On oil palm & deforestation

“If a tree falls in a forest and no one is around to hear it, does it make a sound?” Who cares! As long as the forest is not within the official forest estate map, it doesn’t matter!

At least that is what I think was in the minds of researchers from Bogor Agriculture University (IPB) that recently presented their findings on the link between oil palm and deforestation in Indonesia.

Their research, titled “History of the Land Status, Land Use, and Biodiversity within Indonesia Oil Palm Plantations”, has received major media attention in the past few weeks. All of the national media that covered this research paint oil palm industry in a surprisingly positive light on their headlines; that oil palm is not linked to deforestation in Indonesia (you can check the stories on CNN Indonesia, SindoNews, AlumniIPB, Sawitindonesia).

CNN Indonesia quoted the lead researcher — Prof. Yanto of IPB — that he is very certain that oil palm plantations in Indonesia are not the cause of deforestation. He then argued that most of these oil palm plantations were established in lands that are not coming from forest estate.


Forest estate ≠ forest

Now here comes the tricky part. In Indonesia, the term ‘forest estate’ (kawasan hutan) is a legal term that refers to delineation of land mass that is designated for forest uses — not a term that refers to actual biophysical existence of forests.

The boundary of forest estate was first established in 1982 with the enactment of Ministry of Forestry Regulation on Forest Land Use by Consensus (Tata Guna Hutan Kesepakatan-TGHK).

To simplify, ‘forest estate’ in Indonesia refers to land use designation solely based on Ministry of Forestry map (now Ministry of Environment & Forestry), but the land cover within forest estate can be varied. It can be natural intact forests, natural degraded forests, tree plantations, or even completely non-forested.

Here is a breakdown of simplified land cover categories within forest estate in Indonesia (taken from Ministry of Forestry Statistics, 2013 data):

A quick recap on how Indonesia manages its land, basically, the country’s total land mass can be categorized into two main legal designation: 1) forest estate (64% of total land mass); and 2) non forest estate (36% of total land mass). Non forest estate land is often referred to as “other uses” (Area Penggunaan Lain — APL), as it is designated for land uses other than forestry.

On deforestation definition

As you can see on the chart above, almost a third of the forest estate is currently non-forested. But on the flip side, there are close to 10 million hectares of forested land within the APL area (1.3 mha of intact primary forest, 5.3 mha of secondary forest, and 0.9 mha of plantation forest).

Legally, these 10 million hectares of land can be converted into other land uses anytime; these includes but not limited to public housings, toll roads, or oil palm plantations. And by the way, oil palm is not recognized as forest crop within the country’s regulation, so it is illegal to have oil palm plantation within forest estate. It can only be established within APL area.

Because of this land use designation, 10 million hectares of forested land within APL area will always be prone to conversion. If an oil palm company bought a piece of intact primary forest land within APL area and managed to secure their oil palm license, they can clear cut the area and establish oil palm plantation without breaking any law.

But, more disturbingly, this conversion will not be accounted as “deforestation” to some people, because the legal designation of the land is not “forest estate” in the first place.

So, in theory, Indonesia can cut down 10 million hectares of its forest within APL area, but this will never be marked as deforestation. This is essentially what Prof. Yanto explained in his media interview.

The narrow legalistic definition of deforestation in Indonesia has been one of the most common argument voiced by the proponent of extractive industries. You can read FAO’s definition of deforestation here. Even the Ministry of Forestry has recognized and published statistics that covers deforestation within and outside of forest estate.

By saying that ‘deforestation’ only applies to land use conversion from ‘forest estate’ into ‘APL’, the research has already put a blind eye to millions of hectares of forests within APL that were converted into oil palm plantation.


On research samples

From the executive summary of the research (unfortunately I couldn’t find the full report), there are 8 oil palm plantations managed by companies, and 16 units of oil palm smallholders observed as samples for this research. All of them are located in Riau province, Indonesia.

In terms of size of the total samples, the observed area for company-managed oil palm plantations and smallholders are respectively 46,372.38 Ha and 47.5 Ha.

Now let’s put this in perspective.

According to 2015 statistic provided by Ministry of Agriculture, the total planted area for oil palm in Indonesia is 11.3 million hectares. Unfortunately they do not provide the detailed breakdown of this number by companies and smallholders area. Using this conservative number as a reference, the total observed area for this research is only 0.41% of the total oil palm plantation in Indonesia.

The executive summary also does not provide any clues on how the research selected its samples, but since they are all clustered in Riau, it seems reasonable to think that it is not done randomly.

With this very low number of samples and nonrandomized sampling, surely Prof. Yanto should be more cautious in making a very bold statement that oil palm is not the cause of deforestation in Indonesia.


On validity of oil palm concessions data

Indonesia’s oil palm plantation data is known as one of the most elusive land use dataset. No government agency have the most updated and definitive spatial database of oil palm concessions, both for companies and smallholders. Not even the Ministry of Agriculture. And efforts to push the government and companies to make the spatial database accessible for public remains fruitless.

Why is this the case? Many reasons. But it is mainly due to the fact that there are so many illegal (or unregistered) oil palm plantations in Indonesia. And this is another concern for the methodology being used in this research.

Instead of looking at the actual existence of oil palm plantation in Indonesia (for example, by using satellite imagery), the research only looked at registered oil palm plantations in the form of plantation licenses (Izin Usaha Perkebunan) and land use rights (Hak Guna Usaha).

To demonstrate how flawed this is, I put together an interactive map that overlays oil palm plantations land cover with oil palm licenses in Riau. It is quite clear that registered oil palm plantations area are smaller in size compared to actual existing plantations.

You can access the interactive map here:


Closing note

To recap, from my brief observation, the research cannot establish that oil palm plantation is not the cause of deforestation in Indonesia.

  1. The small number of sample area is significantly inadequate, and with no explanation on sample selection method, it is not possible to generalize the research for the whole country;
  2. The method of using registered oil palm database instead of actual oil palm land cover database do not represent the actual development of oil palm in Indonesia;
  3. The use of narrow legalistic definition of deforestation is questionable. It intentionally left out millions of hectares of forest conversion outside of forest estate.

I will post an update as the full report of the research becomes available.