Biodiversity Part 2: Land Use and Biodiversity
In our biodiversity series, we address biodiversity and its impact assessment. In part 1, we discussed some definitions and units. In this article, we look at land use and how land use and land-use change affect biodiversity. How do we measure the damage, and who pays?
In our previous article on biodiversity, we challenged you to think about using biodiversity as a metric to express the impact of land transformation and to get companies to act. Did you think that’s impossible? Or did you think it is both impossible and necessary at the same time?
It is generally understood that loss of habitat is one of the biggest threat to natural systems and biodiversity. I will try to describe how biodiversity and land use are related.
Land Use and Who Pays for the Damage
There are two aspects to the term ‘land use’:
1. How an area is used at a certain time.
2. The transformation of an area from one way of use to another. Let’s say from forest to agriculture.
In impact assessment and LCA, we keep these two aspects separate: we think that a farmer who uses his land year after year should not be held responsible for the impact his predecessor caused when converting the land. In LCA, therefore, the person who is responsible for converting the land is charged with the conversion and the restoration time, which can be very long. This is different from the rules in the GHG protocol: there, a farmer is charged every year if the transformation occurred in the last 20 years.
Effects of Land Use on Biodiversity
In the figure below, we plot species richness over time, showing what happens when someone converts a piece of nature. During conversion, there is a rapid decrease in species richness. After that, we assume there is a stable situation and that there are no further changes in species richness. We also assume that at some point, the occupation stops and a restoration takes place that results in a gradual increase of species richness.
Whether a natural system can restore itself to its original species richness is unclear, as is the time needed for this to happen. However, we find it reasonable to assume that we need to compare the current impact to a situation where no impact takes place. The surface area of the square and the triangle, multiplying the species lost by the time before they were restored, indicate the total damage we inflict.
Attributing Damage in Impact Assessment
If a farmer uses land that was already converted, we cannot hold him accountable for the impact of the conversion. However, as long as the farmer uses the land, it cannot be restored to nature. So the impact the farmer is held accountable for is the impact that happens while he is using the land, calculated by multiplying the duration of the occupation with the loss of species (measured in potentially disappeared fraction, PDF) and of course the size of the area in question. As discussed in our previous biodiversity article, that results in the unit PDF.m2.yr.
In assessing the impact, we hold the person who converted the land responsible for the damage done during conversion and the damage done during restoration. As the time it takes to convert an area is usually much shorter than the time it takes to restore an area, the first period is negligible. Let’s compare the impact of land use to the impact of conversion. If someone uses a square meter of land to produce a crop for one year, and this keeps the number of species at 80% less than the area’s unconverted state, the damage is 0.8 PDF.m2.yr.
In contrast, if someone converts a natural area that takes 100 years to go back to the original number of species, the damage is much greater, even if the number of species is also reduced by exactly 80%. As you see in the graph above, the assumption is that the species count will gradually increase after the land is abandoned. That is why we divide the total by two, bringing the impact calculation to 0.8*100/2=40 PDF.m2.yr.
What is the Natural State?
While it seems convenient to talk about the natural state of land when assessing the impact of changes in biodiversity, it is in fact not so easy to describe what the natural state is. The figure below is a biodiversity map of the world, showing large differences in the number of plant species. In ReCiPe we decided that the relative number of species is more important than the absolute number. The 20-200 vascular plant species that can be found in the Sahara together form the ecosystem there. Losing half the number of species there is seen as being equally important as losing half the species in the Peru, even though Peru’s ecosystem has more than 5000 species. This implies that losing one species in the Sahara has a much bigger impact than losing one species in Peru.
The next question is how do we know how many species inhabit agricultural and urban land. For agriculture, the answer seems easy: a farmer only wants one species on the land. However, agricultural farmlands are in fact quite rich in species, because there is a rich diversity in the edges, the small unused plots and the pathways. A very detailed inventory is available, made in the UK, where experts counted species on the land itself (the X-plot), the area just inside the fence (the A-plot) and the area just outside (the B-plot).
This inventory shows that the species richness is not really determined by the crop itself, but by the presence of edges, hedges, and small bushes or rows of trees. So it makes a difference if an area is a large-scale monoculture or a small scale traditional agricultural landscape. Unfortunately, that aspect is usually not reported in LCI databases. We use an average species loss of around 40% compared to the reference. To what extent this has global validity is unclear, but we would be surprised if that number is completely wrong in other parts of the world, except when there are huge monocultures.
In the previous first article, we developed a metric for measuring biodiversity as such and assessing the impact of changes in biodiversity. In this article, we now developed an approach to link this biodiversity metric to data about land occupation and transformation. Here, too, we can ask the question: are these metrics reliable enough? The method presented here is somewhat unlike traditional environmental modelling in LCA. For the core part of the model, we need to carefully count species in meadows and edges of meadows. These counts tell us the fraction of species that apparently got lost due to land use. This sure sounds robust, but is it? For instance, can we assume that the fraction of species lost is the same all over the world? Are we right in assuming that the fraction of species is more important than the number of species? Time to make your judgement, and think of a better way.