelle à poitrine blanche (Sitta carolinensis) semblait réagir davantage aux variables climatiques parmi les espèces résidentes (hivers moins froids). Plusieurs espèces qui nichent actuellement dans la portion nord-est des États-Unis pourraient émigrer vers le nord. Il est donc essentiel que suffisamment d’habitats propices soient protégés au Canada pour héberger ces futurs « réfugiés climatiques ». Comme la forêt mixte est le siège d’une intense exploitation forestière, cela milite en faveur d’une attention accrue vis-à-vis du potentiel de conservation des forêts mixtes exploitées.


Key words: bioindicators; breeding bird distribution; climate change; habitat use; sensitivity to climate, Québec



INTRODUCTION


With the continuous increase in atmospheric CO2 and other greenhouse gases since the beginning of the industrial era, the world’s climate has already changed and may change quite considerably before the end of the 21st century (IPCC 2007). The long term management of biodiversity, in terms of both species and ecosystems, requires an adequate understanding of the responses of vegetation and animals to climate change (Kappelle et al. 1999). Changes are being seen in a broad range of taxa, from insects to mammals, and on several continents (UNEP and GRID-Arendal 2009). Birds, for example, are likely to react directly to climate changes such as repeated periods of rain, frost, and heat, and indirectly to changes in the environment that influence such features as food availability, habitat structure, and relationships among organisms. Such responses vary according to each species’ physiological tolerance, and most importantly as nestlings (Hayworth and Weathers 1984, Burton 1995, Thomas et al. 2001, Harrison et al. 2003, Huntley et al. 2006, Hitch and Leberg 2007, Devictor et al. 2008, Virkkala et al. 2008). For birds that are long-distance migrants, climate change may advance the phenology of their breeding areas, e.g., leaf flush, flowering, hatching of pest insect eggs, seed production, etc. (Thomas et al. 2001, Bertin 2008), but the timing of some species’ spring migration relies on endogenous rhythms that are not affected by climate change (Gwinner 1996). Thus, several generations may be required for an optimal adjustment of spring migration to the timing of peak food supply and nestling demand. This mismatching would force poorly adapted species to either advance or accelerate their migration so that they reach their breeding grounds earlier to breed at their period of optimal reproduction (Perrins 1970, Both and Visser 2001, Thomas et al. 2001). The adaptation process may also result in a modification in species’ distribution (Gates 1993, Huntley et al. 2006, Anciaes and Peterson 2006). The synergism of a rapid temperature rise and other stresses, in particular habitat destruction, could easily disrupt the connectedness among species and lead to a restructuring of species assemblages, reflecting different responses among species (Root et al. 2003). It could also lead to numerous extirpations and possibly extinctions (Thomas et al. 2004, Jiguet et al. 2006, Schwartz et al. 2006, Sekercioglu et al. 2008, Lawler et al. 2009).

This study examines the associations between climate and the distribution of breeding birds in Québec. Because birds are highly mobile, their adaptation to climate change may be observed more rapidly than in other organisms, and therefore some bird species may serve as early indicators of the effects of climate change on ecosystems and biodiversity. A quantitative approach was used to provide a statistical description of the relationships between species, climate, and habitat. Pairing data on temporal variation in local climates and bird populations was not possible, therefore the study focused entirely on spatial variation in climate, with the assumption that climatic change across space is equivalent to climatic change through time (Pearson and Dawson 2003). The objective of this study was to determine the relationships between certain climate descriptors and the distribution of breeding birds during the breeding season. A better understanding of these relationships will assist in the development and adaptation of tools for monitoring the effects of climate change on ecosystems and biodiversity.



METHODS

Study area

The study area extends from the southern border of Québec (45°N) to 50° 30'N latitude (~ 500,000 km2; Fig. 1). This area encompasses three main geological regions: the Canadian Shield in the north, the Appalachians in the southeast, and the St. Lawrence Lowlands in between. Elevations range between sea level and approximately 150 m in lowland areas and between 250 m and 750 m on the Shield; however, they are more variable in the Appalachians, ranging from less than 100 m to 1268 m, with 500 m being the average. Six climate types (Litynski 1984) are found in the study area. The inland, and largest portion of the study area, has a continental moderate climate. The St. Lawrence Lowlands and the North Shore and Gaspé coastal areas occur in the continental moderate subhumid and continental subpolar subhumid zones, respectively. Other climate types are found only in small enclaves. Vegetation formations in the study area comprise, from south to north, sugar maple (Acer saccharum), balsam fir (Abies balsamea), and black spruce (Picea mariana) forests. Tundra-type formations are found only at the highest elevations. Most of the populated areas and agricultural lands are in the region dominated by sugar maple stands. The northern third of the study area is characterized by the black spruce-feather moss formation (NRC Atlas of Canada).

Bird, vegetation, and climate data

Dependent variable

Given our incomplete knowledge of the local nesting distribution of several of the breeding bird species of southern Québec, we chose to restrict our analysis to an array of candidate species that were found breeding (presence/absence) in at least 25%, but not more than 75%, of a selection of 601 10 km x 10 km squares well surveyed for the Atlas of the Breeding Birds of Southern Québec (Gauthier and Aubry 1996). Moreover, to maximize the probability of detecting potential climatic influences, we limited our analysis to a subset of 65 species that were selected on the basis of their expected vulnerability to climate extremes. We hypothesized that bird sensitivity to climate is associated with their life history traits and physical characteristics (Table 1). In order to verify this hypothesis we developed a sensitivity index based on 10 life history traits and physical characteristics, e.g., weight, breeding distribution, migration, foraging, etc; see Appendix 1). Of the 65 species, 18 were predicted to be of limited sensitivity to climate, 27 were predicted to be sensitive, and 20 were predicted to be very sensitive to climate (Appendix II; see Morneau et al. 1998 for details).

Independent variables

Climate data were obtained from Environment Canada’s Meteorological Service for the period between 1954 and 1983, thus ending the year before the compiling of data for the Atlas of the Breeding Birds of Québec began. During this period, 212 climatological stations were in operation for at least 20 years, mostly in the settled areas of the province. The exhaustive information provided by the stations was used to calculate the values for a series of daily and monthly climatic variables to which birds are likely to be sensitive, such as the frequency of occurrence of a specific temperature in relation to a specific threshold, the number of degree-days in relation to a temperature threshold, the type of precipitation and frequency, etc. Climatic variable selection was based on scientific literature. A variable was selected if significant correlation or association was found between it and any bird species (Table 2). Other climatic variables likely to affect birds, such as wind, insulation, and solar radiation, were measured at too few stations (<10%) to be used in the analyses.

The number of squares used in the analyses was based on the number of Atlas squares that could be associated with climatic data, and to determine this we used a 25 km range from the centre of an Atlas square as the maximum distance from a climatological station at which climatic data can be inferred to that square. Theoretically, if the topography remains fairly similar within this 25 km range, a climatological station located in the centre of an Atlas square could be used to infer the climatic data for 13 other squares (see Morneau et al. 1998). In practice, not all squares were adequately covered during the bird atlas field work and not all were located ≤ 25 km from a station; in all there were 601 squares that corresponded to these criteria and 171 climatological stations for which the topography and altitude were similar to the Atlas squares with which they are associated (see Morneau et al. 1998). Most of the squares that were retained occur along a southwest to northeast axis that runs parallel to the St. Lawrence River (Fig. 1).

Other environmental variables used were related to habitat, land use, landscape characteristics, and the spatial distribution of the Atlas squares. The characteristics and area of the habitats (10 classes) in each square were obtained from National Oceanic and Atmospheric Administration (NOAA) satellite images (pixels of 1 km²) of Québec obtained in 1989. FRAGSTATS software (McGarigal and Marks 1994) was used to extract additional information from the NOAA images. Specifically, five variables were selected: number of patches, patch size standard deviation, patch richness, Simpson diversity index, and contagion index. NOAA images were also used to determine the elevation of Atlas squares. Elevation was divided into classes and the area of each class was calculated for each Atlas square. To reduce the number of variables, only three variables were retained: lowest elevation, highest elevation, and mean elevation for each square. Each square’s geographic location was described using the latitudinal and longitudinal coordinates of the southwest corner.

To meet the requirements of the statistical analyses, we included only those variables that were as independent as possible. Variable selection was made using the Kendall rank correlation test and was conducted separately for the three groups of variables: climatic, spatial, and vegetation/landscape. Variables representing climate extremes were generally favored. In all, 10 significant climatic variables were retained for the following analyses (Table 3).

Data analysis

Canonical Correspondence Analysis (CCA; Ter Braak 1988) and the method used by Borcard et al. (1992) for partitioning the variance of species abundance were used to determine the effects of climatic variables on the breeding distribution of birds. CCA is a constrained ordination technique that allows identifying which environmental variables, i.e., climatic, land cover, and spatial variables, drive bird species distributions in southern Québec. An inertia value, associated with each dimension, expresses the percentage of the total variance in species distribution attributable to each dimension. The total variation in the species matrix can be divided among eight sources of variation with respect to the three sets of variables taken into account: a) variation due to spatial variables alone, b) variation due to climatic variables alone, c) variation due to land cover variables alone, d) spatial variation shared with climatic variation, e) climatic variation shared with land cover variation, f) spatial variation shared with land cover variation, g) variation shared among all three sets of variables, and h) variation not explained by the independent variables retained.

An initial series of three CCAs were performed (with the CANOCO software package; Ter Braak 1988) on the species matrix, each taking into account only one of the three sets of environmental variables, i.e., climate, land cover, space, in an independent matrix. Environmental variables that were not found to be relevant in the first three CCAs were excluded from subsequent CCAs. A fourth CCA included in a single matrix all the independent variables found to be significant during the first three CCAs (a + b + c + d + e + f + g), allowing the portion of the total variation in the species matrix associated with these variables to be determined (see Morneau et al. 1998 for the combinations of analyses conducted).

In partial CCA, the coordinates of the species along a canonical axis provide a ranking along a given environmental variable (Legendre and Legendre 1998). A series of six partial CCAs were carried out to determine the percentage of the variation in the species matrix accounted for by each combination of two of the three sets of predictors, one set being the independent variables, the other set being the covariables. These additional analyses served to determine the percentage of variation in the species matrix associated with the covariable matrix and the percentage explained by the independent matrix not already explained by the covariable matrix. Subsequently, a series of linear combinations were calculated using linear algebra to determine the percentage of the variation in the species matrix associated with each potential source of variation (a to g above). The amounts of variation explained by the seven components [a] to [g], as well as the amount of unexplained variation [h], were obtained by subtractions from these results.



RESULTS

Significant independent variables

The results of the CCA carried out with climatic variables showed that all 10 variables were significant. The canonical axes for this CCA account for 21.8% of the variation in the species matrix. Axis 1 accounts for 6/7 of the variation explained by the ten canonical axes, corresponding to the ten variables taken into account. The climatic variables that show the strongest correlation with Axis 1 consist of the mean annual degree-days ≥ 10°C in May, June, and July (DD+10), the mean number of days with rainfall ≥ 10 mm and T ≤ 10°C in June and July (MNDR), and the mean annual temperature from December to February (MTDF; Table 4). Thus, the first axis constitutes a temperature gradient. In comparison, Axis 2 accounts for 1/14 of the variation explained by the canonical axes. Two variables were correlated with this axis: the mean annual rainfall in June and July (MARJJ) and the mean annual rainfall, excluding September to November (MAR). Axis 2 thus represents a gradient of precipitation in the form of rain.

The results of the CCA with land cover variables showed that 9 of the 13 land cover variables were significant (Table 4). This CCA accounts for 14.9% of the variance explained in the species matrix; Axis 1 accounts for 7/10 of the variation explained by the nine canonical axes, corresponding to the nine variables taken into account. Four land cover variables were correlated most strongly with this axis: the area covered by agricultural land (AGRI) and human population (POPU), which were inversely correlated, and the area covered by mixed forest (MIX) and by coniferous forest (CONI), which were positively correlated. Axis 1, therefore, appears to represent a gradient of urbanization or area covered by forest. In comparison, Axis 2 accounts for 1/7 of the explained variance. In particular, two variables were correlated with this axis: area covered by deciduous forests (DECI) and area covered by water (HYDR), both of which were inversely correlated.

All three spatial variables were significant in the CCA carried out with these variables. The canonical axes in this analysis account for 22.7% of the variation in the species matrix. Axis 1 accounts for most of the variation (6/7) explained by the three canonical axes, corresponding to the three variables taken into account. The spatial variable most strongly correlated with this axis was latitude (LATI; Table 4); Axis 1 therefore represents a latitudinal gradient. Axis 2 accounts for 1/10 of the variation explained by the canonical axes. It represents an altitudinal gradient, altitude (MOALT) being the variable most strongly correlated with the axis.

In the fourth CCA, the independent matrix consisted of a combination of climatic, spatial, and land cover variables. The canonical axes were found to be associated with 29.1% of the variation in the species matrix. Axis 1 accounted for 7/10 of the variation in the species matrix explained by the canonical axes, and Axis 2 for 1/10. A strong correlation was found between the climatic variables and Axis 1, indicating that these variables best explain the variation in the species matrix. Axis 1 thus represents a gradient of temperature and precipitation, which is mainly latitudinal and to a lesser degree longitudinal (Table 4). In general, this indicates that the highest values for temperature variables (mainly DD+10 and MTDF) and total rainfall (MAR) were recorded in the southern and western parts of the study area. Conversely, the lowest temperatures (DD-25, MNDR) and the highest snowfall (MAS, MASMJ) were associated with the northernmost and easternmost regions. Mixed and coniferous forests are found in the north and east, whereas agricultural areas, human population, and urban areas are concentrated in the south and west. The variables most strongly correlated with Axis 2 were land cover and altitude. Moreover, altitude and the area covered by deciduous forests were negatively correlated with Axis 1 and the area covered by water was positively correlated. High values for deciduous forest area are therefore associated with higher altitudes and high water values with lower altitudes. These two variables seem to be structured by altitude rather than latitude.

Partitioning the variance

The way in which the variation in the species matrix was partitioned suggests that there is a strong association between the three sets of variables (Fig. 2). The variation of the species matrix explained and shared by the three sets of independent variables accounted for 29.1% of the explained variance. A very large portion of the variation explained by climatic variables (5/6) is shared at least partially with spatial variables. Spatial variables accounted for most of the variation in the species matrix, while land cover variables accounted for the least. The fraction of the species-matrix variation not explained by any of the environmental variables used is 70.9%. After correcting for the effect of land cover variables, climatic variables still explain a significant portion (1/9) of the variation in the species matrix. The CCA carried out with climatic variables and correcting for the effect of land cover shows that Axis 1 accounts for 4/5 of the variance explained by the canonical axes. The climatic variables most strongly correlated with this axis are DD+10 (r = - 0.77) and MNDR (r = + 0.68), or the same variables as in the CCA without covariates.

Evaluation of bird species’ sensitivity to climate

Table 5 presents the results of the CCAs carried out, firstly with climatic variables and land cover variables without covariates and secondly with climatic variables corrected for the effect of land cover. The results of the CCA using climatic variables without covariates as independent variables reveal the species that appear to be the most sensitive to climate. Table 6 lists the species for which an important part of the variance (> 33%) was accounted for by the first axis. Species are presented in the sensitivity categories identified during the first stage of the study according to physiological and ecological criteria (see Table 1 and Appendix 1). For these species, variation associated with climate was less influenced by land cover variables.

A comparison of the CCA results, using climatic variables and correcting for the effect of land cover (Table 6), with the evaluation of potential climatic sensitivity, and using criteria taken from the literature (Table 1) shows that our qualitative assessments (Appendix II) are generally supported by the quantitative results. Using a threshold of 20% of the variance on Axis 1 (not shown here), 17 of the 18 (94%) species thought not to be sensitive to climate were effectively found to be weakly associated with climatic variables. Similarly, 38 of the 47 (81%) suspected sensitive species were effectively associated with climatic variables.

To identify species that could be used as indicators of climate change, we determined which species were most strongly correlated with climatic variables, not correcting for the effect of land cover. In nature, species evolve in an environment affected both by climatic conditions and habitat; hence the CCA was performed with climatic variables without covariates (Table 5). Out of the 22 species for which the percentage of explained variation was equal to or greater than that of the entire species matrix for all variables, there were 14 species for which 16% or more of the variation could be explained by climate, correcting for the effect of land cover (Table 6). Among these species, six are at the northern limit of their breeding range in the study area. The inverse can be observed for almost all of the remaining species, that is, they are nearly absent from the southern part of the study area, except for the Dark-eyed Junco (Junco hyemalis), which is frequent there. Of the 14 species, eight are neotropical migrants, five are short-distance migrants and only one is a year-round resident, the White-breasted Nuthatch (Sitta carolinensis; Table 6). The distribution of these species along the first two axes of the CCA with climatic variables shows that six species are correlated above all with the mean annual degree-days ≥ 10°C in May, June, and July (DD+10 in Fig. 3). Five of these are migrants, Great Crested Flycatcher (Myiarchus crinitus), House Wren (Troglodytes aedon), Eastern Meadowlark (Sturnella magna), Warbling Vireo (Vireo gilvus), and Baltimore Oriole (Icterus galbula), and are likely associated with warm summers. In contrast, the White-breasted Nuthatch, a resident species, is probably primarily associated with milder winters (MTDF in Fig. 3). The other species, Ruby-crowned Kinglet (Regulus calendula), Lincoln's Sparrow (Melospiza lincolnii), Swainson’s Thrush (Catharus ustulatus), Tennessee Warbler (Oreothlypis peregrina), Magnolia Warbler (Dendroica magnolia), Bay-breasted Warbler (Dendroica castanea), Wilson’s Warbler (Wilsonia pusilla), and Dark-eyed Junco, all boreal forest species, appear to be more correlated with cooler and wetter summers (Fig. 3).



DISCUSSION

The relative importance of the bioclimate envelope

In this study, climatic, spatial, and land cover factors are strongly associated with one another. The portion of variation in the species matrix explained by both climatic variables and habitat variables suggests that climate may indirectly influence bird distribution by affecting vegetation. Although this is a widely accepted hypothesis (Hayworth and Weathers 1984), these links may also reflect the simultaneous influence of climate on vegetation and bird distribution. In either case, the link between climate and bird distribution is clear and most of the explained variation (21.7%) is probably due to climate in some way. Root (1988b) found that in winter, both climate and vegetation had an effect on the distribution of certain species of birds. In a Tennessee study using atlas data, Nicholson (1991) was not able to find much of a link between temperature and precipitation and species richness in the squares, whereas habitat variables were found to have a greater effect on determining the number of species. According to Telleria et al. (1992), climate is the ultimate determinant of the theoretical number of species that can occupy a given location, because it may directly affect productivity and it plays an indirect role in other cases. The actual number of species in a given area, therefore, is always less than the number that is theoretically possible, i.e., limited by climate alone, because of proximate causes such as habitat components. Consequently, the scale of observation is crucial in separating the effect of climate from that of habitat; too large a scale favors climate over habitat as an explanation of the variation in species richness in a given area (Telleria et al. 1992). To sum up, climate and habitat work together to affect the distribution of bird species. It is already apparent that the scale at which current bioclimatic studies are addressed is of fundamental importance, with effects on the distribution of species being most influential at regional to global scale (Pearson and Dawson 2003). Unfortunately, characterizations of more complex relationships between climate change, land cover change, and Québec bird assemblages are presently limited by a lack of process understanding, data availability at a higher resolution, and inherent climate scenario uncertainties.

Our study’s results suggest that the associations between bird species and climate in the analyses are correct. Nearly 30% in the variation in the distribution of the 65 breeding birds of southern Québec selected for the analysis can be explained by climate, land cover, and spatial variables. The fact that this percentage was not higher may be because of other factors aside from the geographical constraints discussed above, including the accuracy of the Atlas data and the resolution of variables. Because of the limited resolution of the NOAA images (pixels of 1 km ²), the habitat variables used are highly related to the landscape structure variables for most species, which may account for part of the unexplained variation.

Similarly, it is important to keep in mind that the relationships documented here describe the link between the distribution of birds and environmental variables at a given point in time, i.e., 1984–89. The actual situation, however, is dynamic. A recent analysis of the Canadian Breeding Bird Survey (BBS) data from 1967 to 2000 (Downes and Collins 2003) has indicated that the populations of the Swainson’s Thrush and Eastern Meadowlark have declined, those of the Magnolia Warbler and Warbling Vireo have increased, while the population of the Winter Wren (Troglodytes troglodytes) showed precursory signs of decline. Furthermore, certain thrushes and some warblers have been decreasing in numbers over the last 15 years, whereas others, like the Purple Finch (Carpodacus purpureus), House Finch (Carpodacus mexicanus), and Northern Cardinal (Cardinalis cardinalis) have increased significantly.

Population changes could have been caused by a variety of factors including habitat loss in the winter range, proliferation of bird feeders, etc. Whatever the factors involved, they might either have reduced or increased the climatic effects on breeding birds. Despite these problems, we found significant links between the three types of independent variables and the bird species. Our results show that most of the variation in the species matrix due to climatic variables is shared with the variation due to spatial values. Because geographic coordinates and altitude have a strong influence on climate but not the inverse, it can be concluded that part of the portion of the variation in the species matrix shared by these two variables corresponds to spatially-structured climatic variables (Borcard et al. 1992). Therefore, over 10% of the variation in the distribution of species is probably because of climate alone. Given the difficulty of isolating climatic effects on bird distribution from those of other environmental variables in endothermic vertebrates living in nonextreme conditions (Telleria et al. 1992), this is an important finding. Moreover, Johnson (1994) and Currie (2001) found that the contemporary patterns of bird distribution in the conterminous United States covary strongly with summer temperature and moisture. The portion of the variation in the species matrix supposedly explained strictly by climate (11.4%) may be because of the indirect effects of climate on birds. For example, climate may affect birds by influencing insect development or vegetation, variables that were not measured in this study. This is especially plausible given the fact that the main climatic variable used was the number of degree-days greater than or equal to 10°C. It is well known that higher temperatures favor insect activity and development (Gates 1993). For example, flying insects become more active as the temperature rises, which in turn increases the capture success rate by birds (Rodenhouse 1992). Therefore, it is not surprising to find a link between temperature and distribution in the Great Crested Flycatcher, a species that hawks for insects. Furthermore, climate tends to be most limiting on bird distribution during extreme climatic events (Root 1988b); the same is true for dramatic changes in the number of species (Telleria et al. 1992). The use of climatic data measured over a 30-year period in conjunction with the Atlas distribution data (Gauthier and Aubry 1996), which was measured over a shorter period of time, might not have provided a faithful reflection of extreme events in the data set.

Projected climate induced avifaunal change in southern Québec and management

In this study, 14 out of 65 (22%) bird species appear to be sensitive to climate change. Our results show that the White-breasted Nuthatch may be limited by winter temperatures while the Great Crested Flycatcher, House Wren, Eastern Meadowlark, Warbling Vireo, and Baltimore Oriole may be limited by the number of degree-days. The other species, including Wilson’s Warbler, the Bay-breasted Warbler, Tennessee Warbler, Lincoln's Sparrow, and Swainson’s Thrush, require a cool, wet climate. Matthews et al. (2004) projected dramatic shifts northward in the breeding distribution of several northern U.S. bird species, including most of the species just mentioned, under warming climatic conditions or indirectly through dependence on tree species that themselves are limited by warming conditions such as balsam fir, yellow birch (Betula alleghaniensis), sugar maple, red maple (Acer rubrum), and striped maple (Acer pensylvanicum; McKenney et al. 2007). In the case of neotropical migrants, many species that currently breed in the northern portion of eastern United States are likely to move northward into Canada. This is especially true for those species that are associated with the presence of coniferous trees inside the mixed wood forest of north-eastern North America (Matthews et al. 2004). It is thus crucial that enough habitats will be preserved in Canada to accommodate these future “climate refugees”. Since mixed wood forests are often exploited for lumber, additional attention should be placed on their conservation.

From a global warming perspective, birds may be used as early bioindicators of climate change. Birds are highly mobile organisms and can colonize new and suitable areas more quickly than such organisms as trees. They are easy to observe and have generated a large amount of data covering long periods of time. However, it must be remembered, as Morrison (1986) noted, that birds are probably better indicators of secondary changes, i.e., the repercussions brought about by changing conditions, than of primary ones, which act directly on the survival of individuals or the abundance of populations. Although the variables most often taken into account in studies dealing with the links between environmental change and birds are changes in density, abundance, and distribution of avian populations (Temple and Wiens 1989), these may not always be the most appropriate variables. Bird distribution can be an effective indicator of climatic changes only for species that are affected directly by climatic changes. Species that are indirectly affected, that is through habitat or other biotic changes, will react with a delay depending on the speed of the modifications. Therefore, these species are less likely to be effective indicators of climatic changes. Hence, natality, mortality, and dispersal rates, which reflect more directly the bird’s behavioral and physiological responses to environmental change, would be better choices as bioindicator criteria than distributional characteristics. Identifying decline-promoting factors allows scientists to infer mechanisms responsible for observed declines in wild bird populations facing global change, and by doing so allows for a more pre-emptive approach to conservation planning (Jiguet et al. 2007). To the extent that appropriate factors are taken into account, birds are an ideal way of studying the effect of anticipated climate change (Macdonald 1992).




RESPONSES TO THIS ARTICLE


Responses to this article are invited. If accepted for publication, your response will be hyperlinked to the article. To submit a response, follow this link. To read responses already accepted, follow this link.




ACKNOWLEDGMENTS

We would like to thank the managers of the Atlas of the Breeding Birds of Québec data bank, Jean Gauthier, Yves Aubry, and Gilles Falardeau of the Canadian Wildlife Service for giving us permission to use the data files and extract those we needed. We also thank Gérald Vigeant, of the Atmospheric Environment Service, for the use of climatological data and for making his employees available during data extraction. Daniel Lambert and Mario St-Georges from GREBE Inc. did the statistical analyses while Daniel Borcard helped us with the interpretation of the CCA results. David Currie, Mélanie-Louise Le Blanc, Don Thomas, Marc-André Villard, and Richard Cutter provided comments on the previous version of the manuscript. This project could not have been carried out without the financial assistance of the Canadian Wildlife Service (CWS), Québec Region, the Environmental Innovation Program of Supply and Services Canada, the Atmospheric Environment Service (AES), Scientific Services Division, Québec Region, and the ecosystems health component of the “Reducing the Uncertainties” program of the Great Lakes—St. Lawrence Basin (GLSLB) Project.



LITERATURE CITED


Anciaes, M., and A. T. Peterson. 2006. Climate change effects on neotropical manakin diversity based on ecological niche modelling. Condor 108:778-791.

Bednekoff, P. A., B. Biebach, and J. Krebs. 1994. Great Tit fat reserves under unpredictable temperatures. Journal of Avian Biology 25:156-160.

Bertin, R. I. 2008. Plant phenology and distribution in relation to recent climate change. Journal of Torrey Botanical Society 135:126-146.

Bock, C. E., and L. W. Lepthien. 1974. Winter patterns of bird species diversity and abundance in the United States and southern Canada. American Birds 28:556-562.

Borcard, D., P. Legendre, and P. Drapeau. 1992. Partialling out the spatial component of ecological variation. Ecology 73:1045-1055.

Both, C., and M. E. Visser. 2001. Adjustment to climate change is constrained by arrival date in a long-distance migrant bird. Nature 411:296-298.

Brittingham, M. C., and S. A. Temple. 1988. Impacts of supplemental feeding on survival rates of black-capped chickadees. Ecology 69:581-589.

Burton, J. F. 1995. Birds and climate change. Christopher Helm - A & C Black, London, UK.

Currie, D. J. 2001. Projected effects of climate change on patterns of vertebrate and tree species richness in the conterminous United States. Ecosystems 4:216-225.

Cyr, A., and J. Larivée. 1995. Atlas saisonnier des oiseaux du Québec. Les Presses de l'Université de Sherbrooke et la Société du Loisir ornithologique de l'Estrie, Sherbrooke, Québec, Canada.

Devictor, V., R. Julliard, D. Couvet, and F. Jiguet. 2008. Birds are tracking climate warming, but not fast enough. Proceedings of the Royal Society B Biological Sciences 275:2743-2748 doi:10.1098/rspb.2008.0878.

Downes, C. M., and B. T. Collins. 2003. The Canadian breeding bird survey, 1967-2000. Canadian Wildlife Service Progress Notes 219:1-39.

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The birder's handbook. Fireside Book, Simon and Schuster, New York, New York, USA.

Gates, D. M. 1993. Climate change and its biological consequences. Sinauer and Associates, Sunderland, England.

Gauthier, J., and Y. Aubry, editors. 1996. The breeding birds of Québec: atlas of the breeding birds of southern Québec. Association québécoise des groupes d'ornithologues, Province of Québec Society for the Protection of Birds, Canadian Wildlife Service, Environment Canada, Québec Region, Montreal, Québec, Canada

Gwinner, E. 1996. Circannual clocks in avian reproduction and migration. Ibis 138:47-63.

Harrison, P. A., D. P. Vanhinsbergh, R. J. Fuller, and P. M. Berry. 2003. Modelling climate change impacts on the distribution of breeding birds in Britain and Ireland. Journal of Nature Conservation 11:31-42.

Hayworth, A. M., and W. W. Weathers. 1984. Temperature regulation and climatic adaptation in Black-billed and Yellow-billed Magpies. Condor 86:19-26.

Hitch, A. T., and P. L. Leberg. 2007. Breeding distribution of North American bird species moving north as a result of climate change. Conservation Biology 21:534-539.

Holland, P. K., and D. W. Yalden. 1991. Population dynamics of common sandpipers Actitis hypoleucos breeding along an upland river system. Bird Study 38:151-159.

Huntley, B., Y. C. Collingham, R. E. Green, G. M. Hilton, C. Rahbek, and S. G. Willis. 2006. Potential impacts of climatic change upon geographical distributions of birds. Ibis 148:8-28.

Intergovernmental Panel on Climate Change (IPCC). 2007. Summary for Policymakers. In Climate change 2007: the physical science basis. S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor, H. L. Miller, editors. Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK.

Järvinen, A. 1994. Global warming and egg size of birds. Ecography 17:108-110.

Jiguet, F., R. Julliard, C. D. Thomas, O. Dehorter, S. E. Newson, and D. Couvet. 2006. Thermal range predicts bird population resilience to extreme high temperatures. Ecology Letters 9:1321-1330.

Jiguet, F., A.-S. Gadot, R. Julliard, S. E. Newson, and D. Couvet. 2007. Climate envelope, life history traits and the resilience of birds facing global change. Global Change Biology 13:1672-1684.

Johnson, N. K. 1994. Pioneering and natural expansion of breeding distributions in western North American birds. Studies in Avian Biology 15:27-44.

Kappelle, M., M. M. I. Van Vuuren, and P. Baas Magret. 1999. Effects of climate change on biodiversity: a review and identification of key research issues. Biodiversity and Conservation 8:1383-1397.

Ketterson, E. D., and V. Nolan Jr. 1976. Geographic variation and its climatic correlates in the sex ratio of eastern-wintering dark-eyed juncos (Junco hyemalis hyemalis). Ecology 57:679-693.

Lawler, J. J., S. L. Shafer, D. White, P. Kareiva, E. P. Maurer, A. R. Blaustein, and P. J. Bartlein. 2009. Projected climate-induced faunal change in the Western Hemisphere. Ecology 90:588-597.

Legendre, P., and L. Legendre. 1998. Numerical Ecology. Second edition. Elsevier Science BV, Amsterdam, The Netherlands.

Lustick, S., and J. Adams. 1977. Seasonal variation in the effects of wetting on the energetics and survival of starlings (Sturnus vulgaris). Comparative Biochemistry and Physiology 56A:173-177.

Litynski, J. 1984. Classification numérique des climats mondiaux. Organisation météorologique mondiale, Programme climatologique mondial, PCM/WCP-63. Genève Suisse.

Macdonald, I. A. W. 1992. Vertebrate populations as indicators of environmental change in Southern Africa. Transactions Royal Society of South Africa 48:87-122.

Matthews, S., R. O’Connor, L. R. Iverson, and A. M. Prasad. 2004. Atlas of climate change effects in 150 bird species of the Eastern United States. Gen. Tech. Rep. NE-318, U.S. Department of Agriculture, Forest Service, Northeastern Research Station, Newton Square, Pennsylvania, USA.

McGarigal, K., and B. J. Marks. 1994. FRAGSTATS. Spatial pattern analysis. Program for quantifying landscape structure. Version 2.0.

McKenney, D. W., J. H. Pedlar, K. Lawrence, K. Campbell, and M. F. Hutchinson. 2007. Potential impacts of climate change on the distribution of North American trees. BioScience 57:939-948.

Morneau, F., D. Lambert, M. St-Georges, and J.-L. DesGranges. 1998. Évaluation de la vulnérabilité des oiseaux nicheurs du Québec au changement climatique. G.R.E.B.E. for Environment Canada, Québec Region. Canadian Wildlife Service Technical Report Series 299:1-50.

Morrison, M. L. 1986. Bird populations as indicators of environmental change. Current Ornithology 3:429-451.

National Geographic Society. 1987. Guide d'identification des oiseaux de l'Amérique du Nord. Éditions Marcel-Broquet, Laprairie, Québec, Canada.

Nicholson, C. P. 1991. Geographic patterns in species occurrence of Tennessee's breeding birds. Journal of Tennessee Academy of Sciences 66:195-199.

Odum, E. P., and F. A. Pitelka. 1939. Storm mortality in a winter Starling roost. Auk 56:451-455.

Pearson, R. G., and T. E. Dawson. 2003. Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Global Ecology and Biogeography 12:361-371.

Perrins, C. M. 1970. The timing of birds’ breeding seasons. Ibis 112:242-255.

Peterson, R. T. 1980. Field guide to the birds east of the Rockies. Houghton Mifflin, Boston, Massachusetts, USA.

Rodenhouse, N. L. 1992. Potential effects of climatic change on a neotropical migrant landbird. Conservation Biology 6:263-272.

Root, T. L. 1988a. Environmental factors associated with avian distributional boundaries. Journal of Biogeography 15:489-505.

Root, T. L. 1988b. Energy constraints on avian distributions and abundances. Ecology 69:330-339.

Root, T. L., J. T. Price, K. R. Hall, S. S. Schneider, C. Rosenzweig, and J. A. Pounds. 2003. Fingerprints of global warming on wild animals and plants. Nature 421:57-60.

Schwartz, M. W., L. R. Iverson, A. M. Prasad, S. N. Matthews, and R. J. O’Connor. 2006. Predicting extinctions as a result of climate change. Ecology 87:1611-1615.

Sekercioglu, C. H., S. H. Schneider, J. P. Fay, and S. R. Loarie. 2008. Climate change, elevational range shifts, and bird extinctions. Conservation Biology 22:140-150.

Telleria, J. L., T. Santos, A. Sanchez, and A. Galarza. 1992. Habitat structure predicts bird diversity distribution in Iberian forests better than climate. Bird Study 39:63-68.

Temple, S. A., and J. A. Wiens. 1989. Bird populations and environmental changes: can birds be bio-indicators? American Birds 43:260-270.

Ter Braak, C. J. F. 1988. CANACO - a fortran program for canonical community ordination. Microcomputer power, Ithaca, New York, USA.

Thomas, C. D., A. Cameron, R. E. Green, M. Bakkenes, L. J. Beaumont, Y. C. Collingham, B. F. N. Erasmus, M. Ferreire de Siqueira, A. Grainger, L. Hannah, L. Hugues, B. Huntley, A. S. van Jaarsveld, G. F. Midgley, L. Miles, M. A. Ortega-Huerta, A. T. Peterson, O. L. Phillips, and S. E. Williams. 2004. Extinction risk from climate change. Nature 427:145-148.

Thomas, D. W., J. Blondel, P. Perret, M. M. Lambrechts, and J. R. Speakman. 2001. Energetic and fitness costs of mismatching resource supply and demand in seasonally breeding birds. Science 291:2598-2600.

United Nations Environment Programme (UNEP), and GRID-Arendal. 2009. Climate in peril: a popular guide to the latest IPCC reports. UNEP/GRID-Arendal, Nairobi, Kenya and Arendal, Norway.

Virkkala, R., R. K. Heikkinen, N. Leikola, and M. Lueto. 2008. Projected large-scale range reductions of northern-boreal land bird species due to climate change. Biological Conservation 141:1343-1353.


Address of Correspondent:
Jean-Luc DesGranges
Wildlife Research
Sainte-Foy, Quebec, Canada G1V 4H5
jean-luc.desgranges@ec.gc.ca
logo

Up

Home | Archives | About | Login | Submissions | Subscribe | Contact | Search