Trend Analysis and Other Data Analytics for Bamboo

Introduction

Bamboo is a non-wood forest product that is widely used in the Philippines primarily because
of its increasing potential uses and significant contribution to livelihood and economic development.
It is a group of woody perennial evergreen plants belonging to the family of grasses, Graminaceae or
Poaceae. Its size varies from small annuals to giant, solid, timber-like bamboos ranging from 10-40
meters in height. Morphologically, they can be characterized as having woody, usually hollow culms,
complex rhizome and branch systems, petiolate leaf blades and prominent sheathing organs.
Moreover, all the members possess similar anatomical features in the leaf blades (i.e. fusoid cells and
arm cells) which set the bamboo apart from grasses (Roxas, 2012).

Bamboo have the distinction of being the world’s tallest grasses. They grow naturally inside
forests though they also thrive well in lowland areas and along riverbanks. According to De Beer
(1996), bamboos are most abundant in seasonally dry, monsoon forests and favour disturbed areas,
while in primary evergreen rain forests they tend to be found primarily along water courses. The
importance of bamboo is derived from the utility of the long, slender, pliable stem, that is usually
hollow, which makes it suitable for a wide magnitude of uses ranging from simple to sophisticated
ones. These uses according to Razal and Palijon (2009) include: hunting tools (e.g., bow and arrow),
for farming and growing food such as trellises, fishing such as traps as well as for building houses
and fences, animal cages and poultry sheds, in making furniture, tools or cooking utensils, transport
vessels, water or liquid containers, toys, and musical instruments such as flutes and percussion
instruments. And just recently, the use of ethnic and native products has become a trending craze,
that resulted to an increase in the demand for bamboo-made products as evidenced by the resurgence
of bamboo as material for interior design of houses, accessories, and other home, office and
restaurant furnishing. Perhaps, the versatility of bamboo outmatches most of tree species
(PCAAARRD, 2008).

According to Rojo (1996) as cited by Rajal and Palijon, 2009, the Philippines has more than
12 genera of bamboos and about 62 species as compared to the 70 species previously reported by
Aggangang (2015). This increase may be attributed to the establishment of bambuseta in different
parts of the country that encouraged introduction of living specimens of exotic species. Moreover,
according to Virtucion (n.d.) and Aggangan (2015), only 9 are economically important species,
namely, Kawayan tinik (Bambusa blumeana), Bayog (B. merrilliana), Kawayan kiling (B. vulgaris),
Giant bamboo (Dendrocalamus asper), Bolo (Gigantochloa levis), Kayali (G. atter), Buho
(Schizostachyum lumampao), and Anos (S. lima).

The Philippine Bamboo Industry

Bamboo industry in the Philippines has grown incrementally because of the market
potentials of these species. According to PCAAARRD (2011), growing bamboo requires less
time, money and effort. In three to five years, the culms are already matured and can already be
harvested, depending on the intended uses. Besides to its rapid growth, bamboo is a versatile
plant that can be easily worked with and be transformed into various structural uses.

Pole production and processing are the two major components of the bamboo industry
(Angganan, 2015). Bamboo plantation is ready for pole production within a much shorter period
of time as compared to forest trees, thus, leads to bamboo being used as complementary source
of income and a practical economic activity. Harvesting for small bamboo culms can begin on
the fourth year after the plantation establishment while on the sixth to eight year of plantation,
harvesting with the use of culm selection is done wherein mature, over-mature and defective
culms were being selected (Razal and Palijon, 2009).

After harvesting, bamboo culms were usually prepared for treatment to protect bamboos
from decay and other forms of deterioration. Afterwards, the processing begins wherein bamboo
poles where being transformed into furniture, handicrafts, baskets, fishing materials, housing
construction and agricultural implements and many other uses (Razal and Palijon, 2009).
Engineered or laminated bamboos were usually used for housing and construction, as well as in
the furniture-making industry.

In 2010, the Philippine bamboo industry was formalized by virtue of Executive Order
879 that created the Philippine Bamboo Industry Development Council (PBIDC) to promote the
bamboo industry development project and direct the use of bamboo for at least twenty-five
(25%) percent of the desk and other furniture requirements of public elementary and secondary
schools as well as to prioritize the use of bamboo in furniture, fixtures, and other construction
requirements of government facilities.

Role of Bamboo in Livelihood, Economic and Environmental Development

The bamboo industry, through provision of markets to rural farmers and bamboo
producers, contributes significantly to livelihood and economic development. As mentioned
earlier, bamboo’s characteristics of being a fast-growing and versatile plant have introduced
various opportunities both to the producers and consumers. Production of bamboo became
financially profitable and economically desirable due to the fact that bamboo culms can be
harvested from three to four years only depending on intended use, while, merchantable harvest
from first plantation can be obtained only for six years and that inputs for maintenance is
relatively simple and inexpensive. Moreover, bamboo products were also in demand in the
market primarily because of its desirability as raw material for large-scale industrial application,
hence, considered as great substitute for wood.

Like any non-wood forest products, bamboo has an important contribution to household
and national economies, to food security, and to environmental objectives (which include the
ability to store and sequester carbon, and the conservation of biological diversity (Razal and
Palijon, 2009). The contribution to bamboo industry to household population relies on the fact
that many rural communities depend on it for subsistence, either source of food and income, thus
leading to the potential of bamboo for poverty reduction. Likewise, processing of bamboo does
not normally require high labor skill and qualification standard, thus, creating opportunities to
many rural communities and the poor.

In terms of contribution to the environment, bamboo can be grown and harvested without
depletion and deterioration of the soil.
 Overall, the research questions that the paper aims to answer are the status (trends) of
bamboo industry in terms of production, exports, imports and other performance indicators from
the period 1990-2016; determine the bamboo resources in the Philippines; and determine the
potentials of bamboo with regards to carbon sequestration and soil conservation. Other analysis
would have been included, such as the characteristics of bamboo producers and consumers, the
consumers’ preferences on bamboo products, producers’ motives for investing on bamboo and
the feasibility of bamboo production in the Philippines, however, lack of financial
resources/funding constraints the implementation of the mentioned studies.

Methodology

Data Sources and Collection

Secondary data were used to analyze the trend or status of bamboo resources in the
Philippines. The analysis is primarily based on the data provided by the Philippine Forestry
Statistics of DENR, which is the most comprehensive source providing annual data for national
forest inventories. Moreover, for growth curve fitting, the data utilized for analysis was the
average total height and basal culm diameter of four different bamboo species in Pampanga
plantation derived from Virtucio and Roxas (2003).
Data Analysis

Status of Bamboo Industry in the Philippines

Sources of Bamboo in the Philippines

The summarize data for bamboo production from the year 1990-2016 was used for GIS-
based mapping of the sources of bamboo. QGIS was use to produce maps for the analysis.
Probabilistic Projections of Bamboo Production

Using the analysis toolpak and the random variable function in Excel, the data for
bamboo production (yield) in the Philippines was simulated to forecast future trends in yield.
Three approaches in the Montecarlo model was used for probability projections: normal
distribution, triangular distribution, and uniform distribution. Parameters to be defined by the
curve (mean, standard deviation, minimum and maximum values) was set up in the respective
approaches.

Growth Curve Fitting of Four Bamboo Species

Three growth curve models of nonlinear least-squares estimation were utilized to fit
growth curves of average total height and culm diameter of four different bamboo species over
10-year period. The data for kawayan tinik (Bambusa blumeana), kawayan kiling (B. vulgaris),
bayog (Dendrocalamus merrillianus), and bolo (Gigantochloa levis) was retrieved from
Virtucion (2003). Analysis was done using STATA 13’s nl module and the estimated model
were compared by graphical and numerical analysis of the following factors: the coefficient of
determination, R2; the mean square error, root MSE; and the number of significant variables; to
determine the best fit model for the data.

The exponential growth curve used is in the form:

y=b0 +b1 ( b 2x )
wherein the curve was below the asymptote y=b0. This asymptote could indicate the maximum
height or basal diameter interpolated from the data.

Logistic and Gompertz growth curves are both sigmoid function with several varying
properties. Logistic growth curve used in this study is in the form:
b1
y=b0 + ,
1+ e−b ( x−b )
2 3

while for the Gompertz:

−b 2 ( x−b 3 )

y=b0 +b1 e−e .

Both have two asymptotes at y=b0 and y=b0 +b1 in which the maximum height or basal
diameter could be interpolated from the latter asymptote. The inflection point of the logistic
b b
( 2 ) ( e )
curve lies at b 3 , 1 + b0 , while b 3 , 1 + b0 for Gompertz. Though both are sigmoidal in shape,
the Gompertz curve is asymmetrical about the inflection point while logistic is symmetrical. The
slope of the tangent line about the inflection point of the curves was determined as the maximum
growth rate. This can be computed from the first derivative of the function and simplified as
b1 b2 b b
m= and m= 1 2 for logistic and Gompertz, respectively.
4 e

Selection for the best fit model for the culm height and diameter was based on three
factors: how well observations were replicated by the model, measured by the R 2; degree of
deviation between observed values and estimated values, measured by root-mean-squared error,
and; the number of significant variables based on their p-values. Model with the highest R 2,
lowest RSME, and the most number of significant variables was determined as the best fit model
for the data. Using the selected model, culm height and diameter will be estimated with the
harvest age for the four different bamboo species.
Modeling and Simulation

Result and Discussion

Status of Bamboo Industry in the Philippines

Bamboo, being considered as one of the most important non-wood forest products, have
become valuable and oftentimes substitute for wood in industrial application. It has been used by
many countries particularly in the Philippines as major construction materials for houses and is
used to produce mats, baskets, tools, handles, hats, traditional toys, trusses, and fences.
Meanwhile, it also contributes to food sector by utilizing the bamboo shoots as food most
especially in rural areas. Hence, bamboo’s potential for economic and environmental
development has brought its production to increase and even to participate in the international
trade.

For the past 25 years, Philippine bamboo industry has developed simultaneously with its
growing industrial and environmental importance. Figure X shows bamboo production in the
Philippines wherein over 22 million of bamboo culms were produced from the year 1990-2016
(Philippine Forestry Statistics). The figure below depicts the unstable trend of bamboo pole
production. In the first decade, there have been an increase of 1.6 million pole produced from
640,000 poles in 1990. This huge increase in the year 2000 has been the peak of bamboo
production throughout the years. Subsequently, it then declined in the following year by 74%.
The gradual and unsteady rate of bamboo produced continuous in the year 2010 wherein a total
of 929,000 poles where produced. From the year 2011 up to present, slight difference only from
the annual produced poles can be observed, and the latest record from PFS accounts to 883, 000
poles in the year 2016.
Moreover, a linear trend can be regressed from these data points though it has an erratic
behavior with no observable trend. This can be further explained by low R2 (insert here) and high
standard error (insert here) making it difficult to forecast future values.

philippine Bamboo producti on

2500

2000
quantity (in thousands)

1500

1000

500

0
0 91 3 94 95 6 97 8 99 0 1 02 3 04 5 6 07 8 09 0 1 2 3 14 5 6
99 9 99 9 9 99 9 99 9 00 00 0 00 0 00 00 0 00 0 01 01 01 01 0 01 01
1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

year

Figure 1. Bamboo Production in the Philippines (1990-2016)

Thus, time series analysis was utilized to fit models with low standard error and can be
used for forecasting. Time series analysis requires an ordered sequence of values of a variable at
equally spaced time intervals. Bamboo production from 1990-2016 was analyzed using different
methods of time series model fitting.

Moving average generates a “smooth” model by computing mean of successive smaller

sets of data points. From a starting point, “lag” values before the current observation or/and
“lead” values preceding the current observation will be included in calculating the set’s mean.
This process moves forward by one-time unit. Using STATA13’s tssmooth module, moving
average model was generated with 1 lag, current observation included, and 1 lead, notated as (1 1
1). Figure x shows how this moving average model “smoothen” the graph of bamboo production.
 Figure x.

International Trade of Bamboo Products

Globally, the Philippines is known as one of the main exporters of bamboo products.
According to the report of INBAR (2012), the country is the fifth most significant exporter of
bamboo (along with rattan) in the world. Bamboo products such as furniture, seats, poles,
torches, woven bamboo strips, mats, basketworks, wickerwork and other articles of bamboo were
being exported to Japan, Cyprus, France, Germany, United States, Singapore and other countries.

philippine Bamboo export

120

100
quantity (in thousands)

80

60

40

20

0
0 4 0 6 0 2
9 92 9 96 98 0 02 04 0 08 1 1 14 16
9 9 9 9 9 0 0 0 0 0 0 0 0 0
1 1 1 1 1 2 2 2 2 2 2 2 2 2

year

Figure 2. Philippine Bamboo Products Export (1990-2016)

Sources of Bamboo in the Philippines

Based on the report from FAO (2007), the country had 2.4% of forest area in 2005
covered by bamboo which are mostly sympodial species. These species are commonly found in
the public lands such as riverbanks and sloping areas. Figure x shows the diversity of bamboo
species in the Philippines. These accounted species were mostly coming from established
bamboo plantations, ERDB/DENR bambuseta and private bamboo gardens in the country.

Data from Philippine Forestry Statistics shows the regions that produces bamboo in the
Philippines from 1990-2016 (Figure x). A total of 18, 993, 602 bamboos (per piece) has been
produced by the whole country from a 26-year period. The top producing province is Pangasinan
with a total of 5,754,888 bamboo poles produced within the same period. Through bamboo, the
province of Pangasinan were able to boost their town’s economy. Bamboo were used in the
province in charcoal-making, furniture and nipa hats. The second largest to Pangasinan is the
province of South Catobato with a total of 3,155,772 poles. Other provinces that belongs to top
10 bamboo-producing province in the Philippines include Camarines Sur, Davao del Norte, La
Union, Quezon, Batangas, Tarlac, Davao del Sur and Abra (Figure X).

More so, through the National Greening of the government established by the virtue of
Executive Order No. 26 issued on 2011, the DENR committed to establish bamboo plantation.
From 2011-2015, the Philippines was able to establish a total of 16,442 hectares of bamboo
plantations in 15 regions. Among the regions, Central Luzon (Region 3) has the largest
established bamboo plantations under NGP with 4,765 hectares planted. Meanwhile, DENR has
also launched the Bamboo Plantation Development Project (BPDP) that aims to establish 256,
995.90 hectares of bamboo plantations for a period of six years (2017-2022).

Figure. Diversity of Bamboo Species in the Philippines

Figure x. Source of Bamboo in the Philippiens
 Figure x. Top 10 Bamboo-Producing Provinces in the Philippines

Probability Projection of Bamboo Production

Monte Carlo simulation was used to assess the future performance of bamboo production
in the Philippines. Assuming the distribution to be normal, parameters such as mean and standard
deviation was used to guide the generation of the stochastic yield. More so, for triangular and
uniform distribution, the parameters used were min, most likely (mean), max; and min, max;
respectively. Figure x, x, and x shows the result of the Montecarlo simulation for normal
distribution, triangular distribution, and uniform distribution.
 Normal Distribution
1.20

1.00

0.80
Prob
0.60

0.40

0.20

0.00
-1500 -1000 -500 0 500 1000 1500 2000 2500

Triangular Distribution
1.20

1.00

0.80
Prob
0.60

0.40

0.20

0.00
0 500 1000 1500 2000 2500
 Uniform Distribution
1.20

1.00

0.80
Prob
0.60

0.40

0.20

0.00
0 500 1000 1500 2000 2500

Growth Curve Fitting of Four Bamboo Species

Growth curves for average height and basal diameter from four different bamboo species
were modeled using nl module of STATA13 and the coefficients for each model were shown in
Table x and x for average total height and basal culm diameter, respectively.

Table x. Coefficients for Average Total Height of Four Bamboo Species

Bamboo Exponential Logistic Gompertz
Species b0 b1 b2 b0 b1 b2 b3 b0 b1 b2 b3
-
B. 33.97 35.12 0.920 3.835 13.12 1.605 4.969 4.159 12.93 1.170 4.648
blumeana 293 11 294 6 233 497 085 382 843 237 814
-
546.8 546.1 0.997 3.911 11.56 1.033 6.037 4.123 12.00 0.638 5.531
B. vulgaris 732 99 178 773 274 986 255 739 341 476 641
D. -
merrillianu 44.64 44.12 0.962 3.431 9.758 1.323 5.112 3.737 9.678 0.933 4.771
s 184 53 389 212 707 3 224 178 175 412 119
-
36.35 34.69 0.958 4.080 8.964 0.957 5.378 4.348 9.093 0.616 4.859
G. levis 178 24 489 958 924 439 052 897 551 679 969

Table x. Coefficients for Basal Culm Diameter of Four Bamboo Species

Bamboo Exponential Logistic Gompertz
Species b0 b1 b2 b0 b1 b2 b3 b0 b1 b2 b3
-
B. 43.67 44.02 0.944 3.957 12.94 1.714 4.988 4.159 12.93 1.170 4.648
blumeana 947 61 453 339 62 968 028 382 843 237 814
-
527.5 526.8 0.997 3.831 11.65 0.992 6.006 4.123 12.00 0.638 5.531
B. vulgaris 128 19 081 691 908 75 414 739 341 476 641
 D. -
merrillianu 52.61 51.84 0.969 3.530 9.679 1.364 5.171 3.737 9.678 0.933 4.771
s 161 88 597 514 571 062 091 178 175 412 119
-
146.4 144.2 0.991 4.000 9.070 0.912 5.347 4.348 9.093 0.616 4.859
G. levis 051 57 569 021 439 062 26 897 552 679 969

Observed and model values were tabulated in Table and graphically shown in Figure x.
From the graphs, it can be deduced through eyeball estimation that logistic and Gompertz curve
could be the best fit for the observations.

Analyzing the models’ R2, RMSE, and number of significant variables, it was determined
that for the height of Tinik, Kiling, and bolo, Logistic was the best fit curve, while Gompertz
curve for Bayog. For the basal diameter, it was determined to be Gompertz curve for all four
species.
Table. Summary Statistics for Average Total Height of Four Bamboo Species
R2 Root MSE P-value
Bamboo species Exponentia Logisti Exponentia Logisti Gompert Exponential Logistic Gompertz
Gompertz
l c l c z b0 b1 b2 b0 b1 b2 b3 b0 b1 b2 b3
0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0
Kawayan Tinik 0.92 0.99 0.99 1.92 0.55 0.70 1 7 0.00 0 0 0 0.00 0 0 0 0.00
Kawayan 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Kiling 0.93 0.99 0.98 1.34 0.60 0.72 - 0 0.00 0 0 0 0.00 0 0 1 0.00
0.6 0.6 0.0 0.0 0.0 0.0 0.0 0.0
Bayog 0.92 0.97 0.97 1.37 0.87 0.86 0 0 0.00 0 0 2 0.00 0 0 2 0.00
0.4 0.5 0.0 0.0 0.0 0.0 0.0 0.0
Bolo 0.94 0.98 0.97 1.03 0.64 0.75 9 0 0.00 0 0 1 0.00 0 0 2 0.00

Table. Summary Statistics for Basal Culm Diameter of Four Bamboo Species
R2 Root MSE P-value
Bamboo species Exponentia Logisti Gompert Exponentia Logisti Exponential Logistic Gompertz
Gompertz
l c z l c b0 b1 b2 b0 b1 b2 b3 b0 b1 b2 b3
0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Kawayan Tinik 0.90 1.00 1.00 2.17 0.27 0.00 0.48 6 0 0 0 0 0 0 0.00 0 0.00
Kawayan 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Kiling 0.94 1.00 1.00 1.23 0.22 0.00 - 0 0 0 0 0 0 0 0.00 0 0.00
0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Bayog 0.92 1.00 1.00 1.39 0.20 0.00 0.69 9 0 0 0 0 0 0 0.00 0 0.00
0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Bolo 0.96 1.00 1.00 0.85 0.16 0.00 0.89 9 0 0 0 0 0 0 0.00 0 0.00
 Bambusa
Bambusablumeana
Bambusablumeana
vulgaris
12.00
12.00
10.00
10.00
(cm)
(cm)

8.00
8.00
8.00
Height (m)
Height (m)

6.00
Diameter

6.00
6.00
Diameter

4.00
4.00
4.00
2.00
2.00
2.00
0.00
0.00 0 2 4 6 8 10 12
0.00 0 2 4 6 8 10 12
0 2 4 Year6 8 10 12
Year
Year
Basal culm diameter Exponential
Basal
Ave.culm
Logistic totaldiameter
height Exponential
Exponential
Gompertz
Logistic
Logistic Gompertz
Gompertz
Figure 3
Figure 4a

Figure 5 Figure 6
 Dendrocalamus
Gigantochloa
merrillianus
levis Dendrocalamus merrillianus
12.00 12.00
10.00 10.00
(cm)

8.00 8.00
Height(cm)
(m)

Height (m)
6.00 6.00
Diamater
Height

4.00 4.00
2.00 2.00
0.00 0.00
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Year Year

Basal
Ave.culm
totaldiameter
height Exponential
Exponential Ave. total height Exponential
Logistic
Logistic Gompertz
Gompertz Logistic Gompertz
 Based on the selected graphs several characteristics such as maximum height and
maximum growth rate was derived from the model and tabulated in Table x-x. Selected models
were also graphed with their properties shown such as asymptotes, inflection point, and tangent
line about the inflection point.

Maximum height Max, growth rate

Species Model
(m) (m/yr.)
Bambusa blumeana Logistic 16.95793 5.26697
Bambusa vulgaris Logistic 15.47451 2.98893
Dendrocalamus merrillianus Gompertz 13.41535 3.32332
Gigantochloa levis Logistic 13.04588 2.14584

Maximum height Max, growth rate

Species Model
(cm) (cm/yr.)
Bambusa blumeana Gompertz 17.09781 5.57007
Bambusa vulgaris Gompertz 16.12715 2.81939
Dendrocalamus merrillianus Gompertz 13.41535 3.32332
Gigantochloa levis Gompertz 13.44245 2.06300
 Dendrocalamus
Bambusa
Gigantochloa
Bambusablumeana
vulgaris
merrillianus
levis
18
16
14
12
Height (m)

10
8
6
4
2
0
1 2 3 4 5 6 7 8 9 10
Year