Limits...
Change Points in the Population Trends of Aerial-Insectivorous Birds in North America: Synchronized in Time across Species and Regions.

Smith AC, Hudson MA, Downes CM, Francis CM - PLoS ONE (2015)

Bottom Line: We found evidence for group-level change points in 85% of the strata.This group-level synchrony in AI population trends is likely evidence of a response to a common environmental factor(s) with similar effects on many species across broad spatial extents.The timing and geographic patterns of the change points that we identify here should provide a spring-board for research into the causes behind aerial insectivore declines.

View Article: PubMed Central - PubMed

Affiliation: Canadian Wildlife Service, Environment Canada, Ottawa, Ontario, Canada.

ABSTRACT
North American populations of aerial insectivorous birds are in steep decline. Aerial insectivores (AI) are a group of bird species that feed almost exclusively on insects in flight, and include swallows, swifts, nightjars, and flycatchers. The causes of the declines are not well understood. Indeed, it is not clear when the declines began, or whether the declines are shared across all species in the group (e.g., caused by changes in flying insect populations) or specific to each species (e.g., caused by changes in species' breeding habitat). A recent study suggested that population trends of aerial insectivores changed for the worse in the 1980s. If there was such a change point in trends of the group, understanding its timing and geographic pattern could help identify potential causes of the decline. We used a hierarchical Bayesian, penalized regression spline, change point model to estimate group-level change points in the trends of 22 species of AI, across 153 geographic strata of North America. We found evidence for group-level change points in 85% of the strata. Change points for flycatchers (FC) were distinct from those for swallows, swifts and nightjars (SSN) across North America, except in the Northeast, where all AI shared the same group-level change points. During the 1980s, there was a negative change point across most of North America, in the trends of SSN. For FC, the group-level change points were more geographically variable, and in many regions there were two: a positive change point followed by a negative change point. This group-level synchrony in AI population trends is likely evidence of a response to a common environmental factor(s) with similar effects on many species across broad spatial extents. The timing and geographic patterns of the change points that we identify here should provide a spring-board for research into the causes behind aerial insectivore declines.

No MeSH data available.


Related in: MedlinePlus

Geographic Distribution of the Relative Support for 5 models Testing Shared Population Trend Changes among North American Aerial Insectivores.These models include a common group change point for all aerial insectivore species (strata in shades of purple, best model MallAI), separate group change points for flycatchers (FC) and swallows, swifts, and nightjars (SSN; strata in shades of orange, best model one of MFC,SSN, MFC, or MSSN), or no group change point (strata in white, best model MS). Intensity of the purple or orange reflects the relative support for the best model over the next best model represented by another colour (e.g., the best model in the dark orange coloured strata has DIC at least 4 units less than the DIC for either MAllAI or MS).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4493114&req=5

pone.0130768.g002: Geographic Distribution of the Relative Support for 5 models Testing Shared Population Trend Changes among North American Aerial Insectivores.These models include a common group change point for all aerial insectivore species (strata in shades of purple, best model MallAI), separate group change points for flycatchers (FC) and swallows, swifts, and nightjars (SSN; strata in shades of orange, best model one of MFC,SSN, MFC, or MSSN), or no group change point (strata in white, best model MS). Intensity of the purple or orange reflects the relative support for the best model over the next best model represented by another colour (e.g., the best model in the dark orange coloured strata has DIC at least 4 units less than the DIC for either MAllAI or MS).

Mentions: Using the annual indices from the spatial CAR model as data for the Bayesian piecewise linear regression spline model, we found broad support for group-level change points in the population trends of North American aerial insectivores. In 85% of the 153 geographic strata for which we had sufficient data to run a model (annual indices for 2 or more species of both SSN and FC), the best supported model (lowest DIC) included at least one group change point (MAllAi, MFC,SSN, MFC, or MSSN (Table 2); coloured various shades of either purple or orange; Fig 2, S1 Table). The few remaining strata where the model with no group change point (Ms) was best supported (coloured white; Fig 2) were predominantly central and in the west. In these cases, the relative support for the Ms model over the best supported group change point model was relatively weak (mean ΔDIC = 2.4, range [0.03, 5.2]; S1 Table).


Change Points in the Population Trends of Aerial-Insectivorous Birds in North America: Synchronized in Time across Species and Regions.

Smith AC, Hudson MA, Downes CM, Francis CM - PLoS ONE (2015)

Geographic Distribution of the Relative Support for 5 models Testing Shared Population Trend Changes among North American Aerial Insectivores.These models include a common group change point for all aerial insectivore species (strata in shades of purple, best model MallAI), separate group change points for flycatchers (FC) and swallows, swifts, and nightjars (SSN; strata in shades of orange, best model one of MFC,SSN, MFC, or MSSN), or no group change point (strata in white, best model MS). Intensity of the purple or orange reflects the relative support for the best model over the next best model represented by another colour (e.g., the best model in the dark orange coloured strata has DIC at least 4 units less than the DIC for either MAllAI or MS).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4493114&req=5

pone.0130768.g002: Geographic Distribution of the Relative Support for 5 models Testing Shared Population Trend Changes among North American Aerial Insectivores.These models include a common group change point for all aerial insectivore species (strata in shades of purple, best model MallAI), separate group change points for flycatchers (FC) and swallows, swifts, and nightjars (SSN; strata in shades of orange, best model one of MFC,SSN, MFC, or MSSN), or no group change point (strata in white, best model MS). Intensity of the purple or orange reflects the relative support for the best model over the next best model represented by another colour (e.g., the best model in the dark orange coloured strata has DIC at least 4 units less than the DIC for either MAllAI or MS).
Mentions: Using the annual indices from the spatial CAR model as data for the Bayesian piecewise linear regression spline model, we found broad support for group-level change points in the population trends of North American aerial insectivores. In 85% of the 153 geographic strata for which we had sufficient data to run a model (annual indices for 2 or more species of both SSN and FC), the best supported model (lowest DIC) included at least one group change point (MAllAi, MFC,SSN, MFC, or MSSN (Table 2); coloured various shades of either purple or orange; Fig 2, S1 Table). The few remaining strata where the model with no group change point (Ms) was best supported (coloured white; Fig 2) were predominantly central and in the west. In these cases, the relative support for the Ms model over the best supported group change point model was relatively weak (mean ΔDIC = 2.4, range [0.03, 5.2]; S1 Table).

Bottom Line: We found evidence for group-level change points in 85% of the strata.This group-level synchrony in AI population trends is likely evidence of a response to a common environmental factor(s) with similar effects on many species across broad spatial extents.The timing and geographic patterns of the change points that we identify here should provide a spring-board for research into the causes behind aerial insectivore declines.

View Article: PubMed Central - PubMed

Affiliation: Canadian Wildlife Service, Environment Canada, Ottawa, Ontario, Canada.

ABSTRACT
North American populations of aerial insectivorous birds are in steep decline. Aerial insectivores (AI) are a group of bird species that feed almost exclusively on insects in flight, and include swallows, swifts, nightjars, and flycatchers. The causes of the declines are not well understood. Indeed, it is not clear when the declines began, or whether the declines are shared across all species in the group (e.g., caused by changes in flying insect populations) or specific to each species (e.g., caused by changes in species' breeding habitat). A recent study suggested that population trends of aerial insectivores changed for the worse in the 1980s. If there was such a change point in trends of the group, understanding its timing and geographic pattern could help identify potential causes of the decline. We used a hierarchical Bayesian, penalized regression spline, change point model to estimate group-level change points in the trends of 22 species of AI, across 153 geographic strata of North America. We found evidence for group-level change points in 85% of the strata. Change points for flycatchers (FC) were distinct from those for swallows, swifts and nightjars (SSN) across North America, except in the Northeast, where all AI shared the same group-level change points. During the 1980s, there was a negative change point across most of North America, in the trends of SSN. For FC, the group-level change points were more geographically variable, and in many regions there were two: a positive change point followed by a negative change point. This group-level synchrony in AI population trends is likely evidence of a response to a common environmental factor(s) with similar effects on many species across broad spatial extents. The timing and geographic patterns of the change points that we identify here should provide a spring-board for research into the causes behind aerial insectivore declines.

No MeSH data available.


Related in: MedlinePlus