Biodiversity conservation has become a major issue; It concerns all the activities of society, from town planning to agriculture, including transportation, industry and tourism.
However, in this field, not all players always possess the necessary scientific knowledge to deal directly with these issues; Ecology is not their profession and remains a complex science.
Let’s take an example of artificial lighting at night, which is likely to generate light pollution for living organisms; Since they are not biologists or ecologists, professionals in this sector (from manufacturers to network managers) will not have the expertise to accurately assess and correct these impacts.
All too often, executive actors find themselves in a dead end: they have the will, and even the obligation, to minimize their effects, but they don’t really know how to do it. Then they do what they can, within their means, relying on their own experience, or even common sense.
At the same time, there is also the question of potential biases (or even conflicts) if the measures that will be applied to correct an activity harmful to biodiversity are identified by the actors that generate that activity.
We understand here the need for outside expertise, provided by a competent third party.
In search of proven and proven facts
The first option for outsourcing this expertise is to refer to specialists – scientists and ecologists – who will give their opinion based on their knowledge and practice. This is the “expert opinion” we often use in France.
However, the danger of this approach is that it ends up with a subjective or partial result: the best experts cannot know everything or have an overview of everything that science has produced! And how do we make sure that an expert’s opinion is unbiased per se, even in good faith?
To avoid this predicament, another option is to rely on the scientific literature, a written record of research activity. The goal here is to rely on proven facts that have been proven through experiments. This is calledEvidence-based information.
But the scientific literature today is very abundant! Research publishes new articles all the time and we make use of everything that has been published in the past, which is increasingly being distributed in digital form.
For example, the Web of Science Core Collection database, one of the largest in ecology, includes more than 60,000 documents with the phrase “climate change” in its title.
How can this knowledge block be realistically evaluated?
To meet this need, it is customary in the scientific field to make bibliographic summaries – also called “reviews”; It corresponds to scholarly articles which in turn aggregate many articles on a particular topic.
As an illustration, we can cite several bibliographic summaries published in 2020 and 2021 on the effect of night lighting on living organisms.
Read more: How light pollution became everyone’s business
These documents provide valuable “state of the art”, but they generally do not lead to a quantitative result. Moreover, the method is not always disclosed and many biases persist, for example the way in which composite documents are selected.
To deal with the large volumes of literature to be treated, an understandable reaction might be to take only the articles most cited, or by the most famous authors, or published in the best-rated journals.
Using algorithms to sort large amounts of literature is a well-studied but still in its infancy path; It asks the same questions of neutrality: “Who” programs the bots, “How” and “For what purpose”?
The so-called “systematic” reviews
To solve these various problems, a unified approach is recommended in order to make the bibliographic synthesis work as efficient, robust and impartial as possible.
This is the whole purpose of systematic reviews, the aim of which is to achieve the greatest possible comprehensiveness (completeness of the analyzed literature), objectivity (only the most reliable publications are used), transparency (the process is fully described and all decisions are tracked) and repeatability (thanks to this Transparency, anyone can re-exercise and compare the result obtained with the results of the authors).
Systematic review, although less well known, is similar to a “standardized” meta-analysis, particularly in the process that precedes statistical analysis (bibliographic search and screening). In addition, the systematic review includes the original step of evaluating the level of bias in the composite publications, so as to measure the reliability of their findings and, if necessary, weigh them.
The method of systematic reviews was started in the medical field by Dr. Archie Cochran in the UK in the 1970s and has been published in the environment since 2010. Here, it is the Collaboration for Environmental Evidence (CEE), based in Wales, who has formalized and suggested the method that They must be followed by the reinforced guidelines regularly.
In France, where we are more accustomed to the collective experience, things have also changed in recent years. The Foundation for Research in Biodiversity (FRB) is now promoting systematic reviews and is the French center representing the EEC since 2015.
The Joint Service Unit Natural Heritage (PatriNat) produced the first systematic review of transportation infrastructure in 2018, followed by a second review in 2020.
Other work, eg on noise pollution, has been published or undertaken, in particular on climate change and protected areas as part of the Life Natur’Adapt project.
Systematic reviews are part of the interaction between research and operational actors.
They are generally produced in response to a request by an entity – a public authority, a company, a developer, a natural area manager – that will have to make a decision in relation to its activity to assess or reduce its impact on the environment; Or even to stimulate new public policy or legislation (in the case of a ministry).
This tool is particularly useful for topics where there is a lot of literature and contradictory results that can lead to ambiguity in the decision to be made, or even confronting lobbies seeking to exploit certain results in one direction (pro) as in the other. opposite).
To take an example: We know that the effect of artificial lighting on species depends specifically on wavelengths (ie the color or colors in the light).
If we consider the studies separately, we may have the impression of contradictory results: depending on the species and biological functions studied (mobility, physiology, chronology, etc.), they are not the same color (and therefore not the same type of lamp). ) and that would be a problem.
It is usually a topic that needs to be undone, in order to have a reliable world view and to get out of the particular situation. Especially because it is a topical topic with the massive proliferation of new light sources (LEDs) that significantly modify the spectral composition of light pollution.
What guarantees do systematic reviews provide?
Systematic reviews in Ecology are published inEnvironmental Evidence Journal (EEJ), an open access EEC peer-reviewed journal that ensures readers respect EEC standards.
It is also always preceded by a “protocol” article that clearly states which method will be used (such as here or there).
With this transparency and reliability, in addition to informing decision makers, systematic reviews can also respond to the growing crisis of mistrust we are seeing on the part of the general public towards expertise.
However, nothing prevents the author from submitting his systematic review to any journal that accepts it. Then we see more and more systematic reviews appearing that do not or only partially respect the EEC standards (bibliographic search is not exhaustive, article sorting criteria are unclear, no critical analysis was performed, etc.).
To try to address this issue, the EEC has prepared an assessment of the systematic reviews which have not been published in the EEJ, in order to indicate the extent to which the various EEC standards are respected; This is the CEEDER programme. More than 1,000 reviews and meta-analyses published between 2018 and 2021 have already been evaluated to date.
But let us stress that despite these guarantees, the best guarantee remains the reader himself, who should read and not adhere to any publication – whether it is a scholarly article, a systematic review, or a meta-analysis – with a critical eye. His result by a simple principle but because he clearly understands how it was obtained.
To do this, it is more necessary than ever to provide as many people as possible with a minimum of scientific bases. This includes in particular education, information and awareness raising.