Research Paper

INTRODUCTION

The populations of many avian species have been persistently declining over the past century, most of which are either directly or indirectly the consequence of human activity (Pimm et al. 2006, Szabo et al. 2012), indicating a need to better understand the drivers of species’ declines to develop targeted conservation efforts and slow or stop declines. The Acadian Flycatcher (Empidonax virescens) is a Neotropical, migrant songbird that faces multiple environmental stressors across its nesting range, which have led to varying levels of endangerment across parts of this range (Christy 1942, Martin 2007, Environment Canada 2012, Sauer et al. 2017). The species winters in Central and South America and migrates north to breed throughout much of the eastern United States and the Carolinian region of Ontario, Canada (Cheskey and The Twelve Mile Creek Headwaters IBA Steering Committee 2003, Allen et al. 2017). Acadian Flycatchers are considered endangered at the northern range boundary in Ontario (Martin 2007, COSEWIC 2010), and threatened in the states of Minnesota and Wisconsin along the western edge of their nesting range (WDNR 2018, MDNR 2024). In Wisconsin, this is thought to be the result of habitat loss and the spread of tree disease (WDNR 2018), but the Acadian Flycatcher has consistently been rare in Minnesota, the westernmost extent of their range (MDNR 2024).

In Ontario, a combination of threats, including habitat loss, tree disease, invasive species, and nest predation and parasitism have been indicated as threats to the Acadian Flycatcher (COSEWIC 2010, Environment Canada 2012), currently home to an estimated 27–35 breeding pairs (Martin 2007). Reproductive success for the Acadian Flycatcher is believed to be dependent on habitat quality and landscape characteristics, such as fragmentation (Hoover et al. 2006) and proximity to or extent of urban development (Fauth and Cabe 2005, Bakermans and Rodewald 2006, Rodewald 2009, Shustack and Rodewald 2010). This potentially explains why in Ontario, where large patches of high-quality undeveloped habitat are sparse, the average productivity may be lower than in other regions (COSEWIC 2010, Allair et al. 2015). A survey conducted from 2011 to 2015 found that the number of young fledged per nest range from 0.83 to 1.44 in Ontario (Allair et al. 2015), whereas a study in Virginia reported an average of 1.8 successful female fledglings per nest (Fauth and Cabe 2005). To date, peer reviewed research on the status of the small population of Acadian Flycatchers in Ontario has been limited. Here, we use community science (also called citizen science or participatory science) and publicly available land use data to investigate the habitat quality where Acadian Flycatchers are observed across the nesting range, including Ontario.

The Acadian Flycatcher is described as a habitat specialist, particularly in the northern portion of its range, thought to selectively breed in deciduous forests or wooded wetlands with a closed canopy and open understory (Christy 1942, Martin 2007) with low vegetation density (Bakermans and Rodewald 2006). They have been observed to strongly prefer nesting in eastern hemlock (Tsuga canadensis; Becker et al. 2008, Allen et al. 2009), though they also nest in other trees that meet the canopy and understory requirements such as American beech (Fagus grandifolia), or the middle story species flowering dogwood (Cornus florida; Martin 2007, COSEWIC 2010). Further, the species has been shown to avoid habitat edges (Hoover et al. 2006, though see Hazler et al. 2006) and prefer large habitat patches (Environment Canada 2012). For example, unpublished data from B. Woolfenden and B. Stutchbury reported by Environment Canada (2012) suggest that more than 90% of Acadian Flycatcher nesting sites in Ontario are in patches of forest > 25 ha and 56% are in patches > 100 ha.

Invasive species can disrupt forest canopy and understory requirements, food availability, and the risk of nest predation and parasitism, potentially rendering habitats with otherwise preferred conditions unsuitable (e.g., Stoleson and Finch 2001, Rodewald 2009, Schneider and Miller 2014, Filek et al. 2018). According to unpublished data from D. Martin reported by Environment Canada (2012), understories dominated by multiflora rose (Rosa multiflora) and garlic mustard (Alliaria petiolata) appear to deter Acadian Flycatcher nesting, perhaps because invasive vegetation can dominate and overfill understories making foraging more difficult or reducing food availability because of different insect communities supported by invasive vegetation (Litt and Steidl 2010, Massé and Vulinec 2010, Schneider and Miller 2014). Invasive tree pests and pathogens can also degrade the forest canopy and directly impact nesting trees. Hemlock woolly adelgid (Adelges tsugae), beech bark disease (Neonectria faginata), and dogwood anthracnose (Discula destructiva) infect and defoliate eastern hemlock, American beech, and flowering dogwood, respectively, all trees used for nesting. Hemlock woolly adelgid has been shown to deter Acadian Flycatcher nesting (Tingley et al. 2002, Allen et al. 2009), but beech bark disease (Morin et al. 2007) and dogwood anthracnose (Holzmueller et al. 2006) can similarly alter forest structure, potentially reducing the number of suitable nesting sites.

Additionally, the Acadian Flycatcher is susceptible to nest predation by the Blue Jay (Cyanocitta cristata) and nest parasitism by the Brown-headed Cowbird (Molothrus ater), potentially resulting in nest failure (Wilson and Cooper 1998, Hazler et al. 2006, Hoover et al. 2006, Olendorf and Robinson 2008). Hoover et al. (2006) found that 99% of Acadian Flycatcher nest failures were due to nest predation and nearly a quarter of nests were parasitized by Brown-headed Cowbirds. Acadian Flycatchers are most vulnerable to nest predation and parasitism during the egg and nestling stages, which range from early May to late August (J. Allair, personal communication). Nests that are already affected by land use change may be particularly vulnerable, such as near agricultural and urban cover and forest edges (Hazler et al. 2006, Hoover et al. 2006, Rodewald 2009).

The resources and effort required to monitor and track species well is immense, and agencies must often use limited data, expert opinions, and what is known from analogous species and ecological theory to make conservation plans and decisions for threatened species (Doak and Mills 1994, Converse et al. 2011). Community science platforms offer an opportunity to monitor and learn about species across space and time in new ways (Conrad and Hilchey 2011, Dickinson et al. 2012, Adler et al. 2020). These platforms are databases in which members of the public submit organismal observations, creating an online community of individuals contributing data that can be used in research (Pocock et al. 2019, Adler et al. 2020). Community science data is increasingly used to answer important ecological questions, including global change questions assessing species abundance and distributions, introduced biota, the effects of climatic change, land-use change, and the identification of rare species (Dickinson et al. 2012, Pocock et al. 2019, Adler et al. 2020, Chu et al. 2022). Community science data has also, however, been questioned for its reliability as observations are collected by thousands of users with varying levels of experience and the data show biases in space and time based on observer behavior (e.g., Courter et al. 2013, Millar et al. 2019, Arazy and Malkinson 2021, Callaghan et al. 2021). Overcoming the biases in community science data requires accounting for observer behavior and the different processes and guidelines for vetting submitted observations, all of which may impact data quality and reliability (Dickinson et al. 2010, Di Cecco et al. 2021, Strimas-Mackey et al. 2023).

Here, we use a large community science database for bird observations, eBird, to assess observations of Acadian Flycatchers in combination with publicly available land use data and community science observations of associated species. The associated species considered include species that enhance habitat quality, preferred nesting tree species, as well as species that could degrade habitat quality, including invasive understory vegetation and common nest predators and parasites. We address the following questions: (1) Are Acadian Flycatchers observed to co-occur with preferred nesting trees (indicative of preferred habitat) or with species that might degrade habitat quality (invasive vegetation, tree disease, and nest predators and parasites)? (2) Are Acadian Flycatchers being observed in areas where land cover is consistent with their described nesting preferences? We hypothesized that the Acadian Flycatcher would be less likely to be observed in the presence of species that may degrade habitat quality and more likely to be observed in preferred habitat in the presence of known preferred nesting trees both in parts of its range where it is considered threatened and where it is considered common. To account for observer bias, we compared our results from the Acadian Flycatcher analysis against a random set of eBird observations where the Acadian Flycatcher was not detected.

METHODS

Data acquisition

We sourced Acadian Flycatcher observations from within their nesting range during the breeding season (1 May to 31 August) from eBird from 2016 to 2023 (eBird 2024). We filtered eBird checklists to only include complete stationary checklists where all observed species were reported to determine when the Acadian Flycatcher was detected versus not detected (Strimas-Mackey et al. 2023). To reduce the effect of variation in observer effort, we only included checklists that were a minimum of 15 minutes and a maximum of 3 hours long (Strimas-Mackey et al. 2023). We also limited our data to only including checklists with less than 10 observers to discount the disproportionately high effort of large groups (Strimas-Mackey et al. 2023). We compared how the co-occurrence with associated species and surrounding land cover characteristics differed between complete checklists where the Acadian Flycatcher was detected versus complete checklists where the Acadian Flycatcher was not detected. Nest predator and parasite species, Blue Jay and Brown-headed Cowbird, observations were sourced from eBird. We sourced the preferred tree species, American beech, eastern hemlock, and flowering dogwood, observations from a different community science platform that includes observations from a variety of taxa, not just birds, iNaturalist (iNaturalist community 2024). Invasive vegetation species, garlic mustard and multiflora rose, were similarly sourced from iNaturalist as well as the tree pathogens and diseases dogwood anthracnose, beech bark disease, and hemlock woolly adelgid (iNaturalist community 2024). However, beech bark disease and hemlock woolly adelgid observations were supplemented by the Early Detection Distribution Mapping System (EDDMapS), a community science database specific to invasive species and pests, as this provided a more comprehensive record for a type of organism that is harder to notice and identify (EDDMapS 2024, iNaturalist community 2024). Before combining these data sets, we assessed whether there were redundant observations between the two platforms by assessing the overlap between points in Quantum Geographic Imaging System (QGIS 2024). Because we determined there were no redundancies, we kept all observations in the dataset. We sourced land cover information for both Ontario and the United States from the North American Land Change Monitoring System (NALCMS) at a 30-meter resolution last updated in 2020 (NALCMS 2020).

Creating observation buffers

We created buffers around checklists where the Acadian Flycatcher was detected and not detected using complete checklist data from eBird. We treated each stationary checklist coordinate as the centroid of a 150-meter radius circular buffer area, creating sampling regions in which we assessed co-occurrence with associated species and land cover composition. For all analyses, we used the coordinate reference system USA Contiguous Albers Equal Conical Area (EPSG = 5070). We created final buffer geometries around checklists where the Acadian Flycatcher was detected and not detected using QGIS (QGIS 2024). Based on the spatial distribution of Acadian Flycatcher observations, we created a polygon boundary using the Convex Hull tool in QGIS to define a discrete nesting range (Fig. 1). We buffered this polygon by 150-meters to ensure no Acadian Flycatcher observations were at the direct edge of the nesting range, as to ensure all associated species and land cover information is included in observation buffers near the edges of the nesting range. For controls, we selected checklists where the Acadian Flycatcher was not detected only if the observations fell within the nesting range (Fig. 1). This allowed us to make comparisons to checklist data where the Acadian Flycatcher was not detected only within the Acadian Flycatcher range such that both datasets assessed the same region. Associated species observations sourced from iNaturalist and EDDMapS with a positional accuracy greater than 75-meters were excluded from the analysis to only include observations that would fall within the Acadian Flycatcher buffer.

To avoid double-counting individual birds and over-estimating landscape characteristics surrounding observer hotspots, we merged bird observations that are close together (within 30-meters of each other) that occurred in the same month and year. We calculated the centroid of the merged buffers then created a new 150-meter radius buffer from this point. We did this for checklists where the Acadian Flycatcher was detected and not detected separately. In this way, we avoided over-estimating the number of individuals at observer hotspots, maintained consistent buffer geometry, and minimized double-counting individuals while avoiding omitting observations. This resulted in 9432 total Acadian Flycatcher observation buffers and 389,532 buffers in which the Acadian Flycatcher was not detected (Fig. 1). We took a random subset of the eBird checklists where the Acadian Flycatcher was not observed such that the sample size matched the sample size of Acadian Flycatcher observations (n = 9432, Fig. 1). We compared the Acadian Flycatcher’s probability of co-occurring with an associated species and surrounding habitat composition to the random subset of checklists in which Acadian Flycatchers were not observed. Making this comparison controls for results related to observer behaviors as both the experimental and control datasets include comparable observer biases and informs whether our results are occurring simply because of the proportion of different available land use types where birds are observed and random chance in the nesting range.

Associated species co-occurrences

All statistical analyses were conducted in R Statistical software ver. 4.3.1 (R Core Team 2024). To assess the probability of the Acadian Flycatcher co-occurring with associated species, we determined co-occurrence around checklists where the Acadian Flycatcher was detected and undetected using the st_intersection function from the sf package in R (Pebesma and Bivand 2023, R Core Team 2024). We assumed that trees around checklists were present from 2016 to 2023 and not lost to disease or logging during that time. The invasive vegetation and tree diseases and pathogens were filtered by corresponding years with the Acadian Flycatcher observations because of their high rate of spread and active eradication efforts. Blue Jays and Brown-headed Cowbirds, a nest predator and a nest parasite, have the highest movement potential, thus, have the greatest temporal constraints. Because the threats to the Acadian Flycatcher posed by both species are only relevant during the nesting season (J. Allair personal communication), we counted co-occurrences from any complete stationary checklist between May and August with the corresponding year to observation buffers.

Using filtered sets of observations, we assessed (1) the probability an associated species will be reported around checklists where the Acadian Flycatcher is detected or not detected, and (2) whether these co-occurrence probabilities of associated species differ between checklist types. We assessed whether each associated species observation was found within a 150-m radius around checklists where Acadian Flycatchers were detected and not detected, using the filtering criteria outlined above. We created a presence-absence table for co-occurrence where each associated species observation point falling within a buffer has a value of 1 and an observation that does not intersect with a buffer has a value of 0. This looks at co-occurrence with each individual observation, creating a probability of co-occurrence by species relative to the species’ sample size, e.g.:

(1)

For example, there are 29,696 American beech observations and 632 co-occurrences with Acadian Flycatcher buffers (Table 1), so the probability is calculated as 632/29,696. We calculate co-occurrence in this way as fewer observations for an associated species have fewer opportunities to co-occur with an Acadian Flycatcher; calculating co-occurrence relative to the number of buffers, would be highly influenced by sample size:

(2)

In this way, we account for this bias by calculating co-occurrence probability specific to each species relative to sample size, accounting for the effects of species-specific sample size. We then assigned associated species intersections to a category: nest predators and parasites, invasive vegetation, preferred trees, and tree pathogens and disease (Table 1). This allowed us to make comparisons on both the species-specific level and the categorical level. We repeated these methods looking at co-occurrence around checklists where Acadian Flycatchers were not observed. It is important to note that an associated species absence may not be a true absence. Associated species, especially those sourced outside of eBird that do not require complete checklists where all species observed are recorded, may be undetected from a site due to observers not searching for or reporting the species. Our methods for calculating co-occurrence probability relative to associated species’ sample size similarly mitigate the effects of species-specific bias in observer effort.

We assessed whether the probability of co-occurrence differed by the type of associated species and by checklist type (Acadian Flycatchers detected or not detected) using a generalized linear mixed effects model with binomial distribution in the lme4 package in R (Bates et al. 2015). This model uses the category of the associated species observations (Table 1), checklist type, and their interaction as fixed effects, year of the observation as the random effect, and the presence-absence within a buffer as a binomial response variable. We followed this with a model following the same criteria using species rather than category (Table 1) as the fixed effect variable.

We followed this with a concordance analysis to assess whether the overall co-occurrence of associated species is consistent between checklist types. To do so, we conducted a Principal Components Analysis (PCA) of co-occurrence of all categories of associated species for both checklists where the Acadian Flycatcher was detected and those where it was not. We extracted the species scores rather than the site scores from the PCA in the analysis because observation buffer sites were not the same between checklist types. We used a Procrustes rotation to compare these species scores of co-occurrences with associated species from the PCA results between checklist types. To assess whether concordance between checklist types is statistically significant, we conducted a 999-permutation PROcrustean randomization TEST (PROTEST) using the protest function in the vegan package in R (Oksanen et al. 2022, R Core Team 2024). Full results of this analysis can be found in Appendix 1, Section B.

Analyzing habitat preference

We extracted the land cover compositions within buffer regions around checklists where the Acadian Flycatcher detected and not detected from the North American Land Change Monitoring System (NALCMS) raster (NALCMS 2020) using the Zonal Statistics tool in QGIS (QGIS 2024). Using this tool, raster pixels with greater than 50% overlap were counted and pixels with less than 50% overlap were omitted from the buffer composition to prevent overestimation of land cover. This method calculates the average cover around each buffer relative to the specific number of included pixels, resulting in a more precise buffer composition than other tools.

We assessed all land cover types classified in NALCMS that were present within the nesting range and omitted land cover types that were found to be less than 1% of land area around both checklist types from our main analysis (Table SA1; the full land cover analysis can be found in Appendix 1, Section C). We compared the eight remaining land cover types: temperate deciduous forest, temperate needleleaf forest, mixed forest, temperate grassland, wetland, water, cropland, and urban (Table 2). We asked (1) are Acadian Flycatchers more likely to be observed within preferred deciduous, wetland, and mixed forest habitats, and (2) does the land cover composition surrounding Acadian Flycatchers differ from areas where the Acadian Flycatcher is not detected? To answer our first question, we used a Kruskal-Wallis Test to determine if the average land cover within the buffers of a checklist type differ, indicating whether any land cover types are significantly more or less prevalent than the others. We followed this with a Dunn Test with Bonferroni correction to identify which land cover types differed in prevalence. We did this independently for both checklist types.

To answer our second question, we determined if the mean composition of each land cover type differed between checklist types using a Mann-Whitney U-Test for each land use type. We built upon this pairwise analysis by conducting a concordance analysis to determine if overall habitat composition significantly differs between checklists where the Acadian Flycatcher was detected and not detected. To do so we ran a PCA of land cover composition for both checklist types followed by a Procrustes rotation. We again used species scores rather than site scores for the PCA because the same habitat variables were assessed at different checklist sites. To assess statistical significance, we again used a 999-permutation PROTEST using the protest function in the vegan package in R (Oksanen et al. 2022, R Core Team 2024). Full results of this analysis can be found in Appendix 1, Section B.

RESULTS

Associated species co-occurrences

We found a significant interaction between category of associated species and checklist type on the probability of co-occurrence (χ² = 605.98, df = 3, p < 0.001, Fig. 2). Acadian Flycatcher checklists were 25.5% more likely to contain preferred nesting trees checklists without Acadian Flycatchers (z = 19.92, p < 0.001, Fig. 2). Relatedly, Acadian Flycatcher checklists were 7.1% more likely to contain tree diseases associated with preferred trees than checklists where the Acadian Flycatcher was not detected (z = 13.76, p < 0.001, Fig. 2). Acadian Flycatcher checklists were 77% less likely to contain invasive vegetation that may disrupt understory conditions (z = -16.32, p < 0.001, Fig. 2). We found no significant difference in the probability of being found near nest predators or parasites between checklist types.

Acadian Flycatcher checklists are 76% more likely to be found near eastern hemlock (z = 13.94, p < 0.001), 75.4% more likely near American beech (z = 13.98, p < 0.001), and 71.7% more likely near flowering dogwood (z = 9.69, p < 0.001) than checklists where the Acadian Flycatcher was not detected (Fig. 3). The prevalence of hemlock woolly adelgid and beech bark disease did not significantly differ between checklist types, but Acadian Flycatchers were 99.9% more likely to co-occur with dogwood anthracnose than checklists where Acadian Flycatchers were not detected (z = 108.68, p < 0.001, Fig. 3). However, this probability is inflated by the extremely small sample size of dogwood anthracnose observations (Table 1). Garlic mustard was 82.2% more likely to co-occur with checklists where the Acadian Flycatcher was not detected than detected checklists (z = -14.87, p < 0.001, Fig. 3). Co-occurrence probability with multiflora rose, Brown-headed Cowbirds, and Blue Jays did not significantly differ between checklist types.

Categorical and species-specific differences in co-occurrence probability within each checklist type are outlined in Appendix 1, Section D. The concordance analysis results indicated that co-occurrences with associated species were not concordant between checklists where the Acadian Flycatcher was detected and not detected (Appendix 1, Section B).

Land cover composition

Acadian Flycatchers were most commonly observed in areas with temperate deciduous forest, urban, and wetland cover (31.46%, 25.64%, and 10.70%, respectively, Fig. 4). Checklists where the Acadian Flycatcher was not detected were dominated by urban, cropland, and temperate deciduous forest cover (48.69%, 13.83%, and 11.46%, respectively, Fig. 4). We found that the prevalence of all land cover types significantly differed between checklist types (p < 0.05, Fig. 4). The concordance analysis, however, indicated concordance of land cover within checklists where the Acadian Flycatcher was detected and not detected (Appendix 1, Section B). We found that the Kruskal-Wallis test indicated that at least one of the land cover types around checklists where the Acadian Flycatcher was detected significantly differed in prevalence from the others (χ² = 20774, df = 7, p < 0.001). We followed this with a post-hoc Dunn Test, which indicated that all pairwise comparisons between land cover types significantly differed from each other (p < 0.001, Table SA2).

DISCUSSION

By making comparisons between birding checklists where the Acadian Flycatcher was and was not detected, we can interpret the relative preference and avoidance behaviors of the Acadian Flycatcher from co-occurrences with associated species and land cover types. We observed a significantly higher incidence of preferred nesting trees around checklists where Acadian Flycatchers were reported relative to those where they were not detected (Fig. 2), suggesting that the species is largely spending time in habitat that meets its nesting preferences during the breeding season. Eastern hemlock trees, however, were present significantly less than flowering dogwood around Acadian Flycatcher checklists (Fig. 3), despite previous work showing hemlock to be the most preferred species for nesting (Ross et al. 2004, Allen et al. 2009). This may reflect a lack of available eastern hemlock trees, as hemlock woolly adelgid is increasingly prevalent and the North American hemlock population is decreasing (Ellison et al. 2018). However, this could also indicate that iNaturalist users are less likely to report hemlock trees that are common and widespread compared to flowering dogwood. Co-occurrences with tree pathogens, despite the very small number of available records within the Acadian Flycatcher’s nesting range (Table 1), highlights their prevalence and potential importance in degrading Acadian Flycatcher habitat. If tree disease continues to spread, it may become more difficult for the nesting requirements of the Acadian Flycatcher to be met even within otherwise high-quality habitat patches. We did not have a large enough sample size to assess co-occurrence with beech bark disease in either checklist type (Table 1).

Despite many thousands of observations within the Acadian Flycatcher’s nesting range (Table 1), few Acadian Flycatchers were observed in the presence of garlic mustard or multiflora rose. Significantly fewer invasive vegetation observations, specifically garlic mustard observations, were found around checklists where Acadian Flycatchers were detected than not detected, suggesting avoidant behavior (Figs 2, 3). This is consistent with the idea that the presence of these species negatively alters the forest understory structure preferred by Acadian Flycatchers (D. Martin, unpublished data reported by Environment Canada 2012). However, invasive species are more prevalent around disturbed sites than mature forests (Lozon and MacIsaac 1997). Because checklists where the Acadian Flycatcher is undetected are more likely to be found near disturbed urban and cropland habitat (Fig. 4), this result may reflect a lack of suitable forest cover rather than a direct effect of the invasive species.

We found no significant difference in the co-occurrence with nest predators and parasites between checklist types (Fig. 2). However, this threat may increase with land use change. Brown-headed Cowbirds have been shown to be more likely to parasitize nests in fragmented habitats close to edges, particularly in agriculturally transformed landscapes (Cavitt and Martin 2002, Howell et al. 2007). Notably, Acadian Flycatcher vulnerability to predation and parasitism from the Blue Jay and the Brown-headed Cowbird has been observed to be positively related to habitat fragmentation and agriculturally transformed landscapes (Hoover et al. 2006). Thus, as habitat continues to become increasingly fragmented and degraded, the consequences of habitat loss may be compounded by increased nest predation and parasitism, though that does not appear to be the case for the observations assessed here.

Deciduous and wetland cover composed high proportions of buffer regions (Fig. 4) as expected by the Acadian Flycatcher habitat specialization requirements (Martin 2007, COSEWIC 2010). These land cover types are significantly more common around Acadian Flycatcher checklists than checklists where the species was not detected. Additionally, checklists where the species was not detected were found to have more low-quality habitat for Acadian Flycatchers, such as agricultural cover (Fig. 4). This indicates that, relative to checklists where Acadian Flycatchers were not detected, the Acadian Flycatcher is found in habitat meeting its habitat requirements, potentially indicating selective behavior and revealing the importance of habitat preservation for Acadian Flycatcher nesting.

We observed surprisingly high proportions of urban cover around checklists where the Acadian Flycatcher was detected; this was significantly less common than checklists where the species was not detected (Fig. 4). Overall, we found high amounts of urban cover in both checklist types, though this is unlikely to be due to habitat preference because Acadian Flycatchers are known to avoid habitat edges (Hoover et al. 2006). Rather, the higher-than-expected proportions of urban cover is likely an artifact of bias associated with community science data. Community science participants submit observations from areas they can access, such as paths and roads, where birds are more visible and audible to people (Mair and Ruete 2016, Tiago et al. 2017).

Acadian Flycatchers are thought to be habitat specialists, preferring deciduous forests or wooded wetlands with a closed canopy and open understory (Christy 1942, Martin 2007) with additional requirements such as low vegetation density (Bakermans and Rodewald 2006) and preferences for certain trees for nesting (Martin 2007, Becker et al. 2008, Allen et al. 2009, COSEWIC 2010). Overall, the results of this study emphasize that the Acadian Flycatcher has habitat preferences across its range, not only along range edges where it is currently threatened. It also highlights the importance of the preservation of suitable natural habitats to support Acadian Flycatcher populations, primarily deciduous forests and wooded wetlands with a closed canopy and open understory. As land use change and development continues in North America (Alig et al. 2004), remaining suitable habitat for the Acadian Flycatcher will likely continue to shrink. This is especially true in southern Ontario, where the species is considered endangered, and where the landscape is already dominated by agriculture with increasing urbanization (Smith 2015, Eimers et al. 2020), resulting in a lack of availability of the high-quality natural cover the Acadian Flycatcher requires.

The forest composition preferred by Acadian Flycatchers was observed to be disturbed by tree diseases, even with a very small number of available tree disease observations. We observed potential avoidance with invasive vegetation; this behavior may deter visitation and potentially nesting at otherwise suitable locations that meet the species’ canopy and understory requirements. Awareness of these potential trends and habitat information to protect and maintain appropriate habitat may aid the conservation of the Acadian Flycatcher in Ontario and mitigate future declines in the United States, maintaining the Acadian Flycatcher’s status as an abundant songbird.

There have been efforts to conserve the Acadian Flycatcher in Ontario, where populations are currently at risk. One aspect of this work includes promoting environmental stewardship and providing outreach information to landowners, land managers, and forest users with the aim of slowing the spread of invasive species and deterring tree harvesting in native forest patches (Environment Canada 2012). This strategy also entails the continuous identification of degraded habitat areas in need of intervention and conservation (Environment Canada 2012). By increasing our understanding of the specific requirements and potential threats facing this species across its range, conservation efforts can be expanded and implemented across the Acadian Flycatcher range to prevent, not just reverse, population decline.

LITERATURE CITED

Adler, F. R., A. M. Green, and Ç. H. Şekercioğlu. 2020. Citizen science in ecology: a place for humans in nature. Annals of the New York Academy of Sciences 1469:52-64. https://doi.org/10.1111/nyas.14340

Alig, R. J., J. D. Kline, and M. Lichtenstein. 2004. Urbanization on the US landscape: looking ahead in the 21st century. Landscape and Urban Planning 69:219-234. https://doi.org/10.1016/j.landurbplan.2003.07.004

Allair, J., S. Dobney, and M. Falconer. 2015. Forest birds at risk in the Carolinian forest of Southwestern Ontario. 2015. Bird Studies Canada, Port Rowan, Ontario, Canada. https://birdscanada.org/wp-content/uploads/2021/03/2015-Forest-Birds-at-Risk-in-the-Carolinian-Forest-of-Southwestern-Ontario-Summary-Report.pdf

Allen, M. C., M. M. Napoli, J. Sheehan, T. L. Master, P. Pyle, D. R. Whitehead, and T. Taylor. 2017. Acadian Flycatcher (Empidonax virescens), version 2.0. In P. G. Rodewald, editor. The birds of North America. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bna.acafly.02

Allen, M. C., J. Sheehan Jr., T. L. Master, and R. S. Mulvihill. 2009. Responses of Acadian Flycatchers (Empidonax virescens) to Hemlock Woolly Adelgid (Adelges tsucae) infestation in Appalachian riparian forests. Auk 126:543-553. https://doi.org/10.1525/auk.2009.08073

Arazy, O., and D. Malkinson. 2021. A framework of observer-based biases in citizen science biodiversity monitoring: semi-structuring unstructured biodiversity monitoring protocols. Frontiers in Ecology and Evolution 9:693602. https://doi.org/10.3389/fevo.2021.693602

Bakermans, M. H., and A. D. Rodewald. 2006. Scale-dependent habitat use of Acadian Flycatcher (Empidonax Virescens) in central Ohio. Auk 123:368-382. https://doi.org/10.1093/auk/123.2.368

Bates, D., M. Mächler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67. https://doi.org/10.18637/jss.v067.i01

Becker, D. A., M. C. Brittingham, and C. B. Goguen. 2008. Effects of Hemlock Woolly Adelgid on breeding birds at Fort Indiantown Gap, Pennsylvania. Northeastern Naturalist 15:227-240. https://doi.org/10.1656/1092-6194(2008)15[227:EOHWAO]2.0.CO;2

Callaghan, C. T., A. G. B. Poore, M. Hofmann, C. J. Roberts, and H. M. Pereira. 2021. Large-bodied birds are over-represented in unstructured citizen science data. Scientific Reports 11:19073. https://doi.org/10.1038/s41598-021-98584-7

Cavitt, J. F., and T. E. Martin. 2002. Effects of forest fragmentation on brood parasitism and nest predation in eastern and western landscapes. Studies in Avian Biology 25:73-80.

Cheskey, E., and The Twelve Mile Creek Headwaters IBA Steering Committee. 2003. Habitat stewardship for forest birds at risk in the Carolinian zone. Twelve Mile Creek Headwaters Important Bird Area: A Conservation Planning Report. Fonthill, Ontario, Canada. http://www.ibacanada.ca/documents/conservationplans/ontwelvemilecreekheadwater.pdf

Christy, B. H. 1942. Acadian Flycatcher Empidonax virescens. Pages 183-203 in A. C. Bent, editor. Life histories of North American flycatchers, larks, swallows, and their allies. U.S. National Museum Bulletin 179.

Chu, J. J., D. P. Gillis, and S. H. Riskin. 2022. Community science reveals links between migration arrival timing advance, migration distance and wing shape. Journal of Animal Ecology 91:1651-1665. https://doi.org/10.1111/1365-2656.13755

Committee on the Status of Endangered Wildlife in Canada (COSEWIC). 2010. COSEWIC assessment and status report on the Acadian Flycatcher (Empidonax virescens) in Canada. COSEWIC, Ottawa, Canada.

Conrad, C. C., and K. G. Hilchey. 2011. A review of citizen science and community-based environmental monitoring: issues and opportunities. Environmental Monitoring and Assessment 176:273-291. https://doi.org/10.1007/s10661-010-1582-5

Converse, S. J., K. J. Shelley, S. Morey, J. Chan, A. LaTier, C. Scafidi, D. T. Crouse, and M. C. Runge. 2011. A decision-analytic approach to the optimal allocation of resources for endangered species consultation. Biological Conservation 144:319-329. https://doi.org/10.1016/j.biocon.2010.09.009

Courter, J. R., R. J. Johnson, C. M. Stuyck, B. A. Lang, and E. W. Kaiser. 2013. Weekend bias in citizen science data reporting: implications for phenology studies. International Journal of Biometeorology 57:715-720. https://doi.org/10.1007/s00484-012-0598-7

Di Cecco, G. J., V. Barve, M. W. Belitz, B. J. Stucky, R. P. Guralnick, and A. H. Hurlbert. 2021. Observing the observers: how participants contribute data to iNaturalist and implications for biodiversity science. BioScience 71:1179-1188. https://doi.org/10.1093/biosci/biab093

Dickinson, J. L., J. Shirk, D. Bonter, R. Bonney, R. L. Crain, J. Martin, T. Phillips, and K. Purcell. 2012. The current state of citizen science as a tool for ecological research and public engagement. Frontiers in Ecology and the Environment 10:291-297. https://doi.org/10.1890/110236

Dickinson, J. L., B. Zuckerberg, and D. N. Bonter. 2010. Citizen science as an ecological research tool: challenges and benefits. Annual Review of Ecology, Evolution, and Systematics 41:149-172. https://doi.org/10.1146/annurev-ecolsys-102209-144636

Doak, D. F., and L. S. Mills. 1994. A useful role for theory in conservation. Ecology 75(3):615-626. https://doi.org/10.2307/1941720

eBird Basic Dataset. 2024. Version: EBD_relMay-2024. Cornell Lab of Ornithology, Ithaca, New York, USA.

EDDMapS. 2024. Early detection & distribution mapping system. The University of Georgia - Center for Invasive Species and Ecosystem Health, Tifton, Georgia, USA. http://www.eddmaps.org/

Eimers, M. C., F. Liu, and J. Bontje. 2020. Land use, land cover, and climate change in southern Ontario: implications for nutrient delivery to the Lower Great Lakes. Pages 235-249 in J. Crossman and C. Weisener, editors. Contaminants of the Great Lakes. Springer International, Cham, Switzerland. https://doi.org/10.1007/698_2020_519

Ellison, A. M., D. A. Orwig, M. C. Fitzpatrick, and E. L. Preisser. 2018. The past, present, and future of the Hemlock Woolly Adelgid (Adelges tsugae) and its ecological interactions with eastern hemlock (Tsuga canadensis) forests. Insects 9:172. https://doi.org/10.3390/insects9040172

Environment Canada. 2012. Acadian Flycatcher (Empidonax virescens) and Hooded Warbler (Wilsonia citrina): recovery strategy 2012. Environment Canada, Ottawa Ontario, Canada

Fauth, P. T., and P. R. Cabe. 2005. Reproductive success of Acadian Flycatchers in the Blue Ridge Mountains of Virginia. Journal of Field Ornithology 76:150-157. https://doi.org/10.1648/0273-8570-76.2.150

Filek, N., A. Cimadom, C. H. Schulze, H. Jägger, and A. Tebbich. 2018. The impact of invasive plant management on the foraging ecology of the Warbler Finch (Certhidea olivacea) and the Small Tree Finch (Camarhynchus parvulus) on Galápagos. Journal of Ornithology 159:129-140. https://doi.org/10.1007/s10336-017-1481-4

Hazler, K. R., A. J. Amacher, R. A. Lancia, and J. A. Gerwin. 2006. Factors influencing Acadian Flycatcher nesting success in an intensively managed forest landscape. Journal of Wildlife Management 70:532-538. https://doi.org/10.2193/0022-541X(2006)70[532:FIAFNS]2.0.CO;2

Holzmueller, E., S. Jose, M. Jenkins, A. Camp, and A. Long. 2006. Dogwood anthracnose in eastern hardwood forests: what is known and what can be done? Journal of Forestry 104:21-26. https://doi.org/10.1093/jof/104.1.21

Hoover, J. P., T. H. Tear, and M. E. Baltz. 2006. Edge effects reduce the nesting success of Acadian Flycatchers in a moderately fragmented forest. Journal of Field Ornithology 77:425-436. https://doi.org/10.1111/j.1557-9263.2006.00074.x

Howell, C. A., W. D. Dijak, and F. R. Thompson III. 2007. Landscape context and selection for forest edge by breeding Brown-headed Cowbirds. Landscape Ecology 22:273-284. https://doi.org/10.1007/s10980-006-9022-1

iNaturalist community. 2024. Observations of American beech, eastern hemlock, flowering dogwood, garlic mustard, multiflora rose, dogwood anthracnose, hemlock woolly adelgid, and beech bark disease from Ontario, Canada and the United States observed on/between 2016-2023. https://www.inaturalist.org

Litt, A. R., and R. J. Steidl. 2010. Insect assemblages change along a gradient of invasion by a nonnative grass. Biological Invasions 12:3449-3463. https://doi.org/10.1007/s10530-010-9743-6

Lozon, J. D., and H. J. MacIsaac. 1997. Biological invasions: are they dependent on disturbance? Environmental Reviews 5:131-144. https://doi.org/10.1139/a97-007

Mair, L., and A. Ruete. 2016. Explaining spatial variation in the recording effort of citizen science data across multiple taxa. PLoS ONE 11:e0147796. https://doi.org/10.1371/journal.pone.0147796

Martin, D. 2007. Acadian Flycatcher. Pages 344-345 in M. D. Cadman, D. A. Sutherland, G. G. Beck, D. Lepage, and A. R. Couturier, editors. Atlas of the breeding birds of Ontario, 2001-2005. Bird Studies Canada, Environment Canada, Ontario Field Ornithologists, Ontario Ministry of Natural Resources, and Ontario Nature, Toronto, Ontario, Canada.

Massé, R. J., and K. Vulinec. 2010. Possible impact of multiflora rose on breeding-bird diversity in riparian forest fragments of central Delaware. Northeastern Naturalist 17:647-658. https://doi.org/10.1656/045.017.0410

Millar, E. E., E. C. Hazell, and S. J. Melles. 2019. The ‘cottage effect’ in citizen science? Spatial bias in aquatic monitoring programs. International Journal of Geographical Information Science 33(8):1612-1632. https://doi.org/10.1080/13658816.2018.1423686

Minnesota Department of Natural Resources. 2024 (MDNR). Rare species guide: Acadian Flycatcher (Empidonax virescens). Minnesota Department of Natural Resources, Saint Paul, Minnesota, USA.

Morin, R. S., A. M. Liebhold, P. C. Tobin, K. W. Gottschalk, and E. Luzader. 2007. Spread of beech bark disease in the eastern United States and its relationship to regional forest composition. Canadian Journal of Forest Research 37:726-736. https://doi.org/10.1139/X06-281

North American Land Cover (NALCMS). 2020. Landsat, 30-m. Commission for Environmental Cooperation. http://www.cec.org/north-american-environmental-atlas/land-cover-30m-2020/

Oksanen, J., G. Simpson, F. Blanchet, R. Kindt, P. Legendre, P. Minchin, R. O’Hara, P. Solymos, M. Stevens, E. Szoecs, H. Wagner, M. Barbour, M. Bedward, B. Bolker, D. Borcard, G. Carvalho, M. Chirico, M. De Caceres, S. Durand, H. Evangelista, R. FitzJohn, M. Friendly, B. Furneaux, G. Hannigan, M. Hill, L. Lahti, D. McGlinn, M. Ouellette, E. Ribeiro Cunha, T. Smith, A. Stier, C. Ter Braak, and J. Weedon. 2022. vegan: Community Ecology Package. R package version 2.6-4. https://doi.org/10.32614/CRAN.package.vegan

Olendorf, R., and S. K. Robinson. 2008. Effectiveness of nest defense in the Acadian Flycatcher Empidonax virescens. Ibis 142(3):365-371. https://doi.org/10.1111/j.1474-919X.2000.tb04432.x

Pebesma, E., and R. Bivand. 2023. Spatial data science: with applications in R. First edition. Chapman and Hall/CRC, New York, New York, USA. https://doi.org/10.1201/9780429459016

Pimm, S., P. Raven, A. Peterson, Ç. H. Şekercioğlu, and P. R. Ehrlich. 2006. Human impacts on the rates of recent, present, and future bird extinctions. Proceedings of the National Academy of Sciences 103:10941-10946 https://doi.org/10.1073/pnas.0604181103

Pocock, M. J. O., H. E. Roy, T. August, A. Kuria, F. Barasa, J. Bett, M. Githiru, J. Kairo, J. Kimani, W. Kinuthia, B. Kissui, I. Madindou, K. Mbogo, J. Mirembe, P. Mugo, F. M. Muniale, P. Njoroge, E. G. Njuguna, M. I. Olendo, M. Opige, T. O. Otieno, C. C. Ng’weno, E. Pallangyo, T. Thenya, A. Wanjiru, and R. Trevelyan. 2019. Developing the global potential of citizen science: assessing opportunities that benefit people, society and the environment in East Africa. Journal of Applied Ecology 56:274-281. https://doi.org/10.1111/1365-2664.13279

QGIS.org. 2024. QGIS Geographic Information System. QGIS Association.

R Core Team. 2024. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Rodewald, A. D. 2009. Urban-associated habitat alteration promotes brood parasitism of Acadian Flycatchers. Journal of Field Ornithology 80:234-241. https://doi.org/10.1111/j.1557-9263.2009.00226.x

Ross, R. M., L. A. Redell, R. M. Bennett, and J. A. Young. 2004. Mesohabitat use of threatened hemlock forests by breeding birds of the Delaware River Basin in Northeastern United States. Natural Areas Journal 24(4):307-315.

Sauer, J. R., D. K. Niven, J. E. Hines, D. J. Ziolkowski Jr., K. L. Pardieck, J. E. Fallon, and W. A. Link. 2017. The North American breeding bird survey, results and analysis 1966–2015. Version 2.07.2017. USGS Patuxent Wildlife Research Center, Laurel, Maryland, USA.

Schneider, S. C., and J. R. Miller. 2014. Response of avian communities to invasive vegetation in urban forest fragments. Condor 116:459-471. https://doi.org/10.1650/CONDOR-13-009R1.1

Shustack, D. P., and A. D. Rodewald. 2010. Attenuated nesting season of the Acadian Flycatcher (Empidonax virescens) in urban forests. Auk 127(2):421-429. https://doi.org/10.1525/auk.2009.09129

Smith, P. G. R. 2015. Long-term temporal trends in agri-environment and agricultural land use in Ontario, Canada: transformation, transition and significance. Journal of Geography and Geology 7:32-55. https://doi.org/10.5539/jgg.v7n2p32

Stoleson, S. H., and D. M. Finch. 2001. Breeding bird use of and nesting success in exotic Russian olive in New Mexico. Wilson Bulletin 113(4):452-455. http://www.jstor.org/stable/4164400

Strimas-Mackey, M., W.M. Hochachka, V. Ruiz-Gutierrez, O.J. Robinson, E.T. Miller, T. Auer, S. Kelling, D. Fink, and A. Johnston. 2023. Best Practices for using eBird data. Version 2.0. Cornell Lab of Ornithology, Ithaca, New York, USA. https://ebird.github.io/ebird-best-practices/

Szabo, J. K., N. Khwaja, S. T. Garnett, and S. H. M. Butchart. 2012. Global patterns and drivers of avian extinctions at the species and subspecies level. PLoS ONE 7:e47080. https://doi.org/10.1371/journal.pone.0047080

Tiago, P., A. Ceia-Hasse, T. A. Marques, C. Capinha, and H. M. Pereira. 2017. Spatial distribution of citizen science casuistic observations for different taxonomic groups. Scientific Reports 7:12832. https://doi.org/10.1038/s41598-017-13130-8

Tingley, M. W., D. A. Orwig, R. Field, and G. Motzkin. 2002. Avian response to removal of a forest dominant: consequences of hemlock woolly adelgid infestations. Journal of Biogeography 29:1505-1516. https://doi.org/10.1046/j.1365-2699.2002.00789.x

Wilson, R. R., and R. J. Cooper. 1998. Acadian Flycatcher nest placement: does placement influence reproductive success? Condor 100(4):673679. https://doi.org/10.2307/1369748

Wisconsin Department of Natural Resources (WDNR). 2018. Wisconsin Acadian Flycatcher species guidance. PUB-ER-685. Bureau of Natural Heritage Conservation, Wisconsin Department of Natural Resources, Madison, Wisconsin, USA. https://dnr.wisconsin.gov/topic/EndangeredResources/SpeciesGuidance