Beyond the protected area network – Identifying priority areas for conserving biodiversity

Hi everyone, I’m Clarice, a Y2 undergraduate that recently joined the TEE Lab in January! Over the the past six months, I’ve been working on my project for the CY2001 module – identifying conservation areas for biodiversity.

In collaboration with BirdLife International and Nature Society Singapore (NSS), I’ve been systematically sorting through two decades worth of data of bird and butterfly records in Singapore across 78 different sites. This was an overwhelming task as there were many sources I drew from (citizen science data, NSS data etc.), with a grand total of over 600 000 bird records and over 30 000 butterfly records!

Figure 1. A matrix recording bird sightings in each site over the past two decades. There were 300 columns of birds, 355 columns of butterflies, and 78 rows of sites!

After that, I used the prioritizr package and Gurobi solver on R to build and solve conservation problems! Using integer linear programming (ILP), sites are iteratively added to the set in a stepwise manner based on how complementary the sites are in community composition until relative target sets are met. I used a minimum cost objective to achieve the lowest cost of combination of sites to conserve our native biodiversity. Here’s some of the very cool results I achieved:

Figure 2. Sites chosen for conservation of 80% of bird occurrences in Singapore, with a total land area of 111.4km2 needed.

To conserve 80% of bird occurrences in Singapore, we need to protect a minimum of 111.4km2 of green spaces out of Singapore’s total land area of 728.3km2. That is a considerable large land area we need to protect! Looking at butterflies, a significant area of 114.9km2 was also calculated which was unexpected as I thought that butterflies would be more concentrated in certain green spaces due to their small body size and more specific vegetation requirements (e.g., larval host plant availability).

Figure 3. Sites chosen for conservation of 80% of butterfly occurrences in Singapore, with a total land area of 114.9km2.

But here’s an important question – can we conserve birds and butterflies with the same set of sites? Previous studies have found an extremely low coincidence in hotspots of these two taxa (12% overlap). Fortunately, I found that in Singapore, there is a significant overlap of sites which were chosen for the conservation of each taxa!

Figure 4: There are 42 common sites (in red) with high conservation value for both taxa

That’s amazing news – we can conserve both taxa with the same sites which allows for greater efficiency which is crucial for land-scarce Singapore.

Finally, here’s the list of sites which I propose to be prioritised for conservation based on my results:

Unprotected sites that should be prioritised:

Asrama Forest, Bukit Brown, Clementi Forest, Coney Island, Tengah and Ubin Central

Sites within the parks and nature areas network that should be further prioritised:

Bukit Batok Nature Park, Chestnut Nature Park, Chinese Garden, Hampstead Wetlands Park, Kranji Marsh, Mount Faber, Pasir Ris Park, Rifle Range and Telok Blangah Hill Park

Sites that should be protected for their unique species (which cannot be found elsewhere in Singapore):

Ubin East, Western Water Catchment, Zoo (Mandai)

The selection of unprotected sites demonstrates that the current protected area network in Singapore is insufficient for the conservation of both taxa. There is still a lot to be done to achieve conservation goals, and I will be continuing work on this for the next year to hopefully inform future conservation planning in Singapore. I am excited to return with updates in a few months’ time!*

*This analysis was first presented for Clarice's research module (CY2001) in April 2021 and is still a work in progress, with results to be refined