The article has significant flaws in methodology and interpretation, which are discussed in detail at https://geneticliteracyproject.org/2019/12/17/whats-missing-from-claims-.... Some of that discussion is excerpted below:
Yamamuro et al. hypothesize that these declines were due to neonics carried into the lake from nearby rice fields and causing catastrophic declines in "zooplankton biomass" — the invertebrate species of insects and crustaceans that make up much of the fishes' diet.
At first glance, Figures 1 and 2 appear to support that thesis. In the top graph we see neonic quantities rising with time, while, in the bottom graph, zooplankton biomass dramatically collapses. Although this is only a correlation, or association, which would not prove causation, it is highly suggestive.
There's a critical problem with this apparent correlation, however. On closer examination, it is evident that the dates in horizontal axes of the two figures are not aligned; and if both graphs are centered on 1993, the first year neonics were used, it becomes clear that zooplankton biomass had been in precipitous decline for at least a decade before neonics were introduced on rice farms.
The fact that biomass more or less bottomed out in 1993 is clearly the culmination of a trend, a steep and persistent decline that had been going on for over a decade and very likely much longer. The fact that the use of neonics (which began in 1993) overlaps for one year with the tail end of this trend hardly can be said to represent a correlation, especially as the records show negligible quantities during the first six years of usage, from 1993-1998.
To state the obvious, neonics can't be blamed for trends that started long before they were ever applied to farmers' fields. But if one looks for the authors to provide an explanation of what was causing the rapid decline in all those years preceding the advent of neonics, one looks in vain. (More about this below.)
The supposed correlation becomes even less convincing given that Figure 1 doesn't even represent actual measurements of neonic levels in Lake Shinji. The only actual measurements of neonic levels in the lake water were made by the authors in one year, 2018, a quarter century after the 1993 reference year. The authors, in fact, have no idea what neonic levels in the lake were prior to 2018. They are asking us to accept, instead, a kind of surrogate for actual measurements – namely, the total sales volume of neonics in the entire Shimane Prefecture, in which Lake Shinji is located.
Something is obviously very fishy here, which makes one wonder not only about the competence of the investigators but also whether, in overlooking these points and those discussed below, the peer reviewers and editors of Science, the journal in which this study was published, were comatose.
That brings us to the critical issue of dose-response, or in this case, the actual levels of neonics in the water. From April through June 2018, the authors took several measurements in three different sampling sites in the lake. None of these measurements rise above the level of quantification, i.e. the amount that can be accurately determined given the limits of present-day technology, so they are hard to accept at face value. Even so, the one finding reported in the study itself – the highest and only significant concentration, which the authors calculated by adding all the different neonics' levels together – is not high enough to do the widespread damage to insect populations the authors posit.
This is especially true for crustaceans, which form a large part of the zooplankton biomass and are a critical part of the fish diet. As found in a 2018 study, crustaceans are generally much less sensitive to neonics than are the insect species on which the ecological benchmarks, or safety levels, are based.
Things look even worse for the November Yamamuro study if one applies the researchers' surrogate measurement (neonic sales) consistently. Compared to 2016, when sales in the prefecture were about 4,000 kg, sales during the first six years neonics were on the market (1993-1998) were below 200 kg/year; in other words, back then, neonics applied were far below amounts likely to do appreciable harm even to the most sensitive insect species.
Perhaps realizing they had a problem, the authors spent considerable time highlighting neonic concentrations found in the Sagami River as it flows through metropolitan Tokyo. For those not familiar with the geography of Japan, that's about 450 miles away from Lake Shinji, on the other side of at least three massive mountain ranges, and so in a completely different watershed — not to mention, in the middle of one of the largest, most populated cites on the planet. The possible relevance of this was not explained.
Given the rather obvious chronology problems of attributing fish declines to neonics, it might have made sense for the authors to look for other possible causes. They aren't hard to find.
Heavy metal pollution. The lake has been undergoing significant ecological changes since at least 1922, when a canal was dredged between Shinji and nearby Lake Nakaumi, which caused saltwater to flow into the former, turning it brackish. And since 1966, according to the World Lake Database, "efforts have been made to establish a new industrial zone along the coasts of the two lakes" (Shinji and Nakaumi).
One might expect that a half-century of industrial development could lead to serious problems with chemical pollution affecting aquatic organisms. In fact, a 2011 study of chemical contaminants in Lake Shinji found exactly that, noting the lake's sediments "are moderately to strongly polluted with respect to As [arsenic], moderately polluted with Pb [lead] , Zn [zinc], and Cr [chromium]..."
Other researchers (Hook and Fisher 2001; 2002) looking at the effects of zinc and other chemical pollutants on copepods (the small crustaceans that make up a good deal of the zooplankton food supply for fish in Lake Shinji) found that "exposure to contaminated food resulted in assimilation of the metals primarily into internal tissues" and that with "direct exposure, the metals showed up in the exoskeleton" leading "to sublethal effects (e.g., decreased egg production and hatching, ovarian development, and protein concentration in eggs) at concentrations 2~3 orders of magnitude less than lethal concentrations."
Eutrophication. Just as Yamamuro et al. fail to mention Lake Shinji's chemical pollution, they also neglect to alert readers to the well-known, long-standing problems the lake suffers from eutrophication, the depletion of oxygen due to the proliferation of algae and other organisms. As Japan's Ministry of the Environment noted in 1996: "Deterioration of water quality by socio-economic activities in catchment area. Eutrophication with water bloom in summer."
Eutrophication is generally caused by excessive nutrients from sewage, fertilizer and other organic and inorganic pollutants that result in explosive algae growth, the buildup of hydrogen sulfide, and oxygen-depleted water.
Although the researchers say that oxygen and some other measurements in Lake Shingi were similar before and after 1993, the effects of eutrophication on lake ecology are highly complex, can manifest over long time periods, and even to this day are not fully understood. In fact, the study's lead researcher, Ms. Yamamuro, acknowledged these issues repeatedly in a study of Lake Shinji and Lake Nakaumi, published in 2000:
In recent decades, these waters have been strongly affected by eutrophication which is accompanied by oxygen depletion. Detailed study of the migration patterns and growth rates of particular fish species in the area would contribute to improving the management of local fisheries as well as to understanding the reaction of fishes to eutrophication....
The of the bottom-dwelling fish, however, might result from increased eutrophication.... The increase in plantivorous species, K. punctatus and S. zunasi, also might be related to an increase in phytoplankton, although relationships in the lake's food web are not clarified.
Habitat destruction. Other studies have pointed to severe disruptions of aquatic habitat that have caused significant declines in fish populations. For example, the study co-authored by Yamamuro in 2000 pointed to a large "decrease in Carassius sp" which the study attributed to the loss of "spawning and nursery areas, because 75% of the natural coast of the lake has changed with the construction of an artificial wall (Environmental Agency, 1993) during the course of urbanization of the area."
Perhaps it's not a coincidence that this wall appears to have been built around 1993, the authors' pivotal date in this new study when, they claim, the lake's environmental collapse began.