Could something have happened between 2006and 2007?

Posted by rdegoede on 21 Oct 2017 at 11:44 GMT

When studying the trend in figure 2 panel A, it stroked me that something strange might have happened between 2006 and 2007. Based on the raw data obtained from the appendix, I calculated that between the years 1989-2006 the biomass yield (g) per day was relatively constant and was on average 5.8 (CV=34%) g per day. However, after 2007 it dropped to an again relatively constant average of 2.3 (CV=39.2%) gram per day between the years 2006-2016 (the year 2011 had only 4 observations and a relatively high biomass per day). When fitting a regression line for only the years 1989-2006 the slope is -0.14 and the r2 is 0.193. For 2007-2016 this is: slope=0.013 and r2 =0.002. When using the whole time interval between the years 1989-2016, the regression line has a slope of -0.20 with an r2 of 0.579. Furthermore, the weak negative trend between 1989 and 2006 seems very much affected by the first observation year (1989). Without the year 1989, the slope becomes as low as -0.07 with an r2 of only 0.069.

So my alternative interpretation of the data in figure 2 panel A is that it might be that something happened particularly between 2006 and 2007 which caused a sudden dramatic decline in biomass yield per day. Whereas, in the time interval between 1989 and 2006 as well as in the time interval between 2006 and 2016 no significant decline in biomass occurred. If my observation is valid, what could have happened between 2006 and 2007? Was there a real decline in biomass or might there be another (methodological?) explanation?

No competing interests declared.

RE: Could something have happened between 2006and 2007?

dsylvaticus replied to rdegoede on 23 Oct 2017 at 08:09 GMT

One methodological reason for this observation could be that the average (yearly) sampling effort has first steadily risen from 1989 up to 2005 and has been declining since then with a single exception in 2014. A plot of this trend (aggregated by habitat type) is available at https://imgur.com/a/fpnIE

No competing interests declared.

RE: Could something have happened between 2006 and 2007?

rdegoede replied to rdegoede on 23 Oct 2017 at 21:06 GMT

Graphs related to the above given post are available at: http://www.ronecology.nl/...

No competing interests declared.

RE: Could something have happened between 2006and 2007?

GeorgKeckl replied to rdegoede on 16 Jan 2020 at 09:56 GMT

The most likely cause of the decline in insect biomass is the sharp decline in unused arable land around nature reserves. Subsidized, unused fields with very little maintenance attract many insects. The measuring points for the insect mass were all in nature reserves. Nature conservation areas are mostly in arable less fertile areas. In the vicinity of the nature reserves, many arable areas were temporarily no longer used against a subsidy because these areas are also not very fertile. In 1994 1,439,000 hectares were not used (= 12.2% of all arable land), in 2007 648200 hectares, in 2008 309500. The strong promotion of "bioenergy from the fields" to protect the climate has increased the demand for arable land. More data See page 53 ("Stilllegungsfläche") in https://www.umweltbundesa...

No competing interests declared.

RE: Could something have happened between 2006and 2007?

pstoop replied to rdegoede on 13 Jul 2020 at 11:00 GMT

There is a very plausible explanation of this change in Supplementary Data File S1.
If you check the average time between insect collection dates (difference between columns F and E), you'll see that the traps were checked in average every week on 1989-2006 period. and every 2 weeks between 2007 and 2016. This longer time spent in the traps probably influenced the dishydratation rate of the insects, and their wet biomass when they were weighed at last.
I am not sure this sudden change needs further explanations.

Competing interests declared: I work in ITK, a company collaborating with BayerCropscience, Syngenta, and agricultural coopératives like Land o Lakes