Guest Essay by Kip Hansen
Reading widely across the literature of many scientific fields broadens one’s view of the rate of scientific progress — both our increased knowledge and our “steps backwards”. A recent piece in Nature Ecology & Evolution represents both: one of these advances and one of these “steps backwards”.
The article is a Brief Communication titled “Moth biomass increases and decreases over 50 years in Britain” by Callum J. Macgregor, Jonathan H. Williams, James R. Bell and Chris D. Thomas. (a full .pdf is available via SciHub with doi: 10.1038/s41559-019-1028-6). Here’s the abstract:
“Steep insect biomass declines (‘insectageddon’) have been widely reported, despite a lack of continuously collected biomass data from replicated long-term monitoring sites. Such severe declines are not supported by the world’s longest running insect population database: annual moth biomass estimates from British fixed monitoring sites revealed increasing biomass between 1967 and 1982, followed by gradual decline from 1982 to 2017, with a 2.2-fold net gain in mean biomass between the first (1967–1976) and last decades (2008–2017) of monitoring. High between-year variability and multi-year periodicity in biomass emphasize the need for long-term data to detect trends and identify their causes robustly.”
In surprisingly straightforward (and diplomatic) language, the authors start out with a clear explanation of the scientific problem:
Insufficient Comparable and Consistently Collected Long-term Data
“Reports of declining insect biomass give credence to the notion that insects are at the forefront of a ‘sixth mass extinction’. However, some reports have received criticisms for poorly justified conclusions, potential biases and extrapolating beyond the data. Regional abundance and distribution declines have taken place in many individual species but populations of other species are stable or increasing, leaving uncertainty over the consequences for biomass change and associated ecosystem processes. Nearly all existing estimates of biomass change lack continuous, systematically controlled monitoring or sufficient survey sites. Hence, analyses of continuously collected data from multiple sites and environments are necessary to establish the robustness of the conclusion that insect biomass is declining.” [footnote numbers removed — see original for these references — kh]
There is a Press Release related to the article here.
The authors point out: “Contrary to previous reports of insect biomass change, moth biomass increased before it declined and remains higher than in the late 1960s (Fig. 1).”
[click image or here for larger version]
There are several interesting things to notice in the graphs presented in Figure 1 but first and foremost, in my mind, is that in panel “a” we see that Total Moth Biomass varies on an annual basis as much as three orders of magnitude — from a low of 1,000 mg up to 1,000,000 mgs. The Supplemental Information shows the mean varying from 8,000 mgs to 80,000 mgs. That’s a lot of variation.
The author’s main points are obvious in this one four-panel figure: 1) Total moth biomass rose from the mid-1960s to about 1980. 2) from 1980 to 2017, there is a declining trend, but totals remain above 1960s levels. 3) There are “winners and losers” — Panel “b” — the moth family Erebidae has been steadily increasing across the entire time period whereas Geometridae and Noctuidae show the rise-then-decline pattern. and 3) The changes do not seem to be due to land use changes, as the pattern is similar across all four land use categories — arable land, woodland, urban and grassland.
This data supports their conclusions that “analyses of continuously collected data from multiple sites and environments are necessary to establish the robustness of the conclusion that insect biomass is declining” and that “Such severe declines [“insectageddon”] are not supported by the world’s longest running insect population database”. This paper represents a step forward in its own field, and points to other papers that, through their failures, have represented a step backwards.
Even more impressive is the amount of detailed Supplemental Information that is attached at the end of the full .pdf file for this study [use the doi and SciHub to obtain a copy]. There are charts by land use, charts of periodicities in the data and many more.
Callum J. Macgregor, Jonathan H. Williams, James R. Bell and Chris D. Thomas, our authors, are to be congratulated for their extremely well-done paper which is complete, transparent and supplies all the data (or links thereto).
This is Real Science.
And “Insectageddon” ? — probably not.
# # # # #
I encourage readers to obtain a full copy of the .pdf of this paper, for an entirely different reason.
At the end of the paper, after all the author’s work and supplemental information, there is a two-page section headed Reporting Summary:
The very existence of this section represents the hope and future of Science and of Science Journals. The topics covered in the Reporting Summary include:
In this case that the authors followed all the statistical guidelines that applied to their study.
“Data collection: No software was used in the data collection process (data were obtained from Rothamsted Research and other sources as described in the Methods section). Identical datasets can be obtained from these sources on request.”
Software and code:
“Data analysis: All analyses were conducted in open-source software, R version 3.6.1, as stated in the manuscript. All R scripts, from initial processing of datasets to final analyses, are archived online at Zenodo (doi: 10.5281/zenodo.3356841).”
The authors report their study to be in “Ecological, evolutionary & environmental sciences”
Ecological, Evolutionary & environmental sciences study design:
In this section, the authors report their study design in detail:
[click image or here for larger version]
While these NatureResearch steps may not be everything that John P.A. Ioannidis and others who are striving to reform research standards towards replicability and full-disclosure science reporting would demand, it is a beautiful thing to see. Imagine if every published Climate Science paper came with a required, field-appropriate section like the above. Or for that matter, if all science used by the E.P.A., the C.D.C. or the F.D.A. was published to these standards for explicit declaration of study design and full data and code availability.
I had not realized that the Nature group had advanced so quickly, and admirably, on this topic. They have my congratulations and thanks.
# # # # #
On a personal note, I have been away sailing, serving under my youngest son, now a Captain and Master in his own right — from the mid-Hudson Valley of New York (on the Hudson River) to Elizabeth City, N.C. Wonderful trip with three young men, all ranked as Captains.
Insects are a “subject of interest” for me — and I have hundreds of pictures of different “shield spider” species that live in the Dominican Republic — thus my first interest in the paper highlighted above.
In the end though, I found the NatureResearch’s Reporting Summary more significant.
Glad to read your comments on either topic.
# # # # #