[LTER-hbr-im] questions re Hubbard Brook bird population changes (website raw data vs. trends described in your new HB book)

[Richard T. Holmes]

Dear Nora,

Thanks for your message.  I’m pleased that you are interested in the bird trend patterns at Hubbard Brook. Here are some responses to your specific points and questions.

My first comment and recommendation is: Look to the data! The information in the book was an overview and highlight of major findings, and there is a lot more information in the details, which could not be covered in that format. And, most importantly, the last full statistical analysis of bird trends on the long-term plot are those reported in a paper published in 2001 (see attached). The next analysis will take place after the 2018 field season, when the data will cover a 50-year period. So, keep that in mind.

Also, it is important to realize that these data have been collected over a long period, and lots has happened over that time. If you are just looking at numbers in the first and the last 4 years, you may miss what has happened in between. For example, some species were low in abundance early on, then increased, then decreased again, so you really need to look at the patterns to understand what is happening. Also, some species, such as the Philadelphia Vireo, have essentially disappeared from the HB valley, others have declined and/or shifted their distributions within the plot (and the valley) as the vegetation has changed due to forest maturation, disturbance events, and due to the effects of a warming climate. Finally, the study area from which these data come is very small, only a 10-ha grid embedded within continuous forest, which means that statistical power for analyzing trends for the less common species is often limited. As an aside, the initial goal of the research in 1969 was to determine the ecological role of birds in forest ecosystem energy flow and nutrient cycling and to do so, we needed to obtain quantitative estimates of the number and biomass of all birds occupying a given area of forest. Given the great effort this involved, we felt that a 10-ha plot was the largest area we could manage and the results would be sufficient for our stated goals at that time. We had no idea at the time that the plot would be used for a long-term study.  In hindsight, it would have been nice to have had a much larger plot, which would have allowed for more birds and larger sample sizes for statistical analyses. However, this has been remedied since 1999 by censuses of birds throughout the 3000-ha Hubbard Brook valley (see valley-wide data set on the Hubbard Brook Ecosystem Study webpage).

Now for the specifics. Hairy Woodpeckers have fluctuated in numbers over the years, and this species provides a good example of a species where looking at just the first few years and the most recent give a misleading impression of trends – the time scale is important. In contrast, the Swainson’s Thrush may show a decline when we do the 50-year analysis.  It is a species more common in higher elevation, more coniferous forests, and with recent climate warming, its distribution and center of abundance may have shifted to the upper slopes of the valley in recent years. The valley-wide data will give us a chance to check that hypothesis. Downy Woodpeckers may also be a candidate for decline in recent years, but there are so few of them on the plot that statistical power will be limited. The recent decline in the numbers of Black-throated Blue Warblers relates to a thinning of the understory in the 10-ha area; its numbers on a large, 200-ha plot on which we are now studying their demography remains stable, as it do their numbers on the valley-wide scale.  So, again, patterns and trends vary with which temporal and spatial scale are being considered.

In Table 14.1 of the Hubbard Brook book, the left column is labelled ‘declining’ species.  In fact, all 5 of these species no longer occur on the 10-ha study plot, even though some like Least Flycatchers and American Redstarts were once among the most abundant species in that area.  So, ‘declining’ was probably the wrong word to use, as they have, in fact, declined to zero within that patch of forest. Most of these species still occur elsewhere in the Hubbard Brook valley, although very localized and in low numbers.  Only the redstart has been sufficiently common in the last 15 years or so to warrant analyses of their abundance on the valley-wide scale; these analyses (unpublished but currently being prepared for publication) show redstarts declining between 1999 and 2013 at that larger scale. So, I think it is safe to conclude that the redstart and these other species have declined at Hubbard Brook.

Let me know if you have more questions, and I would be interested in hearing more about your project.

With best regards,

Dick Holmes

Richard T. Holmes
Harris Professor of Environmental Biology Emeritus
Department of Biological Sciences
Dartmouth College
Hanover, NH 03755


From: Nora Hanke <nhanke at antioch.edu>
Date: Friday, June 16, 2017 at 8:43 PM
To: "hbr-im at lternet.edu" <hbr-im at lternet.edu>, "Richard T. Holmes" <Richard.T.Holmes at dartmouth.edu>
Subject: questions re Hubbard Brook bird population changes (website raw data vs. trends described in your new HB book)

Hello
I am a graduate student in Environmental Studies at Antioch University and have been enjoying your' wonderful Hubbard Brook book. I am researching some aspects of the HB studies and don't understand what appear to be data discrepancies.

Specifically, I am confused by looking at the raw data published on the HB website in comparison to the species trends documented in Hubbard Brook (2016). (I know the latter are for a period ending in 2013.)
For example, Hairy Woodpeckers are described as fairly stable, but comparing their numbers in the first 4 years of the study with the period 2012 through 2015, they appear to have doubled. Swainson's Thrushes are described as stable, but the average of their no.s in recent years are under a quarter of their abundance in an average of the study's first few years. Philadelphia Vireos are described as declining (rather than disappeared), but they have not been noted in the HB study site since 1982, according to the website's data. Similarly, Veeries are described as declining, but were last counted in 2005.
I am guessing that the HAWO and SWTH number changes are not statistically significant, or else your threshold for defining population change is greater than the population changes noted in the two comparison periods I am studying. But I really don't get the interpretation for the birds that are gone - according to the HB website published abundance data.
I apologise if I am obtuse or show a horrific lack of understanding biostatistics or interpreting raw data. I will be studying biostatistics next semester. In the meantime, I am currently writing a paper on your bird studies and deeply mystified.
I would appreciate some clarification, if you have time to respond.
Thank you!
Nora E Hanke
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lternet.edu/pipermail/hbr-im/attachments/20170617/d07c4153/attachment-0003.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Holmes & Sherry 2001. Auk.pdf
Type: application/pdf
Size: 298252 bytes
Desc: Holmes & Sherry 2001. Auk.pdf
URL: <http://lists.lternet.edu/pipermail/hbr-im/attachments/20170617/d07c4153/attachment-0003.pdf>