FONT SIZE
S
M
L
COLOR
Normal
1
2
3

FRMO Top

Home > International Partnerships > Collaborative Research > Developing a Model to Estimate Biodiversity in Tropical Rain Forests from Satellite Images

Update:September 8, 2020

Main content starts here.

Developing a Model to Estimate Biodiversity in Tropical Rain Forests from Satellite Images

photo01:photos on this research theme

1. Partners

Research Center for Biology, the Indonesian Institute of Sciences (RCB-LIPI), Indonesia

East Kutai Agricultural College School, Indonesia

 

2. Research Period

FY 2014-2017 Grant-in-Aid for Scientific Research (KAKENHI)

3.Lead Researcher

UEDA, Akira

4.Background

The program REDD+ aims to prevent forest decline and associated climate change, considering the importance of both local livelihoods and biodiversity conservation as primary impacts of this goal. Indeed, it is becoming increasingly difficult to conserve biodiversity in fragmented tropical rain forests after deforestation. Because of vast areas of the fragmented and scattered forests, and of difficulties to access, it is difficult to research all areas to manage forests for ecosystem restoration.
On the other hand, recent developments of remote sensing technologies, including satellite images, have been very fast. By developing these methods to evaluate biodiversity from satellite images, it becomes possible to manage vast forest areas efficiently without field investigation. This study aims to develop a method to evaluate biodiversity in tropical rain forests from satellite images.

5.Research Goal

We approached our research objective through the following steps; 1) selecting transects that begin from large intact natural forests and pass through degraded forests with various levels in the lowlands of East Kalimantan, Indonesia, using satellite images and to collect data of vegetation and insect diversity; 2) to describe generalized linear models (GLM) of the relationships between the site index assigned from a satellite image take into account its level of degradation, landscape type, and the diversity data of trees and insects; and 3) to verify what we can estimate from satellite images.

6.Research Strategy

1) Study sites: We carried out studies 10 to 30 km north of Balikpapan in the lowlands of East Kalimantan, Indonesia, including a large intact forest of the Sungai Wain Protection Forest (SWPF).

2) Tree communities: The lowland tropical rainforest in Borneo has one of the richest plant faunas in the humid tropics. This species richness is supported by the complexity of the forest structure. For diversity of trees we selected 0.04 ha (a circle with a radius of 11.3 m) plots with a certain distance on the transects. We measured the diameter of tree trunks at ca. 1.2 m high (diameter of breast height: DBH) of which DBH are larger than 10 cm, and identified the species. We also measured height of the highest tree in each plot using a vertex.

3) Dung beetles: Dung beetles (a coprophagous group of Scabaeoidea) are known to be a reliable indicator of habitat quality and environmental change in tropical rainforests and their surrounding environments. To estimate the diversity of dung beetles, we set pitfall traps baited with raw fish meat and human feces at certain distances along the transects and collected beetles trapped for 5 days.

4) Parasitic wasps: Parasitic wasps, as natural predators of various moths, beetles, flies and bugs, are major agents in regulating forest insect populations. As such, the diversity of parasitic wasps is linked with the maintenance of stable ecosystems within tropical rainforests. We collected parasitic wasps using net sweeping understory vegetation, by walking at every sampling plot along the transect line. A total of 30 sweeps was treated as one collection and we did 12 collections at every plot.

5) Analysis of satellite images and modeling to estimate diversities: We searched satellite images without clouds taken by the satellites Landsat8 and Spot5. We collected data on reflection rates of the observation bands of the satellites and calculated several indices from some reflection rates such as: NDVI (Normalized Difference Vegetation Index), GNDVI (Green NDVI), the value on axis 1 of Tasseled Cap Transformation (TC) (the index corresponding brightness), the value on axis 2 of TC (the index corresponding volume of green leaves), and the value on axis 3 of TC (the index corresponding humidity). We analyzed the relationships between our diversity data and both reflection rates and indices, then ran a GLM to estimate the statistical significance and effect size of the relationships. Finally, we selected models estimating the diversities from satellite images.

 

7.Scientific Achievement

1) Tree communities: We examined tree height, tree thickness, and rate of pioneer tree species as indices of forest degradation and found clear relationships between the highest tree height and the value on axis 1 of TC, the largest tree thickness and the reflection rate of green (wave length 525-600 nm), and between the basal area rate of pioneer tree species and the reflection rate of short wavelength infrared region (2100-2300 nm). These results indicate that we were able to estimate the maturity levels of forests from satellite images.

2) Insect diversity: Most dung beetle species were largely collected in the intact natural forest, and were endemic to Borneo or Sundaland (Peninsular Malaysia, Sumatra, Java, and Borneo). We found a clear relationship between richness of forest species and the value on axis 1 of TC. For the parasitic wasps, we found a clear relationship between species richness and NDVI. These results indicate that we were able to estimate the diversities of forest dung beetles and braconid wasps from satellite images.

 

8.Applications

In this study, we began to develop models estimating the maturity level of forests and diversities of dung beetles and parasitic wasps in tropical rain forests, using satellite images as an example of modeling methods of biodiversity conservation contributing REDD+. Future research will need to investigate species diversity of various organisms within a larger area, such as the whole of East Kalimantan, and to increase the accuracy of the models. Based on our results, it is possible to estimate biodiversity from satellite images without field data in tropical forests in the world, we can manage the fragmented and scattered tropical rain forests aiming restoration of ecosystem efficiently.

9.Publications

Ueda A, Dwibadra D, Noerdjito WA, Kon M, Fukuyama K (2015) Comparison of baits and types of pitfall traps for capturing dung and carrion scarabaeoid beetles in East Kalimantan. Bull For Forest Prod Res Inst 14: 15-28

Ueda A, Dwibadra D, Noerdjito WA, Sugiarto, Kon M, Ochi T, Takahashi M, Fukuyama K (2015) Effect of habitat transformation from grassland to Acacia mangium plantation on dung beetle assemblage in East Kalimantan, Indonesia. J Insect Conserv 19: 765-780

Ueda A, Dwibadra D, Noerdjito WA, Sugiarto, Kon M, Ochi T, Takahashi M, Igarashi T, Fukuyama K (2015) Buffer zones for placing baited traps in grasslands bordering forests and availability of riparian reserves of trees in grasslands: A preliminary study for dung beetle assemblages in East Kalimantan, Indonesia. Bull For Forest Prod Res Inst 14: 125-134

Ueda A, Dwibadra D, Noerdjito WA, Sugiarto, Kon M, Ochi T, Takahashi M, Igarashi T, Fukuyama K (2015) Effects of distance from devastated forests and topography on dung beetle assemblages in burned forests of East Kalimantan, Indonesia. Bull For Forest Prod Res Inst 14: 135-144

Ueda A, Dwibadra D, Noerdjito WA, Sugiarto, Kon M, Ochi T, Takahashi M, Fukuyama K (2017) List of dung beetles (Coleoptera : Coprophagous group of Scarabaeoidea) collected in lowland near Balikpapan, East Kalimantan, Indonesia. Bull For Forest Prod Res Inst 16: 109-119

Ochi T, Kon M, Ueda A (2018) Notes on the Coprophagous Scarab-beetles (Coleoptera, Scarabaeidae) from Southeast Asia XXX. A new Onthophagus species from Borneo. Elytra, Tokyo, New Series 8: 369-372