Field Tests of Carbon Monitoring Methods in Forestry Projects

Field Tests of Carbon Monitoring Methods in Forestry
Projects

Field Test of Carbon Monitoring Methods for Home Gardens in Indonesia¹
Matt Delaney, Program Associate and Jim Roshetko, Program Officer
Winrock International

Introduction
The purpose of this study was to field test methods for monitoring carbon storage in agroforestry
systems in Indonesia. The specific objectives of the study were:
1. Measure the carbon in above-ground biomass, soils, litter, and herbaceous vegetation of home
garden systems in Indonesia.
2. Evaluate the carbon monitoring methods and make recommendations for their improvement.
Study Area
The field test was conducted in North Lampung Province (4% 26.453’ S, 104% 54.768’ E), Sumatra,
Indonesia. The study site is called the North Lampung benchmark area (NLBA) and covers a 42,500
ha area along the lower reaches of the Tulang Bawang river (van Noordwijk et al., 1996). Soils are
well drained, deep, acidic, and of low fertility. Aluminum toxicity is common, especially below a
depth of 15 cm. Root development is possible down to a plinthic layer at 1-1.5 m. The major soil
groups are Oxisols/Ultisols, Inceptisols, and Entisols, covering 64, 29, and 7% of the area,
respectively. Annual rainfall averages about 2,200 mm, with 1-4 months receiving less than 50 mm
of rainfall (van Noordwijk, 1995).
The population in the NLBA has increased dramatically since the early 1980s, when an Indonesian
government program encouraged transmigration from populated areas in Java into North Lampung.
The population can be divided into three main groups. The first is the original inhabitants, who own
fertile lands close to the main rivers that run through the area. The second consists of
transmigrants, who inhabit upland infertile lands. The third group, spontaneous migrants, have
arrived in the area in the last several years, and inhabit any other lands that are available (van
Noordwijk 1995).
Land Use in the North Lampung Benchmark Area
The major land-use types found in the NLBA range from wet rice cultivation to secondary forest.
Home garden systems were chosen for sampling in the NLBA because they form a common land-
use practice in this part of Indonesia. Little information exists about these systems, but they are
believed to offer significantly more carbon benefits than other types of land uses (e.g., agriculture,
grasslands). Home gardens are small household owned plots of land (usually 0.25 ha) that were
converted from secondary forest and planted with food crops such as cassava, maize, and rice.
Spices, medicinal plants, coconuts, and lemon trees are also cultivated (Gintings et al., 1996). A
wide range of tree and plant species are found in home gardens (Table 1). The difference between
a home garden and traditional agroforest systems is not well defined; generally, home gardens tend

---

to be small, close to a farmer’s home, and produce crops usually for family consumption.
Table 1. Species found in 19 home gardens in Lampung, Indonesia
Botanical name
Local name
Acacia auriculiformis
akasia
Acacia mangium
mangium
Aleurites moluccana
candlenut / kemiri
Alstonia spp.
pulai
Anacardium accidentale
cashew / jambu mete
Annona muricata
soursop / sirsak
Anona reticulate
sweet sop / buah nona / jambu nona
Archidendron pauciflorum
jengkol
Arenga pinnata
kolang kaling
Artocarpus heterophyllus
nangka
Artocarpus integer
cempedak
Averrhoa bilimbi
belimbing
Ceiba pentandra
kapok
Cinnamonum parthenoxylon
kayu lada
Cocos nucifera
coconut
Coffea canephora var robusta
kopi / coffee
Erythrina spp.
dadap
Flacourtia rukam
rukam
Gliricidia sepium
gamal
Gnetum gnemon
melinjo
Hevea brasiliensis
karet
Hibiscus tiliaceus
waru
Leucaena leucocephala
lamtoro
Mangifera foetida
pakel
Mangifera indica
mango
Mangifera odorata
kurai / kuweni
Melia azedarach
mindi
Musa spp.
banana / pisang
Nephelium lappaceum
rambutan
Paraserianthes falcataria
sengon
Parkia speciosa
petai
Peltophorum plerocarpum
petean
Peronema canescens
sungkai, sekai
Persea americana
alpokat
Psidium guajava
guava / jambu biji
Schima wallichii
puspa
Spondias spp
kedongdong
Syzygium aceum
jambu air
Tamarindus indica
asam
Tectona grandis
teak / jati
Terminalia citrina
jaling
Theobroma cacao
cacao
Project site description
Local name of site: Villages of Negara Jaya, Tegal Mukti, and Tiuh Baru
Elevation: 100 m
Ecological zone or general site type: Tropical humid zone (Holdridge)
Most common slope class: Gentle (0-5o)
Mean annual rainfall: 2,500 mm
Rainfall regime: Bimodal (5-6 months >200mm, 2-4 dry months ?100mm)

---

Maximum length of dry season: 4 months (?50 mm)
Mean annual temperature oC: 28
Surfacesoil texture (0-5 cm): 61.7% sand, 24.9% silt, 13.4% clay
Sub-soiltexture (5-15 cm): 71.1% sandy, 6.19% silt, 22.7% clay
Soil depth to impermeable layer: >100 cm
Surface soil pH (0-5 cm): 3.74 (KCL), 4.25 (H2O)
Sub-soil pH (5-15 cm): 3.50 (KCL), 4.25 (H2O)

Methods
Criteria for selecting farms
Field work took place February 23-28, 1998. Pre-planning involved talking with local officials, and
researchers with the International Center for Research in Agroforestry (ICRAF). No maps on soil
type or land use existed for home gardens in the area, so no stratification was possible. Instead,
farms were selected for measurement only if: (1) the farmer gave permission, and (2) the home
gardens had a predominant tree cover of species intended for primarily home use. Thus, home
gardens that contained primarily agriculture crops (corn, cassava, vegetables, etc.) or one market-
oriented tree crop (coffee, cacao, coconut, Paraserianthes, etc.) were excluded. Home gardens
that contained rice paddies or fish ponds were also excluded. Even with these limiting criteria, the
species composition and tree density of the home gardens measured varied greatly from farm to
farm. Understory vegetation also varied greatly, from forest-like conditions (containing a lot of
natural regeneration and saplings) to areas where the understory was barren.
Plot establishment and measurement
At each site, the perimeter of the home garden was first measured with a 100-m measuring tape. A
sketch was made of the plot, and uncorrected GPS coordinates were taken in the home garden with
a Garmin 12XL GPS receiver. The methods described in A Guide to Monitoring Carbon Storage in
Forestry and Agroforestry Projects recommend that four subplots be placed in each farm. However,
since the average home garden was 0.18 ha in area, and they were usually 25 m in width and 75 m
in length, it was impossible to locate the plots as suggested in the protocol. We determined it was
best to establish two subplots per farm instead of four. Subplots were laid out perpendicular to the
longest borders, along a line bisecting the reference point (center point) of the home garden
system. If the farm was irregularly shaped, a third plot was established.
From the subplot center, the crew supervisor established four points (N, E, S, and W) 7.9 m away,
or 1 m inside the subplot boundary. At these four points herbaceous material (? 5 cm diameter),
litter (? 5 cm diameter), and soil (to 30 cm depth) were sampled. The herb and litter samples were
collected by clipping all material within a circular aluminum sample ring (0.28 m2). However, if a
juvenile tree or agriculture plant fell within the sampling ring, it was not clipped, because in this
preliminary study, we wanted to minimize any impacts on the farmers’ crops. Soils were sampled
within the aluminum ring after all herb and litter material were collected. In one of the four subplots,
a soil bulk density sample was taken at a depth of 15 cm with an aluminum cylinder. For all trees in
each subplot with a diameter at breast height (dbh) > 5 cm, the diameter was measured and, when
possible, the species name recorded. In cases of trees without a diameter-biomass relationship
(e.g., coconut, banana), heights were measured using a clinometer.
In each farm, at least 2 subplots were established for a total of at least 8 herb, litter, and soil
samples. Each of the herb and litter samples were weighed using a spring scale. All the herb
samples were mixed and a subsample taken for moisture content determination. The same was
done for litter. Soils were sieved through a 5-mm mesh screen, mixed to a uniform color and
consistency, and then a subsample taken for carbon analysis. Walkley-Black analyses for soil
organic carbon were done at a local university. A total of 10 bulk density samples were taken at 15-

---

cm depth. Each sample was dried in an oven at 100o C.
Estimating aboveground biomass
In order to estimate biomass of aboveground vegetation, a general biomass equation was used.
When the general equation was not suitable for the types of vegetation encountered in the plots, an
alternative equation was used, from a study conducted in Puerto Rico on the palm Prestoea
montana (Frangi and Lugo, 1985).
Table 2. Equations used for calculating aboveground biomass in home garden systems.
Species
Equation
Source
R2
General
y=exp{-2.134 + 2.530 * ln(D)
Brown, 1997
.97
Palms
y=4.5 + 7.7 * H
Frangi and Lugo, 1985
.90
where exp {…}means “raised to the power of {…}”
ln means “natural log of (…)”
y = above-ground biomass in kg
H = height in m
D = diameter at breast height (1.3 m)

Results
Carbon Inventory
A total of 19 plots were measured, and the amount of carbon in the aboveground, soils, litter, and
herbaceous material ranged from 10 to 126 t C/ha (Table 3).
Table 3. Carbon stocks by component for home garden systems in the North Lampung benchmark
area, Indonesia
Herbaceous
Plot
Aboveground biomass
Litter
Soil
Roots
Total






(t C/ha)
material

(t C/ha)

(t C/ha) (t C/ha) (t C/ha)
(t C/ha)




HGS 1
61.9
2.4
0.4
65.7
9.3
140
HGS 2
6.3
0.3
0.2
55.3
0.9
63
HGS 3
34.1
2.7
0.1
10.4
5.1
53
HGS 4
17.6
3.2
0.2
52.1
2.6
76
HGS 5
14.4
2.5
0.1
44.9
2.2
64
HGS 6
21.6
1.9
0.3
69.3
3.2
96
HGS 7
23.0
0.2
0.2
41.2
3.5
68
HGS 8
34.8
2.3
0.1
72.5
5.2
115
HGS 9
24.9
1.6
0.3
40.3
3.7
71
HGS 10
22.6
1.2
0.5
72.5
3.4
100
HGS 11
65.3
1.7
0.3
55.3
9.8
132
HGS 12
17.1
2.7
0.4
65.7
2.6
89
HGS 13
21.6
2.8
0.5
51.6
3.2
80
HGS 14
45.4
3.6
0.2
77.5
6.8
134
HGS 15
84.0
0.0
0.0
69.3
12.6
166
HGS 16
56.2
1.4
0.8
76.1
8.4
143
HGS 17
46.3
0.8
0.3
103.7
6.9
158
HGS 18
53.8
4.0
0.1
69.3
8.1
135
HGS 19
19.3
2.9
1.4
62.1
2.9
89
Mean (SD)
35.3 (21.0)
2.0 (1.2) 0.3 (0.07) 60.8 (4.4) 5.3 (0.7) 104 (8.1)
CV
60%
57%
95%
32%
60%
34%

---

Carbon storage data from the 19 home gardens were compared with other data available from
ICRAF (Rosalina et al., 1997), which are included in Table 4.
Table 4. Carbon stocks of different land-use systems in the Northern Lampung bench-
mark area, Indonesia. Results for home garden systems are based on 19 plots sampled.
Land-use type
t C/ha
Primary forest
325
Secondary forest
177
Imperata grassland
3*
Home garden systems
104
*does not include soil carbon
Time required for inventory
The crew took two days to learn the inventory methods. During that period the crew inventoried two
home garden systems per day. Field time per home garden averaged 2 hours, 30 minutes on the
first day and 1 hour, 40 minutes on the second day. After Day Three, inventory time per home
garden system was approximately 1 hour, 15 minutes. ANOVA analysis compared the amount of
aboveground biomass with the amount of time required for sampling, and found no significant
relationship between the two (p > 0.05). The same was true for herb, litter, roots and soil.
There were usually seven people in the crew actively participating in the inventory. This could
probably be reduced to six, with three collecting tree data and three collecting soil/herb/litter data.
Usually one person accompanying the crew knew the community, and thus knew where to look for
home gardens that fit the selection criteria. This person would engage the local residents in
conversation, answer their questions, and reassure farmers that the team would not damage their
crops.
Discussion
Recommendations for refining the methodology
Where agroforestry farms are not a standard size and shape, it is advisable to adjust the monitoring
protocol. This should be done before field work begins or at the first farm site. The home gardens
measured consisted of two basic shapes — rectangular and square — and one standard size, 0.25
ha. Actual size of the home garden systems varied because most farmers had built a house on the
quarter-hectare site, and home gardens lay outside the house area.
A shortcoming of the methods is that the sub-plots as prescribed do not contain species planted
along the borders of agroforestry systems. Observations in North Lampung indicated that bamboo
was a component of approximately one-third of the home gardens inventoried. Yet bamboo was not
sampled in the sub-plots because it was most often planted in the corners or along the back borders
of home gardens.
Planning for carbon inventory work
Apart from time actually spent in the field, the work involves a number of activities that need to be
considered when planning inventories.
These include:
1. Meeting with local village leaders, who in Indonesia are government representatives and
responsible for foreigners in their jurisdiction. It is important to make a social call to explain the
intended work. Visits usually last about 15 minutes, but under some circumstances may require an
hour (travel time, finding the leader, waiting for a meeting).

---

2. Talking with farmers. This is important to put farmers at ease about the inventory work.
3. Evaluating and selecting sites. This often required only a few minutes (5-10). On a few
occasions, though, finding home gardens that fit the selection criteria took up to an hour. This
process can be greatly accelerated by having a crew member who is familiar with the area and by
requesting assistance from village leaders and local farmers, who are often glad to help.
4. Training in methods. As mentioned previously, a crew generally requires approximately two
days to learn the inventory methods and develop a comfortable pace.
5. Establishing of markers, if permanent plots are to be established on small-holder farms. This
could also require time for completing Farmer Interview Forms.
Conclusions
The information compiled on home gardens in Lampung not only helped with the carbon inventory,
but added to the understanding of species composition in home gardens. The carbon-monitoring
methods were generally well suited for sampling home gardens in Indonesia. The modification to
the methods most commonly needed was the use of two subplots per farm instead of four, due to
the home gardens’ small size. The carbon content of the systems were higher than that of Imperata
grasslands, and demonstrates the importance of home gardens to the overall carbon budget of the
region. The field test results also showed the importance of pre-planning, especially the need to
meet with local officials and keep the community informed about field work activities.

References
Brown, S. 1997. Estimating biomass and biomass change of tropical forests: A primer. FAO
Forestry Paper 134. FAO, Rome.
Frangi, J. L. and A. E. Lugo. 1985. Ecosystem dynamics of a subtropical floodplain forest.
Ecological Monographs 55:351-369.
Gintings, A. N. , C. Anwar, I. Samsudin, M. E. Siregar, B. M. Punama, and Kasirin. 1996.
Agroforestry characterization in Pakuan Ratu and Tulang Bawang Tengah, North Lampung District,
Lampung. Pp 59-68 in Proceedings of a workshop Alternatives to Slash-and-Burn Research in
Indonesia, eds. M. Noordwijk, T. Tomich, D. Garrity, and A. Fagi. ASB-Indonesia Report Number 6.
Bogor, Indonesia.
Hairiah, K. 1997. Final report: Carbon stock in various land-use systems in Lampung and Jambi.
ICRAF Indonesia. Pp 2.
Leaving, P and H. de Foresta. 1991. Economic plants of Indonesia: a Latin, Indonesian, French
and English dictionary of 728 species. ORSTROM, SEAMEO, and BIOTROP. Bogor, Indonesia.
Noordwijk, M., Betha Lusiana, Suyanto and Thomas P. Tomich. 1996. Soil and other constraints
to agricultural production with or without trees in the north Lampung benchmark area of the
Alternatives to Slash and Burn project. Agrivita 19:4 136-145.
Noordwijk, M. and P. Purnomosidhi. 1995. Root architecture in relation to tree-soil-crop
interactions and shoot pruning in agroforestry. Agroforestry Systems 30:161-173.
Rosalina, U., Setiabudhi, and A. E. Putra. 1997. Vegetation analysis and database management
system in Lampung and Jambi. International Center for Research in Agroforestry, Southeast Asia
Regional Office.

---

Winrock International
38 Winrock Drive
Morrilton, AR 72110-9370 USA
Tel: 501-727-5435
Fax: 501-727-5242
Email: information@winrock.org
Internet: www.winrock.org
Citation:
¹Delaney, M. and J. Roshetko. 1999. Field Test of Carbon Monitoring Methods for Home Gardens in
Indonesia. In: Field Tests of Carbon Monitoring Methods in Forestry Projects. Forest Carbon
Monitoring Program, Winrock International, Arlington, VA, USA. pp. 45-51.

---