Por:
Oscar Geovani; Kant Martínez Cortés
|
Fecha:
2022
Forest Policy and Economics 138 (2022) 102722
1389-9341/© 2022 Elsevier B.V. All rights reserved.
An analysis of wood availability under six policy scenarios of commercial
forest plantations in Colombia
´Oscar Geovani Martínez-Cort´es a, *, Shashi Kant b, Henrieta Isufllari c
a Graduate Department of Forestry, John H. Daniels Faculty of Architecture, Landscape and Design, University of Toronto, 33 Willcocks Street, Toronto, ON M5S 3B3,
Canada
b Institute for Management & Innovation, and, Graduate Department of Forestry, John H. Daniels Faculty of Architecture, Landscape and Design, University of Toronto,
33 Willcocks Street, Toronto, ON M5S 3B3, Canada
c Comisi´on Nacional de Evaluaci´on y Productividad, Amun´ategui 232, oficina 401, Santiago, Chile
A R T I C L E I N F O
Keywords:
Forest policy
Wood supply: Growth and yield models
Colombia
Commercial plantations
Policy analysis
A B S T R A C T
An empirical analysis which compares Colombia’s 2019 Forest Plantation for Wood Production Value Chain
policy (PFCm policy), aimed at reaching 1.5 Mha of commercial plantations by 2025, with five other policy
scenarios, for the period 2015–2047, is presented. We consider two cases with no expansion, and two alternative
goals for expansion reaching 0.765 Mha and 2 Mha, and compare the projected volume available for supply of
wood with current and projected market demand. The exercise was conducted by designing and using the
Colombian Forest Plantation Growth and Yield Simulator for predicting volume available for the supply of the
country’s unprocessed wood market, and as a consequence, the manufactured wood forest product market. No-expansion
scenarios are non-viable options for the current state of the market, while among the expansion paths
considered, both the PCFm policy goal and the 2 Mha scenario produce desirable results. This empirical exercise
could be useful in quantifying the magnitude of the wood shortage concerns already identified by the PCFm
policy, and provide policy-makers and other stakeholders with a more precise framework for adequate action.
1. Introduction
Colombia is a tropical country endowed with rich forest resources,
including an estimated 25 million ha (Mha) of available area for the
development of commercial forest plantations. However, in the past few
decades, the Colombian forest sector has been underperforming, characterized,
among other things, by high rates of deforestation and natural
forest degradation, and difficulties in supplying from its national sources
the forest products demanded. Currently, the country is redesigning its
forest policy and implementing, with help from the international community,
several forest initiatives to boost the forest sector, but no
analytical tools for the forest sector, such as growth and yield simulation
models, are available to the stakeholders involved in these processes for
analyzing the impact of these policies.
Initiatives to restructure the Colombian forest sector have been in the
works, more intensively, since the start of the current millennium. As
part of the Government initiatives, the Ministry of Agriculture, the
Ministry of Environment and Sustainable Development, the Ministry of
Industry, Trade, and Tourism, and the Department of National Planning
of Colombia released the National Forest Development Plan (PNDF for
its acronym in Spanish), in the year 2000 (MMA et al., 2000). One of the
main objectives of this plan was positioning the country as a key player
in the international arena, with a strong presence and negotiating power
over the most pressing topics in forestry: the use, conservation, and
sustainability of its forest ecosystems, as well as international competitiveness
of its wood forest products (MMA et al., 2000). An important
motivator, at the time the PNDF was announced, was Colombia’s
heavily underused capacity for commercial forest plantations. By 1996,
of the roughly 25 Mha estimated available area for commercial forest
plantations, the country had utilized less than 2% for wood production;
this, together with the degradation of national natural forest resources,
resulted in a substantial negative trade balance of US$ 325 million for
the Colombian forest sector, with exports at around US$ 66 million
(Martínez-Cort´es, 1998). In addition, country wide data on reforestation
in the years preceding 2000, consolidated by Martínez-Cort´es (2022),
indicated low and stagnated rates, contrary to the upward trend
observed worldwide for the same period.
The decades that followed the introduction of the PNDF did not
* Corresponding author.
E-mail addresses: [email protected] (´O.G. Martínez-Cort´es), [email protected] (S. Kant), [email protected] (H. Isufllari).
Contents lists available at ScienceDirect
Forest Policy and Economics
journal homepage: www.elsevier.com/locate/forpol
https://doi.org/10.1016/j.forpol.2022.102722
Received 12 December 2021; Received in revised form 6 March 2022; Accepted 9 March 2022
Forest Policy and Economics 138 (2022) 102722
2
present much change to the state of the forest sector and particularly the
wood forest products industry, due, in part, to the delay with which
some of the objectives and goals of this plan were being executed. Two
examples of this were the slow implementation of the tasks to reach the
goals of 1.5 Mha of commercial forest plantations under sustainable
management, and the modernization and expansion of the wood forest
products industry. As marginal results for these goals were obtained
under the PNDF, and the negative trade balance for the Colombian forest
sector continued to widen for all manufactured wood forest products
(Martínez-Cort´es, 2022), stakeholders called for action in the first half of
the 2010s. This call ended up in the Agricultural Rural Planning Unit
(UPRA – its acronym in Spanish) leading a two-year (2016–2017)
consultation process, which involved more than 50 stakeholders
partially to redesign the Colombian forest policy, resulting into what
was called the Commercial Forest Plantation for the Wood Production
(PFCm) Value Chain policy (PFCm policy - its acronym in Spanish).
The PFCm policy was only signed and officially released by the
Ministry of Agriculture and Rural Development of Colombia (MADR) in
June 2019. This policy is set to be implemented in three time-periods
during the next 20 years: 2018–2022, 2023–2030, and 2031–2038,
each period to be evaluated against the original objectives (MADR,
2019). Comprising a set of policy guidelines and an action plan, the
policy aims to continue to develop and consolidate the PFCm value
chain, an initiative which originated in the middle of the 1950s (Martínez-
Cort´es et al., 2018a) and, by 2016, was the main source of raw
wood for the Colombian forest industry, and within the same year, it
generated about 40,000 direct jobs in forest plantations activities alone
(MADR, 2017a). At that stage, the total available commercial forest
plantation area comprised 0.3 million hectares, translating into a yearly
production volume of around 3.0 million cubic meters of underbark
unprocessed roundwood (Profor, 2017).
One important commitment of this policy, agreed upon through
collaboration with the private sector stakeholders, was to continue with
the expansion of commercial forest plantations until it reached 1.5 Mha
by 2025, the same goal of the PNDF. By focusing on increasing the
supply of wood forest products, this policy plans to address two issues
related to growing the country’s GDP through the forest sector: increase
the international market share of Colombian exports and meet the
projected increase in national demand. Estimates for the country’s per
capita consumption of wood forest products put the increase in demand
for these products at 30% by 2030, and at 50% by 2038, both compared
to 2013 levels (Martínez-Cort´es et al., 2018b). The same study also
estimated an increase of 20% in the productivity upon successful
implementation of the PFCm policy, compared to 2016 levels, as well as
an increase in the quality of its wood forest products, on par with international
levels.
Another important area of the Colombian economy expected to
benefit from the PFCm policy is the labour market, which is projected to
experience an increase of up to 500,000 new formal jobs by 2030. Of
these, around 127,000 will be from the commercial forest plantation
expansion and its sustainable management, and the rest will be indirect
jobs, related to the transportation, the traditional wood forest products
industry, and the commercialization of wood and manufactured wood
forest products. Additional jobs are expected to be generated from the
expansion of innovative wood forest product industries, such as bio
composites, and those belonging to the cosmetics, and the food industry,
to name a few.
The environmental sustainability aspect of the PFCm policy promises
to increase sustainable forest management through a full certification of
its practices related to the establishment, management and harvesting,
and of the chain-of-custody standards of the wood produced in these
plantations. By the end of 2016, the Forest Stewardship Council (FSC)
database reported that only around 143,000 ha of the total commercially
forest planted area was FSC certified. By that year, there were no
other certification programmes for forest plantations in Colombia. In
addition to increasing the commercially planted area to 1.5 Mha, the
PFCm policy plans to reach a certification level of 80% of this area by
2030, and an 80% certification of the chain-of-custody standards for its
forest products by the same year (Martínez-Cort´es et al., 2018b). The
contribution of this level of certification is expected to span from protecting
Colombia from deforestation and illegal logging of its vast natural
forests to improvements in water quality, biodiversity protection, as
well as much needed reductions in the carbon footprint.
The PFCm policy comprises 12 different programs and 30 projects
necessary for its successful execution. The total cost of its implementation,
without considering industry modernization and expansion, was
budgeted at US$ 1.73 billion, 35% of which will come from the public
sector, 64% from the private sector, and the rest from international aid.
The majority of these costs are allocated to the expansion of the current
(2016) commercial forest plantations to 1.5 Mha and their management.
Out of the total cost of US$1.73 billion, the cost of commercial plantations
is estimated US$ 1.7 billion, 65% of which will come from the
private sector stakeholders invested in the initiative, and 35% from the
government budget (Martínez-Cort´es et al., 2018a).
Aside from the importance of such policy to both the private sector
and the economy as a whole, a more detailed look at the implementation
of the 1.5 Mha of the PFCm goal suggests that it was defined without a
clear, transparent, and analytical analysis of the impacts that the
development of forest plantations of such magnitude would have on the
wood forest products markets, and the manufactured wood forest
products industry as a whole.1
A good first step towards better insight on the potential impacts of
such policy would be a rigorous estimation of the additional volumes of
wood added to the market. The heterogenous and multidimensional
composition of the source of wood, and the industry that utilizes it make
the calculation of volume production from the commercial forest plantations
fairly complex. As of 2015, the commercially planted area
comprises uneven aged plantations of more than 20 distinct species,
planted by large, medium, and small landholders (Profor 2017).2 These
species are planted in different geographical areas, and as such, their
productivity levels are subject to variations in soil composition, as well
climate and other environmental conditions. At the management level,
the plantations serve different production objectives, such as wood,
fiber, bioenergy, and non-wood/wood forest products. Given that such
variation is present in both within the same planted area or between two
or more different areas, this, in turn, has generated multiple forestry
management regimes. The PFCm policy plans to achieve the 1.5 Mha
goal by expanding most of the country’s current commercial plantation
areas. It will do so through the planting and management of 16 of more
than 20 existing species, specifically the ones with the highest production
efficiency.3 In doing so, the additional plantations inherit the same
complexity as the current ones.
1 The Colombian wood forest product markets comprise two markets: the
unprocessed wood market and the manufactured wood forest products market.
The manufactured wood forest products industry includes the wood industry,
the furniture industry and the pulp and paper industry. The Colombian wood
industry comprises the sawnwood industry; the wood-based panel industry;
roundwood industry for poles, piling, posts, etc. preserved or not; and the industry
that elaborates a series of manufactured wood products such as
moulding and pallets. In this paper, the traditional manufactured wood forest
products industries of pulp, wood-based panel other than veneer and plywood,
jointly with the innovative manufactured wood forest products industries
(forest bioenergy, bio-composites, and derivates from wood for the chemical,
cosmetic and food industries) are grouped under the acronym of CITPI.
2 For a 2021 up-to-date distribution on commercial forest plantation area in
Colombia refer to MADR (2021). The breakdown of this area by species in 2015
and 2020 is almost the same.
3 Pinus caribaea, P. maximinoi, P. oocarpa, P. patula, P. tecunumanii, Eucalyptus
camaldulensis, E. grandis, E. pellita, E. tereticornis, E. urophylla, Tectona grandis,
Gmelina arborea, Acacia mangium, Cupressus lusitanica, Pachira quinata y
Ochroma pyramidale (Martínez-Cort´es et al., 2018a).
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
3
As a result, there is an imminent need for a dynamic and comprehensive
analysis that uses a systematic and robust way of predicting the
volume increase upon the implementation of the expansionary policy,
which can be utilized by policymakers when evaluating the policy
implementation results. In this paper, we provide such analysis through
an exercise that evaluates the results of the proposed 1.5 Mha expansion
plan of the PFCm policy and five other commercial plantation expansionary
paths, for comparison with the ability of said policy to satisfy
projected quality-based demand from each wood product industry.
In order to conduct our analysis, we developed the Colombian Forest
Plantation Growth and Yield Simulator (SCRPFC, by its acronym in
Spanish) using the site-quality and species-specific data from the existing
commercial plantations. This simulator was designed as part of the
Colombian Forest Sector Model (CFSM) which is the main focus of the
Ph.D. thesis of the first author. The Simulator estimates volume available
for wood supply, and uses it to calculate the expected volume of
wood for different industrial use, that will be useful to policymakers
during their evaluation stage of the policy. We compare the results of six
policy scenarios of commercial forest plantations and discuss their implications
for wood supply to forest industry.
The rest of the paper is organized as follows: Section 2 presents the
literature review. Section 3 introduces the methodology and the simulator.
The theoretical details of the simulator are given in Appendix A.
An empirical exercise comparing six expansion scenarios as well as the
results of the exercise are presented in Sections 4 and 5. We discuss the
findings and conclusions in Section 6 and include some final thoughts
and future research in Section 7.
2. Literature review
Projecting the future availability of wood supply on a national scale
is the key to analyzing a country’s long-term forest industry development
and the supply and demand of forest products markets. This is one
of the reasons why, during the twentieth century, many countries in the
world (especially in Europe and North America), implemented forest
resource monitoring systems, started nation-wide assessments and later
on developed wood availability projection systems that include mostly
country-specific tools, models and simulators (Barreiro et al., 2017a).
The important features of these projection systems are growth and
yield models. These models, which are the representation of an average
growth of trees (Salas et al., 2016), are used to predict stand of tree
development through time (Twery, 2004). These models can be classified
as follows: yield tables, aggregated level (also known as stands-level)
models, and individual-tree models (Munro, 1974; Salas et al.,
2016; Vargas et al., 2016).
The yield tables represent the average value of a stand’s variables
(Salas et al., 2016). These are the oldest approach to predict the growth
of an even-aged forest, with the first yield table published in Germany in
1787 (Vanclay, 2014). With the continuous development of statistics,
yield tables have improved, and by incorporating equations that
describe variables through time, they have become a reliable method
(Vanclay, 2014). On the other hand, stand-level models use the stand as
a single model unit, projecting the number of trees, basal area, or volume
per stand based on information on age, stand density and site index
(Barreiro and Tom´e, 2017). Although these models omit variations
within the stand, they are widely applied around the world (H¨ark¨onen
et al., 2010). An example of the former are the growth and yield prediction
models for various silviculture regimes, based on large databases
of experimental data from large-scale industrial pine plantations in
southern US, developed by the Plantation Management Research
Cooperative at University of Georgia in collaboration with the forest
industry. Aside from considering growth and yield responses to several
silvicultural management regimes, the co-op’s work has extended the
analytical framework of yield prediction to evaluate productivity and
profitability of forest plantations by constructing a profit function for
timber production, which offers additional information about the
production of this commodity (Yin et al., 1998).
Stand-level models are divided in matrix models and models based
on probability density function (Salas et al., 2016). Finally, individual-tree
models project the diameter and height increments of individual
trees, taking into account variables such as tree size, age of stand, stand
density and site index. They can also be classified as spatial or non-spatial
models (Barreiro and Tom´e, 2017).
Another key feature of projection systems are the simulators which
range from standard simulators that combine forest inventories and
yield tables to complex ones that include a national inventory database
and functions/algorithms, which provide helpful information on forest
management and assist policy making decisions (Barreiro and Tom´e,
2017; H¨ark¨onen et al., 2010). An overview of differences between projection
systems implemented in Europe and North America are
described in depth by Barreiro et al. (2017b).
In South America, and specifically in the Neotropics which includes
Colombia, the literature does not report the existence of national projections
systems, but only that of specific simulators for some of a
country’s regions and species available. Among those, one of the most
relevant works has been done in Chile, through two simulators:
Nothofagus (Martin et al., 2020) and AMPL-CPLEX (Büchner et al.,
2019; Instituto Forestal, 2013), aimed at both native forest and plantations
(Pinus radiata, Eucalyptus spp. and Pseudotsuga menziesii),
respectively.
In the Neotropic region, SILVIA from Costa Rica was created to
manage and model the growth of Tectona grandis and Gmelina arborea at
a stand-level or stand group, including species criteria, site index,
thinning programs and growth scenarios. It is open to the public and
comprises many equations that users can choose, and recently it is also
being used for other species from natural forests (Serrano et al., 2008).
In Panama, a simulator using an IPTIM (Integrated Planning for
Timberland Management) Assets software and SIMO (SIMulation and
Optimization framework), where plot inventories are processed to
obtain tree-level volume models, taper curve models and stand-level
yield models, is designed to describe forest dynamics, project growth
and yield, and support decision-making in Teak plantations (Sepp¨anen
and M¨akinen, 2020). Another simulator for T. grandis was built using
SIMILE in Venezuela, with the purpose of obtaining growth of plantations,
harvest volume and carbon sequestration capacity based on
quality index, initial density of plantation and thinning regime (Jerez
et al., 2015).
In Colombia, there exist some simulator-like tools for estimating the
growth and yield of forest plantations, but the majority of them tends to
be not as elaborate, or are limited to a particular species, parameter and
geographical scope. For example, L´opez et al. (2007) have worked on
models for forecasting growth and yield based on basal area and volume,
to determine the optimum rotation age of Pinus caribaea in the Orinoquia
Region in Colombia. Restrepo Orozco (2010), developed an empirical
model based on the Bertalanffy-Chapman-Richards model, aimed at
describing the yield and growth of T. grandis and Pinus patula in the
Norwest Region in Colombia, specifically in Antioquia, C´ordoba and
Sucre Departmentos (an administrative and political divisions of
Colombia is called Departamento, in Spanish), considering age and
environmental covariables.
Barrios et al. (2014) developed equations of total volume and ratio
for Eucalyptus grandis, using data from 101 trees collected in plantations
located in Quindio Departamento (Central Region of Colombia). L´opez
et al. (2015) have developed some models of trunk profiles with an
autoregressive structure for errors for Eucalyptus tereticornis growing on
the country’s Caribbean region and, Melo and Lizarazo (2017) developed
equations to estimate tree volumes by applying a single taper
polynomial function.
The most complete growth and yield simulator for the Colombian
forest plantations is the SimFor v 1.2 Software, which was developed in
2013 by the Institute of Informatics of Southern University of Chile
(UACh) for Colombia’s National Corporation of Forestry Research and
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
4
Development (CONIF), for which growth and yield models were calibrated
with five-year consecutive data coming from 700 permanent
plots, distributed all over the Colombian forest plantations. Models were
developed at a stand level, disaggregated to diameter classes. The
simulation uses three transition functions (number of trees, basal area
and dominant height) as algebraic difference equations. As state variables
are known at any given age, the number of trees under each
diameter class is estimated by recovering the parameters of a Weibull
function. A height-diameter function predicts the height of the mean
tree in each diameter class. Then, a taper function is used to predict the
total volume grouped by products at each diameter class. Although
powerful, there are two main issues that arise with SimFor: the institution
(CONIF) which owns the software does not make it available for
independent and academic research, and the use of a Beta version of the
software with restricted features is not feasible, as the information
needed to perform the simulations (i.e. state variables for each stand
comprising the total mass of commercial forest plantations on December
31st of 2015, or for any other time period is not yet available).
In summary, given that no other viable option exists for the
Colombian forest sector, the development of a proper growth and yield
simulator was instrumental for evaluating wood supply implications of
the proposed 1.5 Mha expansion plan of the PFCm policy and other
commercial plantation expansionary paths.
3. Methodology
The SCRPFC is a tool that facilitates projections of the growth and
yield for any stand in the commercial forest plantations. For the purpose
of the SCRPFC and our paper, a stand is an area of commercial plantation
homogeneously planted with trees belonging to one type of species, all
at the same age, and within the same location. This classification is
intended to minimize the noise in the data originated from all the variations
mentioned in the introduction such as age, species, and site-quality.
The simulator can be described in three stages of operations: (i) the
calculation of total growth at a stand level at any given year; (ii) the
estimation of distribution of each year’s growth by diameter class; and
(iii) the determination of the current volume of available stock appropriate
for harvesting in order to wood supply to different forest industries.
Generally, in the first stage, the transition equations for stand
density, dominant height, and basal area are estimated. In the second
stage, growth parameters and volume are calculated for each diameter
class. In this estimation, first the parameters of a probability function,
Weibull distribution in our case, are calculated to find the number of
trees in different diameter classes. After that the dominant diameter and
the mean height are calculated for each diameter class, and the calculated
value of the mean height and the diameter are used to calculate the
total volume in each diameter class at a given point of time (year). In the
third stage, using the volume taper model, the volume of available wood
of different under-bark diameters (which is commonly used to measure
industrial wood) is calculated. The technical details of all three stages
are given in Appendix A.
The three stages of the simulator represent the simulation steps at a
theoretical level. Given the heterogeneity in species type, tree shape,
and geographical location of the stands, additional adjustments were
required, which include the following: For stage one we used the 52
growth and yield tables generated and already available in the Program
on Forests (Profor) project of the World Bank on Colombia (Profor,
2017); while for stage three, we utilized the species-specific taper
equations developed by L´opez et al. (2011), with data from the permanent
plot network of the commercial forest plantations in Colombia.
4. Empirical exercise
The Profor report, published by the World Bank in 2017, estimated
the existing commercial plantations of 0.3 Mha, by the end of 2015. The
Profor report estimated this area based on Colombian Agricultural Institute’s
(ICA by its acronym in Spanish) database of official registration
of commercial forest plantations (Held et al., 2017). The Profor dataset
is the only database that includes the disaggregated data of commercial
forest plantations (that includes species, location, site quality, municipality,
administrative division (which is also called Departmento in
Colombia), and region for each plantation stand). A new database which
includes all these and other features up to the end of 2021 is now being
prepared under the Ph.D. research of the first author. In addition, one of
the distinguishing features of our Simulator, compared to other simulators,
is that it incorporates the ICA database, which is regularly
updated as per registration requirements implemented in 2019 (MADR,
2019).
The Colombian state reported the commercial plantation area of 0.45
Mha by December 2015 (MADR, 2017b), and about 0.5 Mha as of June
2021. However, the MADR dataset, that reported area of 0.45 Mha, is
disaggregated only to the levels of species and administrative district. In
addition, the MADR results had not been independently reviewed, while
the Profor data was supervised by World Bank specialists. Hence, this
dataset does not have sufficient disaggregation for the use in simulation
exercises.
We, therefore, use the Profor dataset as the base for our analyses of
different scenarios. The Profor dataset is only available for the period
until December 2015, and therefore we use 2015 as the initial year of
our analysis. To account for the reported figure of 0.45 Mha plantations
in 2015, as per the MADR data, we use it as a base for one of our six
scenarios - Scenario 2.
A map of the commercial forest plantations, as per the Profor report
for 2015, in Colombia is presented in Fig. 1. A breakdown of this area
into species, and administrative districts and regions is shown in Figs. 2
and 3, respectively. As mentioned earlier, in Colombia, the administrative
unit is called Departamento in Spanish which we are referring as
department/division. The boundaries of different departments are
shown in Fig. 1.
Next, we identified six scenarios. For all scenarios, we assumed that
the main objective of all commercial plantations is the production of
wood. It was also assumed that all plantations are thinned and subjected
to final harvesting at the age indicated by the growth and yield model of
each forest species and site-class considered. For the first two scenarios,
we assumed that there is no increase in the commercial plantation area
either by public or private sector. In other words, the objectives of the
PFCm policy to increase commercial plantations are not executed. In the
base scenario (Scenario 1), we assumed that 301,146 ha of commercial
forest plantations in 2015, as per the Profor report, is maintained in
perpetuity. In Scenario 2, we assumed that 450,000 ha of commercial
forest plantations in 2015, as per the MADR report, is maintained in
perpetuity. For Scenario 2 modeling, 301,146 ha of Scenario 1 was used
as the base and the difference plantation area of 148,584 ha was
assigned proportionally to the departments and species to the commercial
forest plantations in which plantations of Scenario 1 were
registered in 2015. We call these two scenarios as no growth scenarios.
Next three scenarios are growth scenarios. In these scenarios, the
initial plantation area (0.3 Mha) is incremented, starting in 2016, in
order to reach 0.765 Mha in 2035 (Scenario 3), 1.5 Mha in 2025 (Scenario
4), and 2.0 Mha in 2025 (Scenario 5), respectively. The scenario 3
(0.765 Mha) is the goal for commercial forest plantations for wood
production that Colombia should have under sustained forest rotation,
suggested by Profor (Held et al., 2017). The scenario 4 (1.5 Mha) is the
goal set by the National Forest Development Plan in 2000, expected to
be reached by 2025 (MMA, et al., 2000), which is also the goal for the
2019 PFCm policy (Martínez-Cort´es et al., 2018b), while the scenario 5
is an alternative goal to the scenario 4. In all these five scenarios, the
initial area and the target area are maintained under sustained rotation
(i.e., the initial area and the target area never decrease).
The sixth scenario is about no new plantation and the harvest of the
initial plantation area (0.3 Mha in 2015) starting in 2016 without
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
5
Fig. 1. Geographical distribution of Colombia’s commercial plantations on December 31, 2015. (Source: UPRA, 2017. Not official, without scale.) Yellow dots: the
forest plantation stands. Red lines: major highways. The name of political and administrative divisions of Colombia (departaments) are printed in black capitalized
letter with their corresponding capital city signaled by a red dot.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
6
Fig. 2. Colombia’s commercial plantations on December 31st, 2015, by species, in ha.
Fig. 3. Colombia’s commercial plantations on December 31, 2015, by administrative divisions (departments) and regions, in ha.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
7
replacement; consequently, the commercial plantation area eventually
diminishes to zero. The key features of all six scenarios are presented in
Table 1.
5. Results
We run the Colombian Forest Plantation Growth and Yield Simulator
(SCRPFC) for estimating the expected volume of wood available for
supply from each scenario for the period 2015–2047. Wood available for
supply from commercial forest plantations is the wood that would be
obtained from a stand which is subject to thinnings and final harvest at
the age that the management regime specifies for the species planted in
such stand. The wood logs obtained from the thinnings and final harvest
must comply, at least, with the minimum specification for industrial use.
In 2047, all stands in scenarios S1 to S5 would have completed, at least,
the first rotation period under sustained management (i.e., would have
been subject to final harvest and been planted for second rotation).
A visual representation of the projected expansion area for the six
scenarios is given in Fig. 4. The expansion of areas by species and
geographical location is presented in Appendix B & C.
Simulation results indicate that wood availability from commercial
forest plantations for wood production during the simulation period
would range between 71 and 868 million m3 of underbark roundwood
(Mm3rsc, for its acronym in Spanish), depending on the simulated scenario.
When the initial area for wood production diminishes to zero (S6),
Colombia would only have a one-time supply of 71 Mm3rsc of wood
available for harvesting. When the plantation area increases to 2 Mha
(S5) in 2025, which would then be maintained under sustained rotation,
Colombia would have 868 Mm3rsc of wood during the 2015–2047
period, and a similar quantity every 32 years, starting in 2047. The wood
available for the period of 2015–2047, for the six simulated scenarios is
shown in Fig. 5.
The volumes in Fig. 5 were obtained by strictly applying the management
regime to each stand that makes up the total area of commercial
forest plantation in each one the six scenarios simulated. In
other words, the volume was obtained by subjecting every stand to both
thinning and final harvesting, at an age as per the guidelines outlined in
the usual Colombian silvicultural model for each species considered in
the simulation. It does not reflect any other harvesting done for satisfying
actual or potential market demand for unprocessed wood. In
addition, the calculations reflect volumes of stands which have not
reached full maturity. For stands which, as of December 31, 2015, are
already at full maturity or past it, we suppose that they are harvested at
Table 1
Scenario simulations for expansion of commercial forest plantations in
Colombia.
SCRPFC Scenario Simulation
Description
Scenario Code Observations
NO INCREASE in the existing
area of commercial forest
plantations on December 31,
2015, (BASE SCENARIO at
301,146 ha)
S1:
PROFOR2015
Total area of commercial
forest plantations of 301,146
ha is maintained in perpetuity
(that is, once an area
undergoes the final harvest, an
area of equal size must be
established within forest
plantations).
NO INCREASE in the existing
area of commercial forest
plantations on December 31,
2015, (BASE SCENARIO at
450,000 ha)
S2: MADR2015
For modeling purposes, the
base of 301,146 ha of the
PROFOR2015 scenario was
used and 148,584 ha were
assigned proportionally (in
addition to other criteria) to
the departments and species in
the commercial forest
plantations in which these
were registered in 2015. Total
area of commercial forest
plantations of 450,000 ha is
maintained in perpetuity.
INCREASE of the existing area
of commercial forest
plantations
to 765,000 ha
(Completion date: December
31, 2035)
S3:
PROFOR2035
Area proposed by the Profor
project. Profor’s proposal
includes increases in
plantation yields and the
establishment of 463,584 ha of
new pine and eucalyptus
plantations at an annual rate
consistent with the historical
trend of annual establishment
of commercial forest
plantations in Colombia. For
modeling purposes, the base of
301,146 ha of the
PROFOR2015 scenario was
used and annual stands were
added for pine and eucalyptus
species in the departments
where the plantations were
located in 2015. Total area of
commercial forest plantations
of 765,000 ha is maintained in
perpetuity.
INCREASE of the existing area
of commercial forest
plantations
to 1,500,000 ha
(Completion date: December
31, 2025)
S4: PNDF/
PAPFCm1
Proposed area of the National
Forest Development Plan-
PNDF, to be reached by 2025.
This proposed goal is the same
as the one in the PFCm policy,
developed by UPRA. For
modeling purposes, the base of
301,146 ha of the
PROFOR2015 scenario was
used and, as of 2015, annual
stands were added for 16
species in 14 departments
located in the 3 strategic
development regions for forest
plantations for commercial
purposes defined in the Policy
Guidelines of the PFCm chain.
Total area of commercial
forest plantations of
1,500,000 ha is maintained in
perpetuity.
INCREASE of the existing area
of commercial forest
plantations
to 2,000,000 ha
(Completion date: December
31, 2025)
S5: PNDF/
PAPFCm2
Area proposed in the PNDF /
PAPFCm1 scenario (S4 above)
and an additional 500,000 ha;
the annual planting rates
between 2020 and 2025
revised accordingly. The total
plantation area of 2,000,000
Table 1 (continued )
SCRPFC Scenario Simulation
Description
Scenario Code Observations
ha to be maintained in
perpetuity.
DECREASE of the existing area
of commercial forest
plantations to zero ha
S6:
PROFOR2015-
SR
Scenario 1 without
replacement of annual
harvested area. The total area
of plantations reaches zero in
2035. For the purpose of
calculating the supply of
wood, it has been assumed
that the volume (both thinning
and final harvest) that should
have been harvested until
December 31, 2015, and is still
standing as of that date, will
be harvested in equal quotas
each year between 2016 and
2034.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
8
the same order as they were planted, starting with the earliest one being
harvested in 2016.4
Table 2 presents a more detailed information on the available volumes
for supply for the period of the simulations, showing, among
others, the average total annual volumes, stumpage volumes for
matured, overmatured, and stands where full maturity has not yet been
reached. Other indicators included in the table are the average of the
additional volumes that are added for every expansionary goal after
2015, and volumes if the total forest plantations were a normal forest.
In the Colombian regional context (Fig. 6), under S1 & S2,
approximately half of the wood would be available in the Eje Cafetero y
Suroccidente region, a quarter in the Caribe region, and 15% in the
Orinoquia one. Meanwhile, for S4 & S5, around half of the wood would
be available in Orinoquia, a bit over a quarter in Eje Cafetero y Suroccidente,
and 20% in the Caribe region. In the S3 case, 38% of the total
wood would be available in the Orinoquia region, 36% in Eje Cafetero y
Suroccidente, and 21% in the Caribe region.
With regards to species types, between 55% and 78% of all the wood
available would correspond to the wood of the species of genres Pinus
and Eucalyptus for all the simulated scenarios. If wood from these species
is added to wood of Acacia mangium, T. grandis and G. arborea, they
would account for 87% to 97% of all of the wood available in all six
scenarios simulated.
The annual average availability of wood for different industrial use
from commercial forest plantations for the six scenarios in Colombia is
given in Table 3. The table also includes the details of available wood
Fig. 4. Annual planted area of commercial plantations under six scenarios in Colombia.
4 A stand that has reached the age of final harvest according to the management
regime for a species is considered a mature one. Detailed data on
stands for scenarios simulated and the management regime applied may be
provided as per request to the first author.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
9
from final harvesting and thinning.
6. Discussion, policy recommendations, and conclusions
For convenience, the analysis under this section is conducted in the
absence of the wood supply from other Colombian sources (natural
forests, agroforestry systems and trees outside the forest) and from imports
of unprocessed wood. So, it is important to bear in mind that the
way some findings are presented does not imply that in practice “it must
be the case”, but rather that the figures allow to present the findings in
such fashion. This is an important issue for wood supply and capacity
expansion of the Colombian sawnwood industry, when the levels of
wood supply and expansion are related just to wood coming from forest
plantations, although in reality half of the current supply for this industry
comes from natural forests.
Also, the present analysis is done without taking into account the
possible positive effects on timber supply from gains in productivity that
may be derived from any additional foreseeable intensive plantation
management other than the current ones applied to forest plantations in
Colombia (e.g., additional improvements in competing vegetation control
and fertilization) and any genetic improvement of trees. At present,
this is not a common practice in the Colombian commercial forest
plantation, and should this change in the future, the model presented in
this paper will need to be updated in order to best capture the potential
gains. This can be modelled by following the guidelines on how the
supply of wood from forest plantations can be affected by these two
strategies in Yin and Sedjo (2001) and Yin et al. (1998), respectively.
6.1. Supply and expansion of the manufactured wood products industry
6.1.1. No expansion paths (Scenarios 1 & 2)
Due to constraints related to the property and distance to the industrial
facilities and markets, among others, for scenarios maintaining
initial areas (0.3 Mha and 0.45 Mha) under sustained rotation, the actual
Fig. 5. Availability of wood from commercial plantations under six scenarios in Colombia, 2015–2047 (million m3 of underbark roundwood).
Table 2
Wood available from commercial plantations for the six scenarios in Colombia (Period: 2015–2047; Volume unit: Mm3rsc).
Scenarios S1:
PROFOR2015
S2:
MADR2015
S3:
PROFOR2035
S4: PNDF/
PAPFCm1
S5: PNDF/
PAPFCm2
S6:
PROFOR2015-SR
Total wood available (A) 155 211 303 665 868 71
Annual Average of (A) above 4.8 6.6 9.5 20.8 27.1 2.2
Available volume of matured and over-matured stands on
December 31, 2015
18 22 18 18 18 18
Available volume of not yet matured stands at December 31, 2015 6 7 6 6 6 6
Total availability of wood accumulated during 2015–2047,
without the initial volumes on December 31, 2015 (B)
131 182 279 641 844 47
Annual average of (B) above 4.1 5.7 8.7 20.0 26.4 1.5
Annual average of wood available if the forest under sustained
rotation had been normalized (Normal Forest) in the period
2015–2047
5.1 7.4 13.6 26.1 34.8 N/A
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
10
Fig. 6. Strategic development regions’ potential area for commercial forest plantation development under the PCFm policy. Source: UPRA, 2017. Dark green: the
best potential areas; Light green: the second-best potential areas; and Yellow: the third-best potential areas.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
11
availability of wood would be less than that shown in the simulation
results in Table 2 (less than 155, and 211 Mm3rsc, respectively). Instead
of being, on average, 4.8 and 6.6 Mm3rsc/year (row two of Table 2),
they would be 4.1 Mm3rsc/year and 5.7 Mm3rsc/year (row 6 of Table 2)
for Scenario 1 and 2, respectively.
In Profor, 2017, the installed production capacity of the pulp, wood-based
panels, and sawnwood industries, was estimated at 4.8 Mm3rsc/
year (1.0 Mm3rsc/year for pulp, 0.8 Mm3rsc/year for panels and 3.0
Mm3rsc/year for sawnwood (Martínez-Cort´es et al., 2018b).5 It can be
seen that the volume generated by the initial area of 0.3 Mha is insufficient
for providing wood that meets the 4.8 Mm3rsc/year potential
demand (at a 100% production capacity utilization) of the manufactured
wood primary products industry i.e., woodpulp, wood-based panels and
sawnwood industries. On the other hand, the 5.7 Mm3rsc/year projected
Table 3
Annual average availability of wood from commercial forest plantations for the six scenarios in Columbia (By harvest type and industrial use); All volume values in
Mm3rsc.
Scenario Harvest type Industrial use* Sub-Period Total
2015–2047
Percent
2015 2016–2022 2023–2030 2031–2038 2039–2047
S1: PROFOR2015
Total 0.7 3.0 4.2 3.7 5.3 4.8 100
Final
harvest
High quality sawnwood 0.2 1.0 1.9 1.2 2.3 1.9 39
Low quality sawnwood 0.1 0.6 0.9 0.8 1.2 1.0 21
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.5 0.7 0.7 0.9 0.8 16
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.3 0.8 0.7 0.9 0.9 1.1 23
S2: MADR2015
Total 0.9 3.7 5.9 5.6 7.0 6.6 100
Final
harvest
High quality sawnwood 0.3 1.3 2.5 1.9 3.0 2.5 38
Low quality sawnwood 0.2 0.8 1.3 1.3 1.6 1.4 22
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.6 1.1 1.1 1.3 1.2 18
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.4 1.0 0.9 1.2 1.2 1.4 22
S3: PROFOR2035
Total 0.7 3.0 5.4 9.1 15.8 9.5 100
Final
harvest
High quality sawnwood 0.2 1.0 2.1 2.5 6.5 3.4 36
Low quality sawnwood 0.1 0.6 1.1 1.8 3.6 2.0 21
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.5 0.8 1.3 2.4 1.4 15
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.3 0.8 1.5 3.4 3.2 2.6 27
S4: PNDF/
PAPFCm1
Total 0.7 3.6 15.1 24.6 33.2 20.8 100
Final
harvest
High quality sawnwood 0.2 1.0 3.2 7.2 13.0 6.7 32
Low quality sawnwood 0.1 0.6 2.6 6.1 8.2 4.8 23
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.5 1.9 4.2 5.5 3.3 16
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.3 1.4 7.3 7.0 6.4 6.0 29
S5: PNDF/
PAPFCm2
Total 0.7 3.6 18.2 33.0 45.5 27.1 100
Final
harvest
High quality sawnwood 0.2 1.0 3.4 9.4 18.0 8.7 32
Low quality sawnwood 0.1 0.6 2.8 8.3 11.2 6.2 23
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.5 2.1 5.6 7.5 4.2 16
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.3 1.4 9.9 9.6 8.6 7.9 29
S6: PROFOR2015-
SR
Total 0.7 2.8 2.6 0.9 0.0 2.2 100
Final
harvest
High quality sawnwood 0.2 1.0 1.5 0.5 0.0 1.0 43
Low quality sawnwood 0.1 0.6 0.6 0.2 0.0 0.5 21
Wood-based pannels, pulp, and forest
bioenergy
0.1 0.5 0.5 0.1 0.0 0.3 15
Thinnings
Wood-based pannels, pulp, and forest
bioenergy
0.3 0.7 0.1 0.0 0.0 0.5 20
High quality sawnwood (Min. diameter > 20 cm), Low quality sawnwood (Max. diameter < 20 cm & min. Diameter > 15 cm), Wood-based pannels, pulp, and forest
bioenergy (Max. diameter < 15 cm & min. Diameter > 5 cm.
* Industrial use has been defined based solely on the size of the logs for the industry of traditional primary wood forest products (sawn wood, boards and pulp), and
excludes wood for energy from forest biomass, which is included within of the innovative wood forest products industry. Within the classification, no reference is made
to the industry of traditional higher value-added wood forest products (eg. paper and cardboard, furniture, structures and wood carpentry and other products of
secondary processing), since almost all the products of this industry go through the pulp, sawing and board industries.
5 In 2016 the wood supply of the Colombian woodpulp industry was provided
entirely from commercial forest plantations; for the wood-based panels, it
comprises wood mainly from commercial forest plantations and negligible
amounts of sawndust and other sawnwood industry residues of wood from
natural forests. The supply of wood for the sawnwood industry was made up by
a mix of both sources: natural forests and commercial forest plantations, with
an increasing share of wood from the latter source during the recent years
(Martinez-Cortes et al., 2018b). Production capacity and expansion figures are
referred to the equivalent cubic meters of roundwood excluding bark needed to
produce an amount of manufactured wood products.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
12
from the initial area of 0.45Mha, although enough to satisfy industry
demand, does not allow for any considerable demand increases from the
industry. In addition, as we mentioned earlier, the details of the figure of
0.45 Mha plantation area are not available at the same level as of the
figure of 0.30Mha, and therefore the wood availability of 5.7 Mm3rsc/
year may be questioned. Hence, this makes the dilemma on which initial
area to use of little importance for our discussion.
This shortage of wood was an issue already suspected during the
redesigning process of the PFCm policy in 2017, but given that no actual
calculations were performed, it was only mentioned in a rather
descriptive fashion. One important contribution that we hope to provide
with our simulations is that, now that the figures on wood availability
from the initial area of forest plantation have become available, those
stakeholders that argued that there is a shortfall in wood supply from the
commercial forest plantations in Colombia can now back their argument
by an empirical analysis.
6.1.2. Three expansion paths (Scenarios 3–5)
Maintaining under sustained rotation the areas of 0.765 Mha (S3),
1.5 Mha (S4) and 2.0 Mha (S5), Colombia would have, on average, 9.5
Mm3rsc/year, 20.8 Mm3rsc/year and 27.1 Mm3rsc/year, of available
wood on its commercial forest plantations for wood production (14
Mm3rsc/year, 26 Mm3rsc/year and 35 Mm3rsc/year, respectively, under
the concept of Normal Forest).6 In the following 30 years, this raw
material would allow for expansions of the sawnwood industry (with
estimated current installed capacity of 3.0 Mm3rsc/year) by approximately
two, three and five times, respectively (three, four and six times
when the forest mass in each scenario reach the behaviour of a Normal
Forest). To date, and to the best of our knowledge, there has been no
expansion on installed capacity of this industry or the rest of the
Colombian forest industries. Therefore, the simulations in this paper are
performed using the above rate, without loss of generality. At the same
time, the 9.5 Mm3rsc/year, 20.8 Mm3rsc/year and 27.1 Mm3rsc/year
(14 Mm3rsc/year, 26 Mm3rsc/year and 35 Mm3rsc/year, under a
Normal Forest) average production would allow for the expansion of the
CIPTPI (group of industries of pulp, wood-based panels (boards), and
innovative manufactured wood forest products), by approximately two,
five and six its current (2017) installed capacity, respectively (four, six
and eight times if the forest mass is a Normal Forest).7
Under S3, if the Colombian sawnwood industry remains at its current
installed capacity (3,0 Mm3rsc/year) and operates at full capacity,
starting from the 2023–2030 period, its demand could be entirely
satisfied with wood coming from commercial forest plantations
(currently about 50% of its total supply comes from natural forests). This
industry can also be expanded to a capacity production of 4 Mm3rsc/
year during the period 2031–2038, and then to 10 Mm3rsc/ year during
the 2039–2047 period. For the same time periods, the 2017- CIPTPI
production capacity (1,8 Mm3rsc/year) could be expanded near to 5
Mm3rsc/year and 6 Mm3rsc/year, respectively.
Under S4, the installed capacity of the sawnwood industry could
increase to around 6 Mm3rsc/year during 2023–2030, and to 13
Mm3rsc/year during 2031–2038. At the same time, the CIPTPI could
increase its installed capacity to 9 Mm3rsc/year during the years between
2023 and 2030, and then to 11 Mm3rsc/year during the years of
2031 to 2038.
Under S5, sawnwood industry expansion could reach 6 Mm3rsc/year
during 2023–2030, and up to 18 Mm3rsc/year during the period of
2031–2038. In turn, the installed capacity of the CIPTPI could be
expanded to 12 Mm3rsc/year during 2023–2030, and then to 15
Mm3rsc/year for 2031–2038.
In the previous paragraphs, we have projected industry expansion for
S4 and S5 only up to the period of 2031–2038 while Table 3 shows a
peak of production in period 2039–2047. However, when analyzed
under the focus of a Normal Forest, the figures for average volume of
wood available for supply showed in the last row of Table 2 are closer to
the figures under column corresponding to period 2031–2038 for S4 and
S5. Hence no additional expansion is projected for the period of
2039–2047 for the mentioned scenarios.
The above analysis factors in the different expansion rates for each
scenario, given that they have to be completed by a specific year, as
shown in Fig. 4. For S3, the expansion starts in 2018 with an initial
establishment of 4335 ha, and gradually increases over 17 years at an
average rate of 2450 ha/year, translating into an establishment of
45,884 ha in 2035, when the target area of 0.765Mha is reached. For S4,
the annual planting rate is approximately 10 times that of S3, in order
for the total area under S4 to reach 1.5 Mha by 2025, as indicated in the
PFCm policy. Finally, for S5, the plantation rate between, 2021 and
2025 is even higher than that of S4, since the objective of S5 is to reach a
commercial forest plantation area of 2 Mha by 2025. As shown in the
same figure, the annual planting rate up to 2020 is the same for both S4
and S5, due to the necessary time it takes for changes of such magnitude
to be implemented.
Note that, as Fig. 5 illustrates, the projected volumes for all scenarios
during the period 2016–2022 are pretty close, which makes comparison
of no additional importance to the discussion.
6.1.3. Harvests without replacement (Scenario 6)
If the initial area of commercial forest plantations for wood production
in Colombia (0.3 Mha) were to be harvested without replacement
(S6), and using a harvest rate of area equivalent to 3 Mm3rsc/ year
as per Profor (2017), the initial area would only provide wood for the
years up to 2038, and perhaps a few months in 2039, for the operation at
full capacity of the current industries of pulp, wood-based panels, and
sawnwood. After 2039, the total and average volumes available would
be zero.
6.2. Viability of alternative forest policy goals and their impact on the
forest markets
The above results seem to indicate that no scenario that promotes the
diminishing of the initial area of commercial forest plantations for wood
production in Colombia is a viable one. Decreasing the initial area by
any degree, or bringing it to zero, would put under risk the existence of
the Colombian commercial forest plantation for wood production value
chain (PFCm value chain), an important value chain with more than six
and half decades of generating multiple societal benefits, including,
according to (MADR, 2017a), 40,000 jobs directly related to establishing,
managing and harvesting forest plantations.
Simulation results also point out that scenarios with zero increase of
the 2015 initial area (0.3 Mha and 0.45 Mha), but with uninterrupted
maintenance of existing area under sustained rotation, are also not
viable. Under these scenarios, the 4.1 Mm3rsc/year and 5.7 Mm3rsc/
year of wood availability, respectively, during 2015 to 2047, on average,
would be insufficient for supplying the unprocessed wood market, and
for the expansion of the country’s manufactured wood primary products
industry. The already-delayed industry expansion is much needed to
meet the increasing demand for manufactured wood primary products
in Colombia (pulp, panels and sawnwood). This demand was estimated
to be around 5.5 Mm3rsc of equivalent wood in 2013, with a shortfall in
domestic supply of 1.8 Mm3rsc of wood equivalent, which was covered
by imports (Held et al., 2017). It should be pointed out that almost the
entire demand for primary products is for national consumption; in 2013
6 We also calculated the annual averages of wood availability using the
concept of Normal Forest. The results are given in the last row of Table 2.
7 Innovative manufactured wood forest products include forest bioenergy,
forest bio-composites, and chemical, cosmetic and food products derived from
wood, among others. As of 2017, innovative manufactured wood forest products
industry in Colombia is negligible, so the current (2017) installed capacity
of the CIPTPI is the sum of woodpulp and wood-based panels, i.e., 1,8 Mm3rsc/
year.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
13
Colombia exported only 3% of the mentioned demanded volume (Held
et al., 2017). It is also estimated that 7.3 Mm3rsc of equivalent wood for
manufactured wood primary products will be demanded annually in
Colombia by 2022, and this demand will increase to 9.2 Mm3rsc of
equivalent wood per year by 2030, and up to 10.6 Mm3rsc of equivalent
wood per year by 2038 (these estimations do not include an increase in
exports nor the potential domestic demand generated by the implementation
of the PFCm policy) (Martínez-Cort´es et al., 2018b).8
It seems that a decision of not increasing the initial area of commercial
forest plantations for wood production in Colombia, an economic
activity for which the country has comparative advantages
(Norton et al., 2008), would result in maintaining the negative trade
balance of Colombia’s wood forest products. This balance has been
negative for a long time, due to the continuous imports of pulp, paper
and paperboard, but it has exacerbated in the last two decades due to
substantial increases in imports of products for the pulp and paper industry,
wood-based panel industry and sawnwood industry. By 1996,
Colombia was already reporting a negative trade balance of about 325
million of American dollars (MUSD); this figure came from 391 MUSD in
imports and 66 MUSD in exports, which included the following product
net balances: roundwood (+0.2 MUSD), wood furniture ( 6.9 MUSD),
sawnwood ( 1.5 MUSD), wood-based panels ( 13.0 MUSD), and pulp,
paper and paperboard ( 303.5 MUSD) (Martínez-Cort´es, 1998). Based
on data from Profor (2017), in 2014, this trade balance reached 1044
MUSD (1044.6 MUSD in imports and 0.6 MUSD in exports, with the
trade balance being negative for all products in both monetary and
physical terms). If the trade balance is observed in physical units (volume),
it can be noted that, during the last two decades, Colombia has
had a negative trade balance in volume (i.e. volume of wood forest
products imported greater than volume of wood forest products exported)
and that this deficit has increased substantially in all wood forest
products but specially for panels as well as pulp, paper, and paperboard
products.
The results indicate that scenarios of increasing the 2015 initial area
to 0.765 Mha, 1.5 Mha and 2 Mha, and keeping these target areas under
sustainable rotation, are viable options. Under these scenarios, once the
total mass of forest plantations has reached, on average, the level of a
“normal forest”, the projected amounts, respectively, of 14 Mm3rsc/
year, 26 Mm3rsc/year, and 35 Mm3rsc/year of wood available would
allow for the supplying of the unprocessed wood market with national
wood, the expansion of the manufactured wood primary products industry,
meeting future national demand for these primary products, as
well as the development of the domestic and exports markets for the
manufactured wood products in general. An expansion of manufactured
wood primary products industry usually allows the flourishing of the
manufactured wood secondary products industry (also known as the
major value-added wood products industry); i.e. an industry whose
output is products such as: paper, and paperboard and packaging,
laminated wood, wood furniture, among others. It is also expected that
enough raw material (wood) would influence the development of
innovative manufactured wood products industry such as: forest bioenergy
products, chemical, food and cosmetic wood products, and bio
composites industries. The difference between the first expansion scenario
(S3) and the last two (S4 & S5) lies in the available time and the
magnitude of these endeavours.
Under the 0.765 Mha target (S3) scenario, projected to be reached by
2035, the manufactured wood primary products industry would have
enough raw material from the unprocessed wood market to conduct its
first expansion as early as in the period between 2023 and 2030, when
5.4 Mm3rsc/year of wood is available. However, this expansion would
be insufficient to meet the domestic demand for its products during that
period (i.e 7.3 Mm3rsc of equivalent wood/year). Only during
2031–2038 would the unprocessed wood market have a sufficient
amount of raw material (9.1 Mm3rsc/year), that would allow for the
expansion of said industry to such a size that would make it capable of
completely meeting the demand of the Colombian market for pulp,
boards, and sawnwood (i.e. 9.2 Mm3rsc of equivalent wood per year).
This implies that, up to 2030, Colombia would continue to increase its
imports, and that, up to 2038, all imports and national production of the
manufactured wood primary products industry would be exclusively
dedicated to meeting the domestic consumption needs.9 Only during the
2039–2047 period, when the production in the commercial forest
plantations for supplying the unprocessed wood market would reach 14
Mm3rsc/year (under a normal forest focus), would the manufactured
wood primary products industry be able to expand to the point where it
would be able to both meet the total demand of 10.6 Mm3rsc of equivalent
wood per year (mostly domestic consumption), and marginally
develop the exports markets for the same products.
Under the 1.5 Mha scenario (S4) (the policy goal of the National
Forest Development Plan of Colombia and the PFCm policy to be achieved
by Dec 2025), the final industrial expansion size would be 12
Mm3rsc/year, more than that under the 0.765 Mha scenario. Furthermore,
what happens in this latter scenario for the forest markets between
2039 and 2047 could happen 15 years sooner under the 1.5 Mha
one. For the 2023–2030 period, and then during the 2031–2038 period,
the unprocessed wood market could be supplied with 15 Mm3rsc/year
and 26 Mm3rsc/year, respectively; these amounts would allow for the
expansion of the manufactured wood primary products industry to
entirely meet the domestic consumption during such periods (i.e 7.3
Mm3rsc and 9.1 Mm3rsc of wood equivalent/year). Starting in
2023–2030, Colombia would have an excess of raw material for additional
expansions of the manufactured wood primary products industry,
to account for an expected additional demand to be generated due to the
implementation of the PFCm policy. This additional demand is expected
to be generated especially by fulfilling policy goals related to the
development of an innovative manufactured wood forest products industry
in Colombia (e.g. forest bioenergy, which in 2017 was in its infancy),
an increase of per-capita national consumption of wood and its
manufactured products (whose figures of 12 m3/per capita for 2015 are
deemed to be pretty low (Martínez-Cort´es et al., 2018b)), the recovery of
some wood forest products market segments that the manufactured
wood products industry has lost in the country’s economy (such as those
in the construction sector), and the development of the export markets
for Colombian wood forest products.
For the 2.0 Mha scenario (S5), the quantity of 35 Mm3rsc/year of
wood available for harvesting once the total mass of forest plantations is
considered a “Normal Forest”, represents an additional 9 Mm3rsc/year
potential expansion of the Colombian manufactured wood primary
products industry as compared to S4. The annual volume harvested
under S5 is comparable in order of magnitude to the annual harvest of
Chile, which, with an area of 2.4 Mha of forest plantations under sustained
rotation, harvested 43.6 Mm3rsc in 2015 (Gisling et al., 2017).
In terms of the cost, the average cost of per hectare (establishment
and management for the first four years without including the cost of
land) is USD1,420 (2017). Hence, the total costs of S3, S4, and S5 scenarios
will be 0.66, 1.7, and 2.4 billion USD (2017), respectively. The
socio-economic benefits of these scenarios, in terms of expansion of
forest industry to meet domestic demand, employment potential,
increased export, and contributions to trade balance, are already discussed
earlier. A deeper economic analysis, which is beyond the scope of
this paper, will provide a comprehensive picture of economic ranking of
these scenarios, and should be the subject of future research.
8 Demand figures are preliminary and were estimated by using an analysis
similar to a “gap model” during the redesigning of Colombia’s forest plantation
policy in 2016 and 2017.
9 Recall that demand figures were estimated under the assumption of a no
growth of exports, and in 2013, the base year for the demand estimation,
Colombia exported negligible quantities of wood forest products.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
14
7. Final thoughts: SCRPFC & the Colombian Forest Sector Model
(CFSM)
The apparent usefulness of the SCRPFC in analyzing and evaluating
policy goals was demonstrated in the preceding sections. Information
related to the available volumes for supply provides key elements in
understanding the interactions among some variables affecting the
supply and demand of wood forest products in Colombia: the extension
of the forest plantations, the supplying of wood to the unprocessed wood
market, the current and potential size of the manufactured wood products
industry, and the supplying of the markets of manufactured wood
products. However, the interactions among the above-mentioned variables
are far more complex than that discussed in the preceding sections.
These variables are not working in isolation, but they are affected by
other groups of variables from other parts of the forest sector, such as:
the wood available for supply from natural forests, agroforestry systems
and trees outside the forests; the behaviour of the economic agents in the
unprocessed wood market, which is influenced by what happens in the
manufactured wood products market; and by the behaviour and actions
of economic agents within each market, which are in turn influenced by
the performance of the Colombian and the global economies, to name a
few.
As such, the simulations presented in this paper are the first input
needed to estimate a more complete and powerful benchmark, useful in
conducting the analysis of the PFCm policy: the wood forest product
markets behaviour. The design of the SCRPFC and the volume estimation
exercises are part of the Phase I of the CFSM, a sectoral model
developed by Martínez-Cort´es (2022) as part of his doctoral thesis
research, in order to estimate the supply and demand for wood forest
products in the Colombian context, and estimate the potentially
achievable quantities and prices of wood forest products for the years
that follow.
These and other more complex interactions that happen within and
outside of the forest sector limits are what is simplified into core relations
within the CFSM. The forecasts generated by the CFSM would
provide benchmarks on the required quantities of wood traded in the
unprocessed wood market (a part of which would come from the wood
available for supply from Colombian commercial forest plantations),
and the prices at which the trades would happen, as well as the quantities
of manufactured wood products that economic agents would trade
in the market for these products together with the agreed upon price.
The CFSM analyses that addresses the question that how the expected
quantities and prices for wood and manufactured wood products in the
following years synthetize the expected behaviour of the (wood) forest
markets in the long run, will be presented in the future publications. The
findings of these publications will provide a better benchmark for
feedback and recommendations on the viability of alternative policy
goals, such as: the extension of Colombia’s forest plantation area, the
final size of the manufactured wood products industry, and the development
of the country’s unprocessed wood market as wells as the
manufactured wood products one.
Contributions of authors
Oscar Geovani Martinez-Cortes: Research conceptualization; data
collection and curation; data analysis and model building; programming,
model validation; visualization; and writing - original draft, and
revisions as per comments & editing.
Shashi Kant: Intellectual and theoretical inputs on research conceptualization;
data requirements and curation; data analysis and model
building, model validation; and different version of the manuscript.
Henrieta Isufllari: Support in data collection, policy analysis, preparation
of the first draft, editing, and incorporation of the reviewers’
suggestions.
Declaration of Competing Interest
Authors declare no competing interests.
Acknowledgements
This research was made possible with the financial and non-financial
support of the Graduate Department of Forestry, University of Toronto,
Unidad de Planificacion Rural Agropecuaria de Colombia (UPRA), and
the research grant of the 2nd author from Natural Science and Engineering
Research Council of Canada, [RGPIN-2019-05199].
The authors would also like to acknowledge the valuable inputs/
suggestions of the first author’s Ph.D. supervisory committee members
and William F. Hyde and the valuable support at UPRA received from
Felipe Fonseca, Daniel Aguilar y Alejandro Florez, and Luis Fernando
Parra, specifically in building the simulator and programming in R, and
the comments on the original draft and editing of the paper by Ximena
Laverde. The authors would like to express our appreciation for the
constructive comments and suggestions made by the two anonymous
reviewers. We also thankful to the Editor for his valuable suggestions
and the support during the review process.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.forpol.2022.102722.
References
Barreiro, S., Tom´e, M., 2017. Chapter 3. projection systems in Europe and North
America: Concepts and approaches. In: Barreiro, S., Schelhaas, M., McRoberts, R.E.,
K¨andler, G. (Eds.), Forest inventory-based projection systems for wood and biomass
availability. Springer International Publishing, Cham, pp. 25–47. https://doi.org/
10.1007/978-3-319-56201-8_3. Retrieved from. http://link.springer.com/10.1007/
978-3-319-56201-8_3.
Barreiro, S., Schelhaas, M.J., K¨andler, G., McRoberts, R.E., 2017a. Introduction. In:
Barreiro, S., Schelhaas, M.J., Mcroberts, R., K¨andler, G. (Eds.), Forest Inventory-based
Projection Systems for Wood and Biomass Availability. Managing Forest
Ecosystems, vol 29. Springer, Cham. https://doi.org/10.1007/978-3-319-56201-8_1.
Barreiro, S., Schelhaas, M., McRoberts, R.E., K¨andler, G., 2017b. Forest inventory based
projection systems for wood and biomass availability (Primera ed.). Springer
International Publishing, Cham, Switzerland. https://doi.org/10.1007/978-3-319-
56201-8. Retrieved from. https://basesbiblioteca.uexternado.edu.co:2224/book/10
.1007/978-3-319-56201-8.
Barrios, A., L´opez, A.M., Nieto, V., 2014. Predicci´on de volúmenes comerciales de
Eucalyptus grandis a trav´es de modelos de volumen total y de raz´on. Colombia
Forestal 17 (2), 137.
Büchner, C., Martin, M., Sagardia, Rodrigo, Rojas, Y., Bahamondez, C., Gui˜nez, R., 2019.
Disponibilidad madera en plantaciones de pino oreg´on. Regiones de la araucanía, los
ríos y los lagos (2019–2049) (Santiago de Chile).
Gisling, A., Alvarez, V., Soto, D., Pardo, E., Roger, T., Poblete, P., Bola˜nos, J., 2017.
Anuario Forestal 2016. Chilean Stastistical Yearbook of Forestry. Boletín estadístico
No. 154. Instituto Forestal INFOR, Santiago de Chile, Chile.
H¨ark¨onen, S., M¨akinen, A., Tokola, T., Rasinm¨aki, J., Kalliovirta, J., 2010. Evaluation of
forest growth simulators with NFI permanent sample plot data from Finland. For.
Ecol. Manag. 259 (3), 573–582. https://doi.org/10.1016/j.foreco.2009.11.015.
Held, C., Zapata, F., Casta˜no, C., Nutto, L., Paredes, A., Wittmann, N., Vargas, G., 2017.
Situaci´on actual y potenciales de fomento de plantaciones forestales con fines
comerciales en Colombia. Informe final del programa “Colombia: Reforestaci´on
Comercial Potencial del Banco Mundial / PROFOR (P148233)”. Banco Mundial,
Bogot´a.
Profor, 2017. Situaci´on actual y potenciales de fomento de plantaciones forestales con
fines comerciales en Colombia. Informe final del programa «Colombia: Reforestaci´on
Comercial Potencial delBanco Mundial / Profor (P148233)». Banco Mundial, Bogot´a,
172 pg. Recuperado de. https://www.Profor.info/sites/Profor.info/files/Informe%
20Final%20-%20Plantaciones%20Comerciales%20en%20Colombia_1.pdf>.
Instituto Forestal (INFOR), 2013. Disponibilidad de madera de plantaciones de pino
radiata y eucalipto (2010–2040). Santiago de Chile.
Jerez, M., Quintero, M., Quevedo, A., Moret, A., 2015. Simulador de crecimiento y
secuestro de carbono para plantaciones de teca en Venezuela: Una aplicaci´on en
SIMILE. Bosque (Valdivia, Chile) 36 (3), 519–530. https://doi.org/10.4067/S0717-
92002015000300018.
L´opez, A.M., Barrios, A., Ortega, A., Nieto, V.M., Gasca, G., Salamanca, M., 2007. Empleo
de un modelo de crecimiento y rendimiento para la determinaci´on de la edad ´optima
de rotaci´on de Pinus caribaea var. hondurensis creciendo en la regi´on oriental de
Colombia. Colombia Forestal 10 (20), 119–126.
´O.G. Martínez-Cort´es et al.
Forest Policy and Economics 138 (2022) 102722
15
L´opez, A.M., Barrios, A., Trincado, G., Nieto, V.M., 2011. Monitoreo y modelamiento del
crecimiento para el manejo de plantaciones forestales comerciales. Corporaci´on
Nacional de Investigaci´on Y Fomento Forestal-CONIF-Ministerio de Agricultura Y
Desarrollo Rural-MADR. Bogot´a, Colombia.
L´opez, A.M., Barrios, A., Trincado, G., 2015. Modelos de perfiles fustales con una
estructura de error autorregresiva para plantaciones de Eucalyptus tereticornis en
Colombia. Madera y Bosques 21 (2), 73–88.
MADR (Ministerio de Agricultura y Desarrollo Rural de Colombia), 2017a. Cadena
Forestal Indicadores e instrumentos. Presentaci´on del MADR. Junio de 2017, 18 pg.
https://sioc.minagricultura.gov.co/Forestal/.../002%20-%20Cifras%20Sectoriales%
20-.
MADR (Ministerio de Agricultura y Desarrollo Rural de Colombia), 2017b. Ministerio de
Agricultura y Desarrollo Rural- Reporte de Prensa. Recuperado el 25 de 02 de 2018,
de. https://www.minagricultura.gov.co/noticias/Paginas/historico-crecimiento-fore
stales.aspx.
MADR (Ministerio de Agricultura y Desarrollo Rural de Colombia), 2019. Decreto 2398
de 2019. https://vuf.minagricultura.gov.co/PublishingImages/DECRETO%202
398_27DIC19_movilizaci%C3%B3n%20plantaciones.pdf.
MADR (Ministerio de Agricultura y Desarrollo Rural de Colombia), 2021. Boletín
Estadístico Forestal. Septiembre de 2021, 69 p.
MADR(Ministerio de Agricultura y Desarrollo Rural de Colombia), 2019. Resoluci´on
número 000189 de 2019. 12 de junio de 2019. Por la cual se adoptan los
lineamientos de política para las plantaciones forestales con fines comerciales para la
obtenci´on de madera y su cadena productiva y su plan de acci´on 2018–2038 y se
dictan otras disposiciones. 3 pages and 2 Appendixes. https://www.minagricultura.
gov.co/Normatividad/Resoluciones/RESOLUCI%C3%93N%20NO.%20000189%
20DE%202019.pdf.
Martin, M., Büchner, C., Sagardia, R., Baham´ondez, C., Rojas, Y., Guzm´an, F.,
Barrales, Luis, Gui˜nez, Rodrigo, 2020. Disponibilidad de madera nativa en
renovables de roble-raulí-coihue. regiones de ˜nuble a los ríos 2020–2049. INFOR,
Santiago de Chile.
Martínez-Cort´es, ´O., Fl´orez, A., Fonseca, M., Garc´es, E., Gonz´alez, N., Guti´errez, ´E.,
Toro, ´A., 2018a. Plan de Acci´on para el desarrollo y consolidaci´on de la cadena
productiva de las plantaciones forestales con fines comerciales para la obtenci´on de
madera 2018–2038, 105 pag. UPRA, Bogot´a. Recuperado de. http://www.upra.gov.
co/documents/10184/13821/PLAN+DE+ACCI%C3%93N++PARA+EL+DESARRO
LLO+Y++CONSOLIDACI%C3%93N+DE+LA+CADENA++PRODUCTIVA+DE+LAS
+PLANTACIONES+FORESTALES+CON+FINES+COMERCIALES+PARA+LA+
OBTENCI%C3%93N+DE+MADERA++2018-2038/e81189fe-aa13-4b3e-be8e-084
fcf63ebc5.
Martínez-Cort´es, ´Oscar Geovani, 1998. Boletín SIEF - Sistema de Informaci´on Estadístico
Forestal. Proyecto Establecimiento e Implantacion de un Sistema de Informacion
Estadístico Forestal (A˜no 1 No 2). Bogot´a, Colombia. Ministerio del Medio Ambiente
- Organizaci´on Internacional de Maderas Tropicales. Recuperado el 2016, de. http
://www.itto.int/files/itto_project_db_input/2216/Promotional/BOLETIN%20SIEF%
20Octubre%20de%201998.pdf.
Martínez-Cort´es, ´Oscar Geovani, 2022. The Colombian Forest Sector Model CFSM. An
Analysis of Forest Plantation Policy in Colombia. Unpublished Doctoral Thesis for
PhD in Forestry. University of Toronto, Toronto, Canada.
Martínez-Cort´es, ´O., Castro, L., Fl´orez, A., Fonseca, M., Garc´es, E., Guti´errez, ´E., Toro, ´A.,
2018b. Plantaciones forestales con fines comerciales para la obtenci´on de madera y
su cadena productiva: lineamientos de política, 78 pgs. UPRA. Recuperado de,
Bogot´a. http://www.upra.gov.co/documents/10184/13821/PLANTACIONES+
FORESTALES+CON+FINES+COMERCIALES+PARA+LA+OBTENCI%C3%93N+DE
+MADERA+Y+SU+CADENA+PRODUCTIVA/051c6fbc-ae53-4bf6-8
e45-a0d64939c391>.
Melo, O.R., Lizarazo, I., 2017. Estimaci´on del volumen de madera en ´arboles mediante
polinomio único de ahusamiento. Colombia Forestal 20 (1), 55.
MMA, MADR, Mincomex, Mindesarrollo, DNP, & CONIF, 2000. Plan Nacional De
Desarrollo Forestal. Bogot´a, Colombia: Ministerio del Medio Ambiente, Ministerio de
Comercio Exterior, Ministerio de Desarrollo Economico, Departamento Nacional de
Planeacion, Ministerio de Agricultura y Desarrollo Rural.
Munro, D., 1974. Forest growth models–a prognosis. In: Fries, J. (Ed.), Growth Models
for Tree and Stand Simulation, 30, pp. 7–21.
Norton, R., Argüello, R., Samac´a, H., Martínez, H., Martínez-Cort´es, ´O., 2008. In:
Aldana, C. (Ed.), Una perspectiva de la competitividad forestal de Colombia (a
perspective of the forest competitiveness in Colombia, published in Spanish), Vol. II.
USAID y Ministerio de Agricultura, Bogota.
Restrepo Orozco, H.I., 2010. Estimaci´on del rendimiento forestal con informaci´on
espacialmente explícita y an´alisis financiero de plantaciones forestales en Colombia.
Retrieved from. https://www.openaire.eu/search?q=&Search=.
Salas, C., Gregoire, T.G., Craven, D.J., Gilabert, H., 2016. Modelaci´on del crecimiento de
bosques: estado del arte. Bosque (Valdivia, Chile) 37 (1), 3. https://doi.org/
10.4067/S0717-92002016000100001. Retrieved from. https://explore.openaire.
eu/search/publication?articleId=od______3056::c1cc5dd38cb7c46ec6a550b
02face28f.
Sepp¨anen, P., M¨akinen, A., 2020. Comprehensive yield model for plantation teak in
Panama. Silva Fennica (Helsinki, Finland : 1967) 54 (5). https://doi.org/10.14214/
sf.10309.
Serrano, M., Campos, J., Villalobos, R., Galloway, G., Herrera, B., 2008. Evaluaci´on y
planificaci´on del manejo forestal sostenible a escala de paisaje en hoja ancha, Costa
rica. San Jos´e, Costa Rica.
Twery, Mark J., 2004. Modelling in forest management. In: Wainwright, John,
Mulligan, Mark (Eds.), Environmental Modelling: Finding Simplicity in Complexity.
John Wiley & Sons, Ltd, London, pp. 291–301.
Vanclay, J.K., 2014. Forest growth and yield modeling (Primera ed.). John Wiley & Sons,
Ltd, Chichester, UK. https://doi.org/10.1002/9781118445112.stat07703. Retrieved
from. https://onlinelibrary.wiley.com/doi/abs/10.1002/9781118445112.stat0770
3.
Vargas, B., Corral, J., Aguirre, ´O., Nagel, J., 2016. Modelos de crecimiento de ´arbol
individual: Aplicaci´on del simulador BWINPro7. Madera Y Bosques 16 (4). https://
doi.org/10.21829/myb.2010.1641162.
Yin, Runsheng, Sedjo, Roger, 2001. Is this the age of intensive management? A study of
loblolly pine on Georgia’s Piedmont. J. For. 99, 10–17.
Yin, Runsheng, Pienaar, Leon, Aronow, Mary, 1998. The productivity and profitability of
Fiber farming. J. For. 96, 13–18.
´O.G. Martínez-Cort´es et al.
Fuente:
Biblioteca Virtual Banco de la República
Formatos de contenido:
Artículos
