White-winged Juncos in Colorado (LONG)
83 views
Skip to first unread message
colorad...@aol.com
Feb 5, 2016, 3:51:46 PM2/5/16
to cob...@googlegroups.com
All:
Since one of my favorite birds has come up as a topic in this venue, I find it impossible to not jump in. In my experience, White-winged Junco's primary winter range is the Ponderosa Pine belt on the east side of the Front Range and Wet Mountains. It is nowhere numerous and is nearly always greatly outnumbered by other junco taxa. This is not surprising considering that the breeding range of White-winged Junco is so small and its winter range is so much larger. So, while the taxon is fairly readily findable in winter at the foothill-plains interface (particularly at a few favored locations, such as Red Rocks Park), White-winged Junco is typically found there only in tiny numbers. Hiking uphill is the way to find them, such as illustrated by Ted's checklist.
Below, I provide links to the eBird distribution maps of White-winged Junco for select winters:
Current winter
2014-2015
2013-2014
2012-2013
As can be seen on the above maps, overall distribution in CO across recent winters does not differ in any appreciable way. As comparison to a decade ago, note the eBird map for winter 2005-2006. HOWEVER, many, many fewer birders used eBird ten years ago, so old data like these cannot be considered to accurately represent actual occurrence patterns. That leaves us with the Christmas Bird Count (CBC) as the only long-term data set available for looking at such questions. Unfortunately, there is a huge gob of methodological problems with the CBC (both CBC policy and individual-CBC mechanics) that make analyzing the data an exercise best left for a Ph.D in statistics. Though I will not delve into those problems here, a very obvious one is the effect of deep-sixing the participation fee within the past ten years ago and the resultant great increase of participation (both in observers/circle and in number of circles), which has to have had a large impact on data and their analysis. However, as a very rough indication of patterns for taxa found in sufficient numbers, it does a reasonable job; White-winged Junco is NOT such a taxon.
For the below discussion, I queried the online National Audubon Society's CBC database and obtained the previous 20-year history for White-winged Junco in Colorado and for five select CBCs. The span of CBCs queried was from the 2005-2006 season through the 2014-2015 season (the 2015-2016 data set is not complete, nor available for query). I chose five Colorado CBCs as the best representative counts to analyze Colorado White-winged Junco abundance in order to cover as much of the spread of occurrence in the state as possible while also maintaining sufficient sample size for useful analysis. The five count circles, all with extensive acreage of suitable habitat for the taxon are, from north to south: Boulder, Evergreen-Idaho Springs, Black Forest, Pikes Peak, and Penrose. The graphs use number of individuals/party-hour in order to normalize the data for, at times, highly variable number of observers across years. I made screen captures of the graphs and placed them on my Flickr site.
In Colorado as a whole, the last half of the 20-year period seems to show a decline (but see caveat above). However, the most striking feature of the graph is the great variability in abundance, particularly in the first ten years, with a fairly regular cycle of boom and bust. While weather might also explain it, my experience with that variable is that winter weather regimes are not so precisely cyclical as to explain this result. Other passerine species are known to exhibit similar boom-and-bust cycles, with Common Redpoll and Pine Siskin being two of the best examples, and these two species also match the short temporal pattern of the apparent White-winged Junco cycle. The two finch species are thought to be responding to specific food sources; the Pine Siskin to the availability of pine seeds. Considering that White-winged Junco has a high preference for pine forest, perhaps it, too, is responding to localized food resources, perhaps with larger proportions of the taxon staying closer to the breeding grounds in some years, and having to travel farther afield in others. Inter-annual variation in breeding success might also be a causative factor behind the cycle. Documentation of age structure in winter in Colorado and in the Black Hills might provide useful data at getting at the cause of the cycle.
Regardless of the cause(s) of the apparent population cycle, the system seems to have broken down in the last ten years or so, with four consecutive years of decline, from the 2008-2009 season to the 2011-2012, with a return to something of a cycle in the last four years, albeit at a considerably lower abundance value.
Looking at individual CBC results (see below), the five selected CBCs throw a huge monkey wrench into the analysis, one that might be explainable with a more-in-depth analysis, but which lies outside the scope of this post. In the below graphs, we can see that no two of the five CBC circles present results similar to each other, nor to the general Colorado results!
Boulder
Evergreen-Idaho Springs
Black Forest
Pikes Peak
Penrose
Table 1 illustrates a general north-to-south trend of decreasing reliability of detection, suggesting that there is a geographic correlate of abundance as it affects detection probability, thus of frequency, and that southern Colorado is closer to the edge of the winter range than is northern Colorado. That suggestion seems borne out by the overall eBird winter-distribution map. Somewhat contrastingly, both maximum abundance and average abundance reached their highest values in central Colorado. However, the three central counts all have a higher percentage of their circles occupied by suitable White-winged Junco habitat than does either the Boulder CBC, a large portion of which is occupied by urbanized plains habitats, or Penrose, a large portion of which is occupied by low-elevation habitats. Table 1 suggests that the southern part of the Front Range and the associated Palmer Divide is the epicenter of Colorado White-winged Junco abundance, with the Black Forest CBC circle being especially so. However, the small sample size of Colorado CBCs incorporating extensive suitable habitat (n=5) makes that assumption somewhat problematic.
Table 1. Averages of White-winged Junco abundance on five
Colorado Christmas Bird Counts, 1995-1996 to 2014-2015.
=======================================================
# years Abundance Avg.
CBC detected Min Max abundance*
-------------------------------------------------------------------------------------------------------------
Boulder 20 0.0122 0.3447 0.1309
Evergreen-Idaho Springs 20 0.0898 0.6612 0.3848
Black Forest 16 0 4.4615 0.6287
Pikes Peak 11 0 1.1195 0.1918
Penrose 11 0 0.6387 0.0826
-------------------------------------------------------------------------------------------------------------
*Abundance is generated by summing annual values of
birds/party-hour and then dividing by 20.
Table 2 presents the range of variation of participation for each of the five CBCs. Note that the Black Forest and Pikes Peak CBCs have been subject to much higher variability in participation that, depending upon deployment of observers, might impact the results across years. More and more-consistent participation would be beneficial for these two counts (particularly Pikes Peak).
Table 2. Range of the number of participants for five Colorado
Christmas Bird Counts, 1995-1996 to 2014-2015.
============================================
CBC Min* Max % difference
----------------------------------------------------------------------------------------
Boulder 81 137 69.1
Evergreen-Idaho Springs 42 70 70.0
Black Forest 14 39 179.9
Pikes Peak 9 21 133.3
Penrose 19 30 58.9
----------------------------------------------------------------------------------------
*Number of participants for individual CBCs on which zero
White-winged Juncos were detected are not readily
available without more effort, so values presented may not
represent actual minima and maxima for those counts.
Sincerely,
Tony
Tony Leukering
Largo, FL
http://cobirds.org/CFO/Resources/Columns.aspx?id=2
Since one of my favorite birds has come up as a topic in this venue, I find it impossible to not jump in. In my experience, White-winged Junco's primary winter range is the Ponderosa Pine belt on the east side of the Front Range and Wet Mountains. It is nowhere numerous and is nearly always greatly outnumbered by other junco taxa. This is not surprising considering that the breeding range of White-winged Junco is so small and its winter range is so much larger. So, while the taxon is fairly readily findable in winter at the foothill-plains interface (particularly at a few favored locations, such as Red Rocks Park), White-winged Junco is typically found there only in tiny numbers. Hiking uphill is the way to find them, such as illustrated by Ted's checklist.
Below, I provide links to the eBird distribution maps of White-winged Junco for select winters:
Current winter
2014-2015
2013-2014
2012-2013
As can be seen on the above maps, overall distribution in CO across recent winters does not differ in any appreciable way. As comparison to a decade ago, note the eBird map for winter 2005-2006. HOWEVER, many, many fewer birders used eBird ten years ago, so old data like these cannot be considered to accurately represent actual occurrence patterns. That leaves us with the Christmas Bird Count (CBC) as the only long-term data set available for looking at such questions. Unfortunately, there is a huge gob of methodological problems with the CBC (both CBC policy and individual-CBC mechanics) that make analyzing the data an exercise best left for a Ph.D in statistics. Though I will not delve into those problems here, a very obvious one is the effect of deep-sixing the participation fee within the past ten years ago and the resultant great increase of participation (both in observers/circle and in number of circles), which has to have had a large impact on data and their analysis. However, as a very rough indication of patterns for taxa found in sufficient numbers, it does a reasonable job; White-winged Junco is NOT such a taxon.
For the below discussion, I queried the online National Audubon Society's CBC database and obtained the previous 20-year history for White-winged Junco in Colorado and for five select CBCs. The span of CBCs queried was from the 2005-2006 season through the 2014-2015 season (the 2015-2016 data set is not complete, nor available for query). I chose five Colorado CBCs as the best representative counts to analyze Colorado White-winged Junco abundance in order to cover as much of the spread of occurrence in the state as possible while also maintaining sufficient sample size for useful analysis. The five count circles, all with extensive acreage of suitable habitat for the taxon are, from north to south: Boulder, Evergreen-Idaho Springs, Black Forest, Pikes Peak, and Penrose. The graphs use number of individuals/party-hour in order to normalize the data for, at times, highly variable number of observers across years. I made screen captures of the graphs and placed them on my Flickr site.
In Colorado as a whole, the last half of the 20-year period seems to show a decline (but see caveat above). However, the most striking feature of the graph is the great variability in abundance, particularly in the first ten years, with a fairly regular cycle of boom and bust. While weather might also explain it, my experience with that variable is that winter weather regimes are not so precisely cyclical as to explain this result. Other passerine species are known to exhibit similar boom-and-bust cycles, with Common Redpoll and Pine Siskin being two of the best examples, and these two species also match the short temporal pattern of the apparent White-winged Junco cycle. The two finch species are thought to be responding to specific food sources; the Pine Siskin to the availability of pine seeds. Considering that White-winged Junco has a high preference for pine forest, perhaps it, too, is responding to localized food resources, perhaps with larger proportions of the taxon staying closer to the breeding grounds in some years, and having to travel farther afield in others. Inter-annual variation in breeding success might also be a causative factor behind the cycle. Documentation of age structure in winter in Colorado and in the Black Hills might provide useful data at getting at the cause of the cycle.
Regardless of the cause(s) of the apparent population cycle, the system seems to have broken down in the last ten years or so, with four consecutive years of decline, from the 2008-2009 season to the 2011-2012, with a return to something of a cycle in the last four years, albeit at a considerably lower abundance value.
Looking at individual CBC results (see below), the five selected CBCs throw a huge monkey wrench into the analysis, one that might be explainable with a more-in-depth analysis, but which lies outside the scope of this post. In the below graphs, we can see that no two of the five CBC circles present results similar to each other, nor to the general Colorado results!
Boulder
Evergreen-Idaho Springs
Black Forest
Pikes Peak
Penrose
Table 1 illustrates a general north-to-south trend of decreasing reliability of detection, suggesting that there is a geographic correlate of abundance as it affects detection probability, thus of frequency, and that southern Colorado is closer to the edge of the winter range than is northern Colorado. That suggestion seems borne out by the overall eBird winter-distribution map. Somewhat contrastingly, both maximum abundance and average abundance reached their highest values in central Colorado. However, the three central counts all have a higher percentage of their circles occupied by suitable White-winged Junco habitat than does either the Boulder CBC, a large portion of which is occupied by urbanized plains habitats, or Penrose, a large portion of which is occupied by low-elevation habitats. Table 1 suggests that the southern part of the Front Range and the associated Palmer Divide is the epicenter of Colorado White-winged Junco abundance, with the Black Forest CBC circle being especially so. However, the small sample size of Colorado CBCs incorporating extensive suitable habitat (n=5) makes that assumption somewhat problematic.
Table 1. Averages of White-winged Junco abundance on five
Colorado Christmas Bird Counts, 1995-1996 to 2014-2015.
=======================================================
# years Abundance Avg.
CBC detected Min Max abundance*
-------------------------------------------------------------------------------------------------------------
Boulder 20 0.0122 0.3447 0.1309
Evergreen-Idaho Springs 20 0.0898 0.6612 0.3848
Black Forest 16 0 4.4615 0.6287
Pikes Peak 11 0 1.1195 0.1918
Penrose 11 0 0.6387 0.0826
-------------------------------------------------------------------------------------------------------------
*Abundance is generated by summing annual values of
birds/party-hour and then dividing by 20.
Table 2 presents the range of variation of participation for each of the five CBCs. Note that the Black Forest and Pikes Peak CBCs have been subject to much higher variability in participation that, depending upon deployment of observers, might impact the results across years. More and more-consistent participation would be beneficial for these two counts (particularly Pikes Peak).
Table 2. Range of the number of participants for five Colorado
Christmas Bird Counts, 1995-1996 to 2014-2015.
============================================
CBC Min* Max % difference
----------------------------------------------------------------------------------------
Boulder 81 137 69.1
Evergreen-Idaho Springs 42 70 70.0
Black Forest 14 39 179.9
Pikes Peak 9 21 133.3
Penrose 19 30 58.9
----------------------------------------------------------------------------------------
*Number of participants for individual CBCs on which zero
White-winged Juncos were detected are not readily
available without more effort, so values presented may not
represent actual minima and maxima for those counts.
Sincerely,
Tony
Tony Leukering
Largo, FL
http://cobirds.org/CFO/Resources/Columns.aspx?id=2
Reply all
Reply to author
Forward
0 new messages