|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
COMMUNITY-BASED TIMBER MANAGEMENT IN ACRE, BRAZIL, AND ITS
IMPLICATIONS FOR SUSTAINABLE FOREST MANAGEMENT
CARA A. ROCKWELL
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
Cara A. Rockwell
This thesis is dedicated to my friends and colleagues in PAE Porto Dias, who welcomed
me into their homes, and to my husband, Chris Baraloto, for coming this far in the
I would like to express my sincerest gratitude to my advisor and committee chair,
Karen Kainer, for all of her unfailing support. The past two and a half years have been a
truly enjoyable experience. My committee members, Marianne Schmink and Jack Putz,
have been excellent sources of encouragement and knowledge. I extend my full
appreciation to their involvement in the development of this thesis.
Thanks go to the School of Forest Resources and Conservation, the Tropical
Conservation and Development Program and Sigma Delta Epsilon/Graduate Women in
Science for the funding and support of this research.
The execution and completion of this project could not have been accomplished
without the participation of many key partners in Brazil. First, I must thank the families
of PAE Porto Dias, the Associagdo Seringueira Porto Dias, and the Projeto Manejo
Florestal Comunitario de Multiplo Uso for their patience as well as their kindness during
my stay in their community. I must particularly thank the following families for
welcoming us inside their homes; such hospitality will never be forgotten: the family of
Francisco Correia da Cunha and Raimunda Feitosa do Nascimento, the family of Adilton
Ferreira de Souza and Maria Alves Barboza, the family of Juscelino da Silva Correia and
Sebastiana do Santo Ven6ncio, and the family of Lazaro da Silva Salgueiro and
Marinalva Souza Amora Salgueiro. Also, a special obrigada goes to the following
community members for their participation in the study: Luzia Amora Salgueiro, Luziani
Amora Salgueiro, Juciclei Ven6ncio Correia, and Juciane Ven6ncio Correia, and thanks
go also to Luciana Roncoletta, who was conducting her own research in Porto Dias, but
still found time to assist with my project. Undoubtedly, the staff of the Centro dos
Trabalhadores da Amaz6nia (CTA) was there for me from beginning to end. They are
exemplary models of how collaborative partnerships can work. I especially would like to
thank my friends and colleagues, Nivea Marcondes, Magna Cunha dos Santos, Evandro
Araujo, Patricia Roth, and Pedro Bruzzi. I received valuable information and support
from the staff of the Federal University of Acre herbarium as well as faculty members
Marcos Silveira and Evandro Ferreira. Lucia Wadt from the Empresa Brasileira de
Pesquisa Agropecudria (EMBRAPA) provided support both in terms of research interest
and in friendship. And across the border in Bolivia, Marielos Pefia-Claros of the Instituto
Boliviano de Investigaci6n Forestal (IBIF) gave me great insight and suggestions
pertaining to the implementation of this study.
I never would have gotten through the crazy moments without the support from so
many fellow students, both in Gainesville and in Acre. If I miss any names, it is only
because there have been so many in the last couple of years. My warmest regards and
thanks go out to the girls of the Kainer lab: Diana Alvira, Jamie Cotta, Rosa Cossio
Solano, Amy Duchelle, Shoana Humphries, Christie Klimas, and Joanna Tucker. I thank
fellow RPCV Paraguayers who have seemed to find their way to University of Florida,
for all of their kind moments, advice, and plentiful terere: Mandy Baily, Maria DiGiano,
Tom Henshaw, and Katie Painter. The hospitality of Rich Wallace and Ana Puentes in
Acre will never be forgotten. I thank Samantha Stone for her introduction to Porto Dias
and CTA. She has set a high standard for all to follow. Other fellow SFRC students who
have given great support along the way and have always been willing listeners are: Julie
Clingerman, Robin Collins, Trina Hofreiter, Louise Loudermilk, George McCaskill, and
Other faculty and staff at the University of Florida have often provided much-
needed counsel and direction. Their efforts are duly appreciated: Cherie Arias, Hannah
Covert, Jon Dain, Eric Jokela, Ramon Littell, Christine Staudhammer, and Daniel Zarin.
I must not neglect to thank those who assisted both in terms of hospitality and
technical assistance during the writing process in French Guiana: Sabrina Coste, Paul
Arrive, Eric and Darouny Marcon, Francois Mornau, and Jean-Yves Goret.
I never could have gotten to this stage without the support and love of my parents
and my family. And there are no words sufficient to thank my husband, Chris Baraloto.
My appreciation will probably never make up for all of the cuts and stings suffered in the
tabocal forest of Acre while he acted as my botanist and trail cutter, but it has
nonetheless been an incredible journey. His love of the forest has been one of the greatest
educations in this process.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iv
LIST OF TABLES ....................................................... ............ .. ............ ix
LIST OF FIGURE S .................................................. ... ...... .............. ....
ABSTRACT .............. .................. .......... .............. xi
1 G EN ER A L IN TR O D U C TIO N ......................................................... .....................1
2 ECOLOGICAL LIMITATIONS OF REDUCED IMPACT LOGGING AT
COMMUNITY AND SMALLHOLDER SCALES ......... ......................................8
Introdu action ............... ... .................... ......... ..... ........................ ....... ..8
Sustainable Forest Management (SFM) and Reduced-Impact Logging (RIL)
M o d e ls .............................. .......... ..... ... .................... . ........... .1 1
Community/Smallholder Scale vs. Industrial Operations .......................................13
To What Extent Does the RIL Model Ensure SFM at the
Sm allholder/Com m unity Level? .......................................... ......... .... ............... 19
Pre-harvest Forest Inventories, Botanical Identification and Selection of Seed
Trees................ ...................................... 23
M inim um D iam eter Cuts ............................................................................. 27
Low Harvest Intensity ............. .... ................................... 30
Skid T rail P planning .............................. ........................ .. ........ .... ............32
R otation C ycles ............................................... ........ ...... 33
L ian a C u ttin g ...............................................................34
Logging Block Size and Implications for Pre-harvest Inventories ................ 37
M o n ito rin g ................................................................3 9
M moving beyond R IL ........................... ......... ............................. ............... 40
3 EVALUATING LOGGING DAMAGE TO FUTURE CROP TREES IN ACRE,
BRAZIL: IMPLICATIONS FOR MANAGEMENT OF A CERTIFIED
C O M M U N IT Y F O R E ST ................................................................ .....................44
Introduction .............. ..... .. .......... ......... ........... .... ..... ........ 44
Site D description ...................... ...... ...... ........................ ........ .. 47
Forest Management and Harvesting Operations...................................................49
M methods ...................................................... ................................5 1
Future Crop Tree (FCT) Selection ............................ ...... ...................... 51
Experimental Design and Pre-logging Inventory ..............................................52
A rea D istu rb ed ......................... .......... ....... ..... ........................... 52
FCT Location in Bamboo (Guadua)-dominated Forest............... ............... 53
Liana Presence ....................... ......... .......................... .... ....... 53
H harvest Intensity .................. ..................... ......... .......... .. ................53
M parking Treatm ent .................. ...................... ......................... 54
Assessment of FCT Damage ............ ........................................ ....................54
S statistical A n aly ses........... ...... ................................................ .. .... ...... 5 5
Results ...................... .................. ........ .........................55
Overview of the Logging O operation ............................... ................................. 55
F C T D a m a g e ................................................................................................. 5 6
A re a D istu rb e d ............................................................................................... 5 7
Discussion .............. ..................................... .........57
FCT Damage in Guadua-dominated Forest .................................... ..... 59
Impact of Lianas on FCT Damage ........................... ......................... 61
Impact of Harvest Intensity on FCT Damage..........................................62
Impact of Marking Treatment on FCT Damage................................ ..............63
C o n c lu sio n .................................................. .................. ................ 6 4
4 C O N C L U SIO N ......... ......................................................................... ........ .. ..... .. 78
R E F E R E N C E S ............................................................................ 8 1
B IO G R A PH IC A L SK E TCH ..................................................................... ..................94
LIST OF TABLES
2-1 Contrasting tree population parameters from inventories at two different spatial
scales comm only used by sm allholders.. ...................................... ...............42
2-2 An example of spatial aggregation among neotropical tree species harvested by
com m unities in A cre, B razil.......................................................... ............... 43
3-1 Commercial timber species inventoried during 2004 logging operations in the
PAE Porto Dias, Acre, Brazil. For those species harvested in 2004-2005, total
volum e (m 3 ha 1) is given ............................................................. .....................67
3-2 Classification of damage sustained by FCTs (future crop trees) during logging
operations (modified from Jackson et al. 2002 and Krueger 2004).........................69
3-3 Summary of harvesting impacts in 2004 for the three logging blocks at PAE
P orto D ias, A cre, B razil. ........................................ ............................................70
3-4 Results of a logistic regression for the probability of damage to future crop trees
(FCTs) at Porto Dias, Acre, Brazil during 2004 logging operations......................71
LIST OF FIGURES
3-1 Photograph taken in PAE Porto Dias, Acre, Brazil, demonstrating the re-curved
thorn on a branch of a Guadua culm ........................................... .. ........... .......... 72
3-2 PAE Porto Dias, with the four landholdings inventoried in the study, including
the three (in blue) harvested in 2004 ............................................. ............... 73
3-3 Logging block of Barrinha 1, showing skid trails, harvested trees, and location
of future crop trees (FCTs) in the zones of impact .............................................74
3-4 Logging block of Barrinha 3, showing skid trails, harvested trees, as well as
location of future crop trees (FCTs) in the zones of impact...............................75
3-5 Logging block of Palestina, showing skid trails, harvested trees, as well as
location of future crop trees (FCTs) in the zones of impact...............................76
3-6 Relationship between size of felled trees and the surface area of the
corresponding logging gap created during felling......................... .................77
Abstract of Thesis Presented to the Graduate School
of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Master of Science
COMMUNITY-BASED TIMBER MANAGEMENT IN ACRE, BRAZIL, AND ITS
IMPLICATIONS FOR SUSTAINABLE FOREST MANAGEMENT
Cara A. Rockwell
Chair: Karen Kainer
Major Department: Forest Resources and Conservation
In recent years, there has been much debate about whether conservation goals can
be achieved by logging old-growth tropical forests. Even so, the global demand for
tropical hardwood is not likely to subside anytime soon, making good management of
forests outside of parks and other protected areas essential. Some researchers have
suggested that rural communities may be the most ideal candidates to engage in
sustainable timber management, since they generally harvest at low intensities. Yet, it has
also been noted that the impact of natural resource exploitation at this scale is still not
well known, and that sustainable forest management (SFM) guidelines will necessarily
vary, depending on the location in which they are being implemented.
Field studies were conducted in Acre, Brazil to assess the impacts of the Porto Dias
community timber management project. The main objectives of this investigation were
(1) to determine the extent of area disturbed (treefall gap and skid trail construction) in a
small-scale certified logging operation, (2) to determine a difference in damage incidence
to future crop trees (FCTs) > 20cm dbh between forest with or without bamboo (Guadua
spp.), (3) to determine a difference in damage incidence to FCTs > 20cm dbh between
those associated with lianas and those that are not, (4) to determine to what extent
harvesting can be conducted more intensely (m3ha1), without incurring greater damage to
FCTs > 20cm dbh, and (5) to determine to what extent marking of FCTs diminishes
damage during logging. Related to this experimental study, an examination of key
reduced-impact logging (RIL) harvesting techniques and their potential influences on
population dynamics of tropical tree species in small scale timber operations was also
initiated, with the goal of addressing potential future management strategies for the
The results from the study indicate that although the effects of marking, liana
presence and location in Guadua-dominated forest were not significantly associated with
FCT damage, the high values reported for area damaged merit attention. Previous studies
have demonstrated that low intensity harvests are often associated with increased damage
levels per tree due to skid trails, suggesting that the gap mosaic of the forest will often
change, even when low-intensity RIL guidelines are followed. Given the series of poorly
understood parameters outlined in this thesis, it is recommended that systematic post-
harvest monitoring accompany logging operations, so that the benefits of recommended
practices may be evaluated by managers at the appropriate scale.
Deforestation is one of the greatest environmental challenges of the past few
decades (Food and Agricultural Organization (FAO) 2005). Brazil, which claims 63% of
the Amazon basin (Lentini et al. 2003), has long been at the center of the tropical
conservation debate. The Brazilian Amazon constitutes approximately one-third of the
world's extant tropical forests (Uhl et al. 1997). Since the 1960s, however, the country
has been plagued by destructive development policies that have opened up the region to
industrial logging, mining, ranching, and large-scale colonization projects, contributing to
a 14% forest cover loss to date (Nepstad et al. 2002). Compounding the problem of forest
cover loss, Brazil is simultaneously the world's largest producer and consumer of tropical
timber, most of which is generated through land clearing or unplanned selective logging
(Verissimo and Barreto 2004).
Selective logging is the harvesting method in which individual trees (rather than an
entire group) are removed. When conducted on a rotational basis, selective logging is
generally regarded to have a low impact on the forest (Uhl and Buschbacher 1988, but
see Sist et al. 2003 a,b, Asner et al. 2005). Unplanned selective logging, however,
typically excludes the use of a systematic timber inventory or management plan and often
harvests the same area over short, repeated intervals, frequently removing high volumes
of timber. The primary goal in this case is to extract the largest and most valuable
commercial species with little regard to surface area damaged or future harvests (Schulze
2003). In many instances, this has led to acute (up to 50% or more) canopy reduction,
severely disturbed mineral soils and damage to the residual stand (Uhl and Vieira 1989,
Verissimo et al. 1992, Johns et al. 1996). In response, scientists and practitioners alike
have proposed implementing sustainable forest management (SFM) practices that could
limit canopy cover loss and protect ecosystem function and biodiversity while
maintaining the economic value of the forest (Putz 1994, 1996, Pearce et al. 2003).
Definitions of SFM differ, but the term generally implies the sustenance of many forest
products and services over long periods of time (Pearce et al. 2003). Yet, despite the
recent conservation interest in SFM, there are many who would argue that to link
conservation with "sustainable forest management" is erroneous, and that the only way to
curb deforestation is outright protection (Rice et al. 1997, Bowles et al. 1998, Terborgh
Recently, there has been a massive undertaking in Brazil to reduce the negative
environmental impacts of regional silvicultural practices, much of which is generated by
large scale industrial operations. Forest management research in the region was first
initiated by FAO in the 1950s (Oliveira 2000), but it was not until the early 1990s that the
Brazilian federal government began advocating reduced-impact logging (RIL) techniques
as a means to combat rampant predatory logging in the Amazon basin (Holmes et al.
2002, Schulze 2003). RIL is an essential component of SFM if timber is to be harvested
from the forest (Pearce et al. 2003), typically employing a pre-harvest forest inventory,
road and skid trail planning, pre-harvest vine cutting, and directional felling (Dykstra and
Heinrich 1996). Indeed, in the absence of secure prices for non-timber forest products
(NTFPs) and infrastructural support for ecotourism efforts, selective logging under the
RIL model at low intensities may be one of the few assured income-generating activities
that can preserve the structure of the forest without compromising the entire ecosystem
The SFM model has generated considerable interest for community managers and
smallholders in the region, with the last few years witnessing the creation of a number of
innovative community-based conservation experiments founded on SFM principles
(Kainer et al. 2003, Stone 2003). Extractive reserves, or tracts of forest under community
control with secure usufruct rights, figure prominently among these community
development efforts. The remarkable and often violent history that preceded the
emergence of such initiatives as the extractive reserve (RESEX) program immediately
gave these forest communities widespread international support, but their conservation
goals have since become the focus of intense scrutiny (Nepstad et al. 1992, Redford and
Stearman 1993, Terborgh 1999, 2000). Many researchers, although generally not
contesting the social justice value that such endeavors provide to a historically
disenfranchised population, have questioned the wisdom in holding up the extractive
reserves as models for biodiversity conservation (Redford and Stearman 1993, Terborgh
2000). This issue has become especially pertinent now that many community forests have
shifted from local economies based on non-timber forest products (NTFPs) to that of
timber extraction (Kainer et al. 2003, Schmink 2004). The general consensus among
some researchers is that external economic and cultural pressures, in concert with a
tendency to abuse the resource commons (see Hardin 1968), will hinder the ability of
community forest managers to protect biodiversity (Redford and Stearman 1993,
In Brazil, though, voluntary forest certification has provided a strong incentive for
both community forest managers and commercial logging companies to adopt SFM
practices (May 2002, Kainer et al. 2003). As of April 2004, Brazil ranked fourth in the
world in number of certified forests and eighth in area (slightly under 1.6 million ha,
529,079 ha of which are considered natural forest) (May 2002). The failure of
government policies to curb illegal logging gave birth in the 1990s to the certification
instrument, which endorses forest products originating from sustainable forestry
operations, a process often executed by a third-party organization (Molnar 2003). Forest
certification's role as a community-based conservation mechanism has received
widespread support (Molnar 2003). This has been particularly true of the Forest
Stewardship Council (FSC), an international organization that has had a strong influence
in the development of certification standards in the Brazilian Amazon (Humphries 2005).
Ultimately, the certification process could provide access not only to the global certified
forest product market but also to silvicultural evaluation and assistance, thereby creating
motivation to conserve the resource base (Putz 1996, Putz and Viana 1996, Putz and
Despite their positive associations with the conservation movement, a number of
researchers have criticized forest certification bodies and biologists for producing SFM
guidelines and criteria that fail to recognize the problems that such a system might create
for smallholders (Pretty and Hine 1999, Molnar 2003). Even though community timber
management operators may have the knowledge, "human nature and economics" can
offset the desire to sustainably manage the forest (Salafsky et al. 1998). Smallholders and
community forest managers in the rural tropics have limited financial and technical
resources, so implementing additional silvicultural treatments and monitoring may be
unfeasible for them, unless the proposed activities are seen as adding to their
productivity. Furthermore, given intense competition from less sustainable commercial
industrial timber enterprises and illegal logging, the long-term impact that certification
will have on the industry and deforestation in general is questionable, particularly in
Latin America (Sierra 2001). In Brazil, for example, certified wood is currently meeting
only 2% of the annual demand for logs (Lentini et al. 2003).
Community and smallholder forest enterprises are generally thought to be more
environmentally benign than their industrial counterparts (Salafsky et al. 1998), making
them seemingly ideal candidates to implement the SFM and RIL models. There are,
however, a number of inherent ecological limitations associated with small-scale timber
operations that are rarely addressed in the management literature, with few ecological
investigations regarding the impact of timber extraction having been conducted at the
level of the community or smallholder level (Oliveira 2000, Walters 2005). Although RIL
is, in general, ecologically beneficial to the residual stand, there is evidence that the
magnitude of those benefits varies, depending on the forest being exploited (Fredericksen
and Putz 2003, Schulze 2003) and the specific RIL techniques being used. RIL was
initially developed for large scale industrial operations where harvest intensities and
investment potential are high (Oliveira et al. 1998). Nevertheless, community forest
managers have eagerly adopted RIL standards, even though the ecological impacts of
RIL at this scale are not well-known. Given that these impacts are poorly understood,
there is a need to explore both the limitations and the benefits of RIL for these
community and smallholder-based systems rather than embracing what is held to be the
ideal by researchers working with larger operations.
In this thesis I examine both RIL guidelines as generally implemented at the
community and smallholder levels as well as a specific case study in PAE Porto Dias,
Acre, Brazil, testing the harvesting impacts generated by the activities of the "Multiple-
Use Community Forest Management Project" on the residual stand. This approach
potentially addresses some of the challenges of community timber operations in Acre as
well as examining future crop tree (FCT) damage as sustained during RIL operations.
In Chapter Two, I examine some of the ecological constraints associated with
timber operations at the community and smallholder scales. To do so, I evaluate several
assumptions of selected RIL recommendations and their implications for smallholder
management. After a brief introduction to SFM and RIL, particularly as they relate to the
Brazilian Amazon, I discuss eleven key differences between industrial and
community/smallholder operations. I then ask to what extent the RIL model contributes
to SFM at the smallholder/community level, with an examination of RIL harvesting
techniques and their potential influences on the population dynamics of tropical tree
species in smallholder systems, with a focus on the Brazilian Amazon. I conclude the
paper with a discussion of potential future management strategies for smallholders.
In Chapter Three, I present research investigating the impacts that forest type
(bamboo/non-bamboo), liana presence, harvesting intensity, and marking have on FCT
damage during logging operations in PAE Porto Dias. The study focuses on the adult and
sub-adult size classes (> 20 cm dbh) of commercial timber species in the hopes of
capturing information about the availability of timber at the time of the next harvest (30
This thesis has been organized such that the second and third chapters are two
individually and fully structured papers for publication. Each one of these chapters has its
own conclusions, while Chapter 4 summarizes the main findings of the entire study and
provides a discussion of the future directions for small-scale timber management
ECOLOGICAL LIMITATIONS OF REDUCED IMPACT LOGGING AT
COMMUNITY AND SMALLHOLDER SCALES
Researchers and practitioners alike have proposed that large tracts of tropical
forests will not be conserved without the involvement of local forest residents (Scherr et
al. 2002, Shanley and Gaia 2002). Already, approximately one quarter of the world's
forests are community-owned or managed, a proportion that is expected to double in the
next 15 years (Molnar 2003). Some have advised that these forests will stand a greater
chance in surviving when placed under the jurisdiction of communities, assuming they
will have a desire to conserve their forest resource base (Camino and Alfaros 2000). In
concert with a growing smallholder interest in formal timber production and a need for
sustainable forest management (SFM) practices outside of protected areas (Putz et al.
2000), it is likely that community and smallholder forests will play an increasingly
important role in global conservation.
For various reasons, including an international market demand for sustainably
harvested timber products, many of these community and smallholder operations are
becoming third-party certified (Molnar 2003). This voluntary process emphasizes
rigorous social and environmental standards for production forests, potentially providing
a strong economic incentive for local forest managers to value the resource base (Putz
1996, Putz and Viana 1996). The reduced impact logging (RIL) management models
favored by the certification instrument, however, have failed to develop situation-
appropriate strategies that accommodate the ecological and social complexities of small-
scale commercial timber extraction. Instead, the tendency has been to impose
management practices of large commercial operations onto smallholder systems (Oliveira
2000), causing some forest operations to take on corporate features usually relegated to
industrial timber enterprises (Putz 2004). It has been suggested that further research and
operational training are needed to evaluate how variations in forest type and size of
logging operations affect the performance of RIL systems (Holmes et al. 2002).
Without much empirical information suggesting otherwise, it is often assumed that
because many smallholders harvest at low intensities and utilize animal traction (Oliveira
2000) or small machines, they are more environmentally benign than industrial logging
enterprises (Salafsky et al. 1998). Nevertheless, there are a number of inherent ecological
limitations associated with small-scale timber operations and low intensity harvests,
which have been shown to have substantial impacts on forest structure, composition and
regeneration (Smiet 1992, Gullison and Hardner 1993, Boot and Gullison 1995, Whitman
et al. 1997, Awasthi et al. 2003, Walters 2005). Community and smallholder-owned
forests generally encompass smaller areas compared to their industrial counterparts (<
500 ha) (Lentini et al. 2003). When managing for timber in small, non-contiguous
management units, ecological constraints to regeneration, such as insufficient pollen
transfer and/or inbreeding (Ghazoul et al. 1998), dispersal limitation (Dalling et al. 2002),
and predator satiation (Schupp 1992) can become exacerbated, as population sizes of all
species are correspondingly small (Putz et al. 2000).
RIL's emphasis on environmental sustainability is often assumed to be a natural fit
for rural forest dwellers (Braz and Oliveira 1996). However, it is becoming increasingly
clear that the shift from traditional management systems to that of third-party certified
commercial timber extraction is challenging. Smallholders and communities are typically
required to implement RIL techniques to meet certification standards, compelling them to
invest in both equipment and human capital not normally required for their daily
subsistence needs (Stone 2003).
Although the RIL model has been demonstrated to have many ecological benefits
in tropical forests, there is a need to explore both the limitations and the benefits of RIL
specifically for smallholder systems. In this paper, I evaluate key assumptions of several
RIL recommendations and their implications for smallholder management. I use the
terms smallholderr" and "community" interchangeably throughout, though there is a
distinction (albeit broad) between smallholders living on individual lots in colonization
projects and traditional or indigenous forest-dwelling communities, such as those living
in or near forests under some form of protection. As well, when referring to the terms
"community" and smallholderr" management in this paper, I am specifically addressing
locally-run operations, not to be confused with commercial logging concessions on
smallholder properties (e.g. Nepstad et al. 2004). After a brief introduction to SFM and
RIL, I discuss eleven key differences between industrial and community/smallholder
operations. I then ask to what extent the RIL model ensures SFM at the
smallholder/community level, with an examination of RIL harvesting techniques and
their potential influences on the population dynamics of tropical tree species in small
scale systems, with a focus on the Brazilian Amazon. I conclude the paper with a
discussion of potential future management strategies for the smallholder system.
Sustainable Forest Management (SFM) and Reduced-Impact Logging (RIL) Models
Selective logging when applied on a long rotational basis and at low harvest
intensities is generally regarded to have a low impact on the residual stand (Uhl and
Buschbacher 1988, but see Sist et al. 2003a,b, Asner et al. 2005). Selective logging,
however, can be one of the first steps in forest conversion, providing a subsidy for
pasture formation as in the Brazilian Amazon case (Oliveira 2000). Additionally, very
few commercial timber operations in this region employ sustainable selective logging
techniques, with an estimated 95% of all timber extracted through unplanned
conventional logging (Verissimo et al. 2002).
The extent to which selective logging causes permanent damage can be controlled
in some ways by planned management activities, as has been suggested by the SFM
model. SFM can be defined based on the following assumptions: (1) management can be
exercised in a manner compatible with the maintenance of biodiversity, (2) management
of tropical forests is economically viable, and (3) management can result in sustained
timber yield over the long term (Bawa and Seidler 1998). Translating these concepts into
practice, Bruenig (1996) maintains that when discussing timber extraction, SFM should
encompass the following: "canopy openings should be kept within the limits of natural
gap formation; stand and soil damage must be minimized; felling cycles must be
sufficiently long and tree marking so designed that a selection forestry canopy structure
and a self regulating stand table are maintained without, or with very little, silvicultural
manipulation; (and) production of timber should aim for high quality and versatility".
Bawa and Seidler (1998) conclude, however, that too often, SFM is promoted without
rigorous examination of the interrelatedness between biodiversity maintenance, economic
viability, and long-term yields, causing some researchers to doubt the validity of SFM
and calling for the outright protection of forested regions instead (see Rice et al. 1997,
Bowles et al. 1998, Terborgh 1999). As well, some researchers have questioned whether
timber harvests in at least some forests can be sustained without silvicultural intervention
(Daube et al. 2005). Nevertheless, the SFM philosophy has been embraced by multiple
forest actors. In Brazil, for example, the federal government, state institutions, NGOs and
research organizations have actively promoted SFM, hoping that it will quell unchecked
land conversion and predatory logging activities (Schwartzman et al. 2000, Nepstad et al.
2002, Fearnside 2003, Kainer et al. 2003).
RIL is considered a necessary component to achieve the goals of SFM, though its
implementation alone cannot maintain biodiversity and ecosystem functions nor even
secure sustained timber yield (referring to the objective of maintaining timber production
for future harvests) (Webb 1997, Putz et al. 2000, Sist et al. 2003b). RIL implies a series
of pre- and post-logging guidelines (see Pinard et al. 1995, Dykstra and Heinrich 1996)
which are designed to protect advanced regeneration (i.e., seedlings, saplings and larger
future crop trees), minimize soil damage, limit damage to non-target populations
(i.e.wildlife and non-timber plants species), protect water quality, and preserve ecosystem
processes (i.e. carbon sequestration) (Putz et al. 2000). Operations under these guidelines
typically employ long-term cutting cycles, pre-harvest forest inventories, road and skid
trail planning, pre-harvest vine cutting, and directional felling (Dykstra and Heinrich
1996). RIL has been shown to reduce damage to the residual stand by 30-50% and has
also demonstrated an increased efficiency of logging operations (see Johns et al. 1996,
Pinard and Putz 1996, Bertault and Sist 1997, Barreto et al. 1998, Sist et al. 1998, Holmes
et al. 2002).
Under pressure from both domestic and international concerns to improve the
ecological sustainability of the tropical hardwood industry, Brazil began to adopt RIL
techniques in the early 1990s, following the model that was originally developed in
Southeast Asia and Suriname a decade earlier. Since that time, many Brazilian timber
companies have agreed to abide by RIL regulations at the behest of federal government
regulations, and in some cases, third-party forest certification standards, hoping to tap a
high-end niche market (Schulze 2003).
Community/Smallholder Scale vs. Industrial Operations
The SFM model (and implicitly, RIL) has generated considerable interest for
community forest managers and smallholders (Braz and Oliveria 1996). Researchers
often point to local peoples' "long-term commitment to place" and their intimate
knowledge of ecosystems (Schmink 2004) as evidence of their ability to both manage and
(when necessary) restrict access to forest resources (Agrawal and Gibson 1999). Yet, it
has also been noted that the implementation of a systematic management regime can
often be overwhelming for communities and smallholders, requiring significantly
different conditions for long-term success than large-scale commercial enterprises
(Schmink 2004). Indeed, to mitigate the challenges faced by these operations, some
suggest that community and smallholder forest management systems must be considered
within case-specific ecological, socio-economic, political, and cultural settings (Berkes
and Folke 2001, Schmink 2004). This viewpoint recognizes the unique character of
individual community and smallholder systems and also further emphasizes the need to
distinguish these enterprises from their industrial counterparts.
Several general operational characteristics distinguish community and smallholder
timber enterprises from the large-scale commercial timber industry, building on a list of
five proposed by Salafsky et al. (1998): (1) Ownership: Industrial operations usually buy
logging rights via a concession, or the timber company may own the land. More than
likely, the proprietors live offsite. In contrast, community and smallholder timber
harvesting rights are owned or, at the very least, held by local inhabitants (Salafsky et al.
1998). For example, in the Brazilian Amazon, one third of the land area is inhabited by
smallholder and indigenous communities with the rights to manage their own forests
(Shanley and Gaia 2002). Due to a strong sense of ownership and multigenerational
responsibilities, community forest managers may implement more sustainable
management practices than their industrial counterparts (Walters et al. 2005). Indeed, in
those community cases where secure land tenure is lacking, there appear to be fewer
incentives to manage the resource sustainably (Banana and Bombya-Ssembajjwe 2000,
Gibson and Becker 2000). (2) Timber as integrated livelihood component: Community
and smallholder forest managers rely on the forest for a suite of goods and services,
viewing timber management as one component of an integrated system rather than a sole
income-generating activity (Salafsky et al. 1998). Since timber harvesting is a seasonal
activity (usually during the driest months), local managers pursue other livelihood
activities throughout the remainder of the year (Oliveira 2000). Even so, timber
management and other livelihood activities are not always compatible, especially when
harvesting is conducted by those from outside the community. In one Eastern Amazonian
case, community members from the Tapaj6s-Arapiuns Extractive Reserve believed that
commercial logging negatively impacted NTFP collection (due to incidental damage) and
hunting (through easier game access by logging roads) (Menton 2003). Researchers also
have reported a lack of documented systems explicitly integrating timber and NTFPs in
management plans (Bawa and Seidler 1998, but see Salick et al. 1995, Romero 1999).
Local forest managers who wish to protect a multitude of forest services (NTFPs, game
species, watershed) may be more inclined to adopt RIL guidelines geared to protect
components of the forest beyond future timber harvests. (3) Scale: Community and
smallholder enterprises generally operate on a smaller scale than their industrial
counterparts, the latter tending to rely on capital intensive techniques and heavy
machinery rather than local labor (Salafsky et al. 1998). Many techniques developed
specifically for communities have been initiated in secondary forests, where the use of
animal traction is ideal for the extraction of small diameter logs (Pinedo-Vasquez et al.
2001). Smallholder logging in primary forests has become more common in recent years,
particularly in the Brazilian state of Acre (see Kainer et al. 2003, Stone 2003), but the
need for heavy machinery in these systems and the subsequent competition with larger
enterprises has constrained community participation in this type of extraction (Oliveira
2000). However, smaller scale operations may facilitate more intensive management for a
suite of services and products, something that might be impractical in a larger
management unit. Also, in Brazilian community timber efforts, harvest intensities tend to
be low, usually no more than 10 m3 (or two trees) ha-1 (Oliveira et al. 1998), which (in
theory) should be compatible with RIL guidelines. In comparison, most Brazilian
conventional forest management systems harvest anywhere from 30 to 40 m3 ha- yr-
(Oliveira 2000). (4) Added-value: Industrial logging operations harvest roundwood logs
that are then transported to a centralized offsite-processing mill. Conversely, community
and smallholder operations often seek to add value to the lumber onsite, transforming the
raw material into finished or semi-finished products (Salafsky et al. 1998). This
potentially provides greater income to local residents, such as the case of at least one
certified community operations in Acre, Brazil that is producing artisan objects from
locally harvested wood (pers. obs.). (5) Local reinvestment: In community-based
systems, capital is generally reinvested locally and there is a greater long-term incentive
for sustainability (Salafsky et al. 1998). In comparison, industrial timber enterprises tend
to move profits to other localities or sectors of the economy, thereby reducing any
potential social benefits from forestry activities (Salafsky et al. 1998). For example, in
Oaxaca, Mexico, Bray et al. (2003) note community use of logging profits to encourage
more benign forest enterprises (i.e., water bottling, ecotourism, and resin-tapping) and
build more community assets (i.e., potable-water networks, schools, and clinics). And
even though timber extraction has subsidized cattle pasture expansion in the Brazilian
Amazon (Oliveira 2000), Summers et al. (2004) found that smallholder timber managers
in Rond6nia tended to clear less forest and had smaller cattle herds than their neighbors.
(6) Adaptive management/traditional governance: Communities and smallholders
sometimes develop adaptive management systems that permit local users to identify
internal and external threats to the forest (Margoluis and Salafsky 1998, Salafsky et al.
1998). Some researchers have recognized this as a benefit of traditional governance, or
locally-defined institutions and regulations that guide decision making processes (Banana
and Gombya-Ssembajjwe 2000, Becker and Le6n 2000, Bray et al. 2003). For instance,
Becker and Le6n (2000) found that the Yuracare of the Bolivian Amazon have clear,
locally-defined rules regarding forest resources, upon which they can control both access
and use of the forest at multiple scales. (7) Value of forest to future generations:
Intergenerational values or investment in the forest for future generations also distinguish
community timber enterprises from their industrial counterparts, a factor often
overlooked (see Bowles et al. 1998) when discussing excessive production costs of
community forests (Bray et al. 2003). In several studies conducted in the Brazilian
Amazon, rural respondents identified silvicultural activities as positively impacting the
future of their children and grandchildren (Walters et al. 2005). (8) Traditional
ecological knowledge: Rural peoples who have often lived in the same geographic
region for generations have accumulated considerable knowledge about the natural
resource base through daily observations in their subsistence activities, usually
demonstrating a comprehensive understanding of ecosystem linkages (Folke et al.1998,
Berkes 1999, Calheiros et al. 2000, Berkes and Folke 2002). For example, it has been
noted that despite the highly technical silvicultural expertise available in Finland, the
main factor explaining the excellent forest health in that country is that conservation
practices have been based on TEK passed from one generation of forest managers to
another (Banuri and Apffel-Marglin 1993). Accordingly, TEK can be applied to a
systematic forest resource management plan, providing details on timber quantity and
location, regeneration requirements, and other requisites for successful implementation of
RIL guidelines. (9) Commitment to forest conservation: Mexican communities in
Oaxaca and Quintana Roo consistently logged below the authorized volume (some by as
much as 37%), suggesting a willingness to reduce the volume of extraction when
inventories indicate external threats to the resource base. Some of these same
communities also prohibit or limit hunting in local forests, displaying an increasing
commitment to biodiversity conservation (Bray et al. 2003) or perhaps, a clear
understanding of sustainable game harvest limits. Other forest managers in Acre, Brazil
have developed local indicators for advanced regeneration as determinants for harvest.
By identifying a sufficient number of what they term as "daughters" and
"granddaughters" of the tree(s) to be harvested, they acknowledge the importance of
leaving several individuals of differing size classes of a particular species in the
management unit, not just reproductive adults (P. Roth, pers. comm.). (10) Participatory
monitoring: Local participants, especially those termed "parabiologists" or
"parataxonomists", have been instrumental in monitoring environmental impacts and
planning formal conservation initiatives, and, in some cases, have actually improved the
data quality of the academically-trained researcher (Janzen 2004, Sheil and Lawrence
2004). In one Brazilian case, local participants monitored the populations and phenology
of selected fruit tree species in collaboration with professional biologists, eventually
determining that the environmental impact of timber extraction was too great a risk to the
long-term NTFP value of a particular fruit species (Shanley and Gaia 2000). These
examples suggest that local forest dwellers are very aware of forest dynamics and capable
of executing environmental monitoring (especially those post-logging surveys
encouraged by the RIL paradigm) when provided with some capacity building and
equipment (Sheil and Lawrence 2004). (11) Influence/funding from outside
organizations: Particularly since the early 1990s, a network of community forests in the
Brazilian Amazon have received financial and infrastructural support of a number of
NGOs and governmental institutions (Stone 2003). This support has led to the voluntary
third-party certification of at least 7 community operations in the region, fortifying the
competitiveness of community forest management in a demanding market dominated by
industrial giants. Indeed, two of the principal barriers faced by community and
smallholder managers are the small scale production and consequent marketing
difficulties, limitations that certification may be able to improve, primarily through
access to a niche market (May 2002). Yet, Stone (2003) also warns of the increasing
reliance of certified community timber projects on outside funding and technical
assistance, mostly because of the complex requirements inherent with RIL adoption, such
as investment in expensive equipment and the writing of management plans.
To What Extent Does the RIL Model Ensure SFM
at the Smallholder/Community Level?
Prior to the recent emphasis on the RIL model, traditional smallholder forest
exploitation in the Brazilian Amazon was based on low technology methods that
produced minimal financial returns, but also resulted in low environmental impacts
(Oliveira 2000). Indigenous inhabitants have been managing forests in the region since
pre-Columbian times (see Heckenberger et al. 2003) for a suite of subsistence products
and services (Summers et al. 2004), a pattern that has continued to the present. Rural
Amazonian inhabitants typically integrate wood production into their existing swidden-
fallow systems, resulting in a highly patchy landscape and allowing smallholders to
maintain favorable environments for several different species (Pinedo-Vasquez et al.
2001). It was not until the early 1990s, amidst a deepening international concern for
global deforestation and the creation of the Program for the Conservation of the Brazilian
Rain Forest (PPG7) in 1992, that the formal establishment of SFM projects for
community and smallholder forests emerged (Stone 2003), eventually leading to the first
RIL project for privately-owned smallholder forests in the state of Para in 1993 (Amaral
and Neto 2000). By the mid 1990s, the Brazilian federal government also began to take
an interest in community-based SFM models and in 1998, the Brazilian Institute of the
Environment and Renewable Resources (IBAMA) (the federal agency directly
responsible for the execution of federal forest policy) implemented the PMFSimples
(Simple Sustainable Forest Management Plan) protocol, allowing smallholders to manage
up to 500 ha for timber (Stone 2003). It also enabled community forest initiatives to
submit one management plan for all participating properties, rather than do this
individually. Since 1995, at least 7 smallholder and community forest projects have been
established in the region, under a wide range of property regimes extractivee reserves,
national and state forests, indigenous reserves, and agricultural colonization projects)
(Stone 2003). Community and smallholder interest in timber has been motivated in great
part by a need to increase and diversify livelihoods, fueled in particular by low net returns
for NTFPs (Kainer et al. 2003).
However, as the sophistication and intensity of community and smallholder forest
management systems increases, it has also become evident that silvicultural prescriptions
appropriate to local conditions are needed, especially those that take into account a
shortage of investment and human labor (Oliveira 2000). The Brazilian RIL model was
initially developed for large scale industrial operations where investment potential is high
(Oliveira et al. 1998). For example, since 1995, training in RIL techniques for Brazilian
operations has been a primary focus of the Tropical Forest Foundation (TFF) and its
Brazilian subsidiary, Fundacgo Floresta Tropical (FFT) (Holmes et al. 2002). This effort
has resulted in the development of several model systems throughout the Amazon basin,
but most of these examples are considered large-scale operations. Not surprisingly, these
types of models have been eagerly adopted by community forest managers, even though
the RIL ecological impacts on the overall forest structure at this scale are not well-known
(Oliveira et al. 1998).
In its original conception, one of the basic tenets of RIL mandated that logging
should mimic the so-called gap-phase dynamics of the neotropical rainforest (see Brokaw
1982) by creating small-scale disturbances around the extracted crop tree (Hartshorn
1989). Natural small-scale disturbances (usually involving the death and subsequent fall
of a neighboring dominant tree) permit access to limiting light (and other) resources. By
attempting to mimic this type of disturbance during RIL operations, it is hoped that
residual stand damage will decrease and that growth rates of surviving trees will increase.
Natural disturbances, however, are qualitatively and quantitatively different from those
logging-induced disturbances (Boot and Gullison 1995, Vitt et al. 1998, Dickinson et al.
2000). Boot and Gullison (1995) note that mortality rates of trees in forests with only
natural disturbances are on the order of 1-2% (see Swaine et al. 1987), while
conventional selective timber harvests can damage as much as 55% of the residual stems.
Additionally, they point to other sources of residual stand degradation, such as soil
compaction, erosion, and changes in drainage, that are not part of a natural small-scale
Although RIL techniques are designed to minimize this type of aggravated impact,
some species of commercially-valuable trees require larger disturbances for reproduction
than is mandated by minimal impact logging techniques, which usually emphasize low
harvest intensities (Fredericksen and Putz 2003). Careful planning may maintain timber
volume but the removal of scarce, high value logs could result in replacement by less
important species, ultimately reducing forest biodiversity and dooming the second cutting
cycle to timber not yet recognized by the market (Schulze 2003). In this case, the gap-
phase management paradigm could be in direct opposition to the ecological role of
catastrophic natural and anthropogenic disturbances in forests, conditions to which
particular species have adapted over time (Deneven 1992, Bush and Colinvaux 1994,
Fredericksen and Putz 2003). Fredericksen and Putz (2003) point to examples of
neotropical trees that may not benefit from a low-impact management regime, such as
Swietenia macrophylla, Bertholletia excelsa and Cedrela spp., species which appear to
establish in large disturbed areas.
Another of the underlying assumptions of many RIL guidelines is that advanced
regeneration of valuable commercial timber species is abundant and exists uniformly
throughout the stand (Schulze 2003). However, given the suite of differing growth rates,
wood densities, population dynamics, and modes of seed dispersal, both between and
within commercial species in the Amazon basin (Schulze 2003), it may be that
knowledge of tree autecology is one of the greatest limiting factors in implementing SFM
(Hammond et al. 1996, Guariguata and Pinard 1998). Currently, there is a lack of
sufficient ecological information to help guide forest management decisions in the
Brazilian Amazon, which is one of the reasons forestry methods have not advanced much
beyond RIL (Schulze 2003). Yet, it is prudent to evaluate small-scale, RIL-based logging
operations to date, and discuss how some of the basic tenets of RIL (pre-harvest forest
inventories, selection of harvest and seed trees, minimum diameter cuts, low harvest
intensity, skid trail planning, rotation cycles, and liana cutting) have worked at the
smallholder level and how they can be modified.
Pre-harvest Forest Inventories, Botanical Identification, and Selection of Seed Trees
A full pre-harvest inventory provides managers with detailed information on
distribution, quantity, and size of potential harvest trees and future crop trees (Schulze
2003). Unfortunately, for many small operations (as well as some large ones), 100%
inventories are cost prohibitive, especially when the entire management unit (as is the
case of those community management projects in the extractive settlement projects and
extractive reserves in Amazonia) is as large as 300 ha. In at least one study in Quintana
Roo, Mexico, local rural people identified time and finances as two of the biggest
constraints for the inventory process (Lawrence and Roman 1996).
In response, it has been common to inventory only the logging block for the following
year, as the inventories are generally carried out one year prior to harvesting activities.
Although this may eliminate the need to pay for a costly comprehensive inventory, it
limits the manager's ability to harvest with respect to species population biology (see
Table 2-1). For example, in a 10 ha logging block chosen from a 300 ha landholding,
several individuals of a particularly valuable species are identified during the 10 ha
inventory. For management purposes, these individuals are then the sole representatives
of that species within the 300 ha landholding, since a full inventory of adult trees has not
been carried out. Many commercial tropical tree species are characterized by highly
clumped spatial distributions (Hubbell 1979; Table 2-2) and are self-incompatible (Bawa
et al. 1985), so harvesting even half of the inventoried individuals could potentially lead
to inbreeding and genetic drift (Jennings et al. 2001).
Ghazoul et al. (1998) demonstrated that the reproductive success of ,\l/,i',t
siamensis was limited by compatible pollen transfer at a reduced population density after
logging. This Allee effect (see Allee 1949) can increase the probability of extinction,
particularly in small populations. For those species occurring in widely spaced clumps, a
reduced density in aggregations may result in a decrease in pollen exchange between
clumps, severely limiting genetic variability within aggregations and thus, their ability to
adapt to changing conditions (Ghazoul et al. 1998).
When asked about this type of problem, local forest managers often respond that
they know their forest well enough to have a general understanding of species
distribution patterns (pers. obs.). This might be true in many cases, but at least one study
conducted in Para, Brazil revealed that during Rapid Rural Appraisal interviews, local
residents grossly overestimated the densities and distributions of game-attracting fruit
species, presumably because they were favoring forest groves where the tree species
occurred in high densities (Shanley and Gaia 2002). Even if the landowner or community
cannot afford to pay for a 100% inventory, other forms of surveys (like rapid-assessment
type inventories) should be tested to give the forest manager a more accurate idea of the
availability of species and their distributions across the landholding. For example, in the
ejido system of Quintana Roo, Mexico, Lawrence and Roman (1996) report that local
inventory specialists prefer small circular plots as opposed to a 100% inventory design.
Although these circular plots may lack some of the statistical accuracy of a full-scale
inventory, they satisfy local needs for ease of implementation and available financial
Botanical identification during the inventory is crucial for making the most basic
management decisions and should be considered one of the first steps in SFM. In
Brazilian community forest settings, much of the identification is carried out by a forest
manager or tree identifier (mateiro) from the community. Usually in these cases, initial
identifications are recorded as common names. Thanks to the substantial botanical
knowledge base in the community, most mateiros are able to identify species to common
name with few difficulties, using a series of morphological features (i.e., bark, slash, and
odor). The greater problem arises when the inventory sheet is sent back to the office of a
local NGO or state agency and entered into a database. At this point, the common names
are converted to scientific epithets based on lists (often created by a state or federal
agency distant from the management region) that have been circulated between agencies.
Too often, these lists do not correspond correctly to all common names, as the master list
has been generated by a different set of mateiros. The assumption that these common
names will then correspond with each other is highly optimistic, as even mateiros in the
same community will often use different names for the same species. In Lacerda and
Griffiths' (2005) study from the Tapaj6s forest in Para, for example, more than 90% of
common names were technically correct, but once the inventory sheet arrived in the
office and was converted to corresponding scientific names, botanical accuracy was
reduced to less than 40% correct. This type of error can have serious repercussions for
both the selection of harvest trees and post-harvest population dynamics.
The pre-harvest inventory and subsequent botanical identification have strong
implications for the selection of seed trees. Some RIL guidelines require that "mother
trees", or seed trees, are maintained in the management unit at the "appropriate" spacing
and density, based on means of dispersal, individual size, and tree diameter at first
reproduction, as suggested by Guariguata and Pinard (1998). The seed tree concept does
not elude forest managers on the ground. One landowner in a community timber
management project in Acre, Brazil expressed his reluctance at cutting down large-
crowned individuals ofDipteryx spp. (even though such trees would bring an excellent
price on the market), believing them to be good seed sources (as well as being concerned
about the subsequent large gap size created when felled). He would instead take a smaller
conspecific, or what is locally known as a "daughter" tree (F. Correa da Cunha, pers.
Target mother tree densities are often around 3-4 trees ha-1, or (as in accordance
with federal Brazilian law) 10% of the harvest compartment, but given that
reproductively mature conspecific densities in many neotropical forests are much lower
(Lieberman and Lieberman 1994, Guariguata and Pinard 1998), these target goals may be
inadequate. As well, if individuals cannot be properly identified to species, their selection
as seed trees remains dubious. For example, current community timber operations in Acre
identify Dipteryx spp. to one common name, cumaruferro, even though at least 7 species
of this genus exist in the region (C. Baraloto, unpublished data). Consequently, in a
hypothetical situation where 30 mature cumaruferro in a 10 ha logging block have been
left as seed trees, it is highly probable that this group entails several species that may be
isolated from other conspecific reproductive adults.
In general, the dilemmas of implementing a pre-harvest inventory, identifying
scientific species, and selecting seed trees can occur at both the community and industrial
scales. Brazilian federal law requires all companies to survey their logging concessions
prior to harvest, regardless if they are certified or not. The difference for community and
smallholder operations is that a pre-harvest botanical inventory will require long-term
budgetary commitments not normally required on the part of a community-based
organization or forest-dwelling individuals (Janzen 2004). Local parataxonomists, during
the time of the inventories, are taken away from other responsibilities with families and
subsistence activities, potentially compromising their situations at home. As such, they
may need to be compensated financially by the forest management group, which can
cause resentment among untrained community members who do not benefit directly from
the botanical survey.
Minimum Diameter Cuts
The most common harvesting method in the tropics is selective logging, in which
only the largest, most valuable individuals are removed from the forest (Webb 1997).
One of the justifications used to support the practice of employing a minimum diameter
cut is that smaller individuals of the same species will grow to replace the harvested trees.
However, application of minimum-diameter cuts (often termed "high grading", or taking
the biggest and the best) is riddled with undesirable outcomes, and has thus been
criticized as a "short-term gain for long-term loss" (Nyland 2002). As minimum diameter
cuts are often defined by market demands rather than tree biology, Sist et al. (2003a)
suggest that diameter limits be based instead on the structure, density and diameter at
reproduction of target species. Minimum diameter cuts are not necessarily an inherent
part of RIL, but their implementation is linked to the polycyclic cutting cycle, which is a
common RIL recommendation.
High grading typically dooms the next harvest cycle to small, poorly-formed trees
of lesser commercial value or, at the very least, leaves inferior genetic stock to produce
the next generation (Smith et al. 1997, Fredericksen and Putz 2003). For example, Vester
and Navarro-Martinez (2005) point out that selective logging of the largest trees may
deplete a population of the most favorable genotypes, eventually leading to genetic
erosion. The extent to which this selection will affect populations depends on a number
of factors, including harvest intensity, the contribution of the residual trees to
regeneration, and the degree to which undesirable traits are genotypic rather than
phenotypic (Jennings et al. 2001). As well, strong correlations have been drawn between
tree size and fecundity (Chapman et al. 1992), so the removal of a stand's largest
individuals often eliminates the most prolific seed producers, important for both wildlife
as well as regeneration (Thiollay 1992).
Most tropical rain forests have more individuals below the minimum diameter
cutting limit than above. When the reverse is true, the population is often assumed to be
in decline (Jennings et al. 2001). One forest type representing the latter category is the
arborescent bamboo (Guadua spp.)-dominated forest of southwestern Amazonia, which
has been overlooked by the literature for management considerations (see Chapter 3).
Griscom (2003) hypothesizes that if Guadua achieves dominance in managed forests,
either the majority of tree stems will become competitively excluded, or a substantial
proportion of the remaining stems will become deformed as a result of mass loading
(referring to the added physical mass of the bamboo, which can displace aboveground
tree structure (Griscom 2003)). He notes that it is not necessarily the inability of juvenile
trees to germinate and establish in Guadua stands that distinguishes the reduced above-
ground biomass in these forests, but rather the failure of seedlings and saplings to recruit
into larger size classes. These factors present a few challenges to the minimum diameter
cut, such as lack of advanced regeneration (one of the requisites for RIL) and removal of
a substantial portion of what little canopy exists in the form of large trees (pers. obs.). As
such, the stand dynamics of this particular type of forest might lead to an increased
probability of high grading. Yet many smallholders, because they are limited to their
landholdings, may be forced to harvest in these areas.
Another major drawback of the minimum diameter cut affecting both large and
small-scale operations is that knowledge of tree size at first reproduction (which is
essential for maintaining adequate regeneration levels) is very limited in the neotropics
(Guariguata and Pinard 1998). In the Yucatan forests of Mexico, for example, an original
60 cm dbh diameter limit for Swietenia macrophylla (later reduced to 55 cm) was based
not on ecological constraints, but rather, market demand (Snook 1998). Even with the
new 55 cm limit, it is assumed that it will take 75 years for the next generation of S.
macrophylla to reach that minimum, an assumption that has been met with much
skepticism (Vester and Navarro-Martinez 2005).
Brazilian forest law requires that all harvested timber trees must be > 45 cm
diameter at breast height (dbh) (Schulze 2003). This is considered to be a fairly
conservative lower limit by many forest managers, and some impose their own higher
diameter limits on certain species, based on what they think is the ideal sapwood-
heartwood ratio of the tree (F. Correa da Cunha, pers. comm.). Whereas a few species are
characterized by little sapwood (e.g. Peltogyne spp.) and thus are considered worthwhile
to harvest at diameters as small as 45 cm dbh, many species are not harvested unless they
are much larger. In one community operation in Acre, Brazil, Tabebuia spp. and
Hymenaea spp. must reach 70 cm dbh before they are cut, whereas other taxa, such as
Dipteryx spp., must attain 80 or 90 cm dbh before they can be considered for harvest.
This management strategy is also employed in operations that manage vast forest
concessions, and in either case, such a minimum diameter limit could remove many of
the largest trees on the site. Schulze (2003) suggests that by increasing the minimum
diameter cut to 60 cm dbh, one runs the risk of removing the most vigorous individuals
on site, not necessarily the oldest. In this scenario, it would be best to employ a
proportional cutting across adult size classes rather than concentrate efforts on a
minimum harvestable size, thereby ensuring the maintenance of fast-growing individuals
for the next cutting cycle.
Low Harvest Intensity
Low harvest intensities (few trees harvested per hectare) are common in
smallholder systems, and indeed, low intensity selective harvesting is attractive to those
conservationists promoting SFM (Whitman et al. 1997). In small community
management units in Acre, Brazil, the desire to maintain NTFP resources coupled with
local investment in timber certification initiatives have encouraged communities to try
and create as little damage as possible to the residual forest during logging operations
(pers. obs.). This aspiration is also apparently reflected in Oaxaca and Quintana Roo,
Mexico, where community forest managers generally log below the authorized volume
(Bray et al. 2003).
However, low harvest intensities associated with these smallholder timber
operations may actually aggravate logging disturbance. Several studies have documented
that as intensity of harvest increased, fewer skid trails and logging roads were required,
resulting in a declining per-tree damage rate (Gullison and Hardner 1993, Whitman et al.
1997, Panfil and Gullison 1998). Panfil and Gullison, in their 1998 study of short term
impacts of experimental timber harvest intensities in the Chimanes Forest of Bolivia,
cautiously argue for an intensive (albeit at a fixed volume), localized harvest rather than
an extensive extraction over a larger area. They suggest that the primary benefit of doing
so is the reduction in mortality of smaller size classes due to skid trail construction and
use. The tradeoffs involved with this scenario include the following: first, by logging one
area of a large management unit more intensely, there is an increased chance of reducing
local population levels. Second, even if the area damaged due to skid trail and road
construction is decreased, there will be more concentrated disturbance in the treefall area
in the form of large gaps, which are more susceptible to small-scale disturbances, such as
windthrow and fire (Webb 1997). Third, the threshold values for the release of pioneer
and vine species after such harvests is unknown for many forests, necessitating
monitoring plots for this potentially site-specific information (Panfil and Gullison 1998).
Fourth, the impacts of logging damage should ideally consider biodiversity maintenance
in the event of a more intensive and localized extraction (Panfil and Gullison 1998).
Tradeoffs associated with low harvest intensities will necessarily vary, depending
on site and operational constraints. One of the primary factors that should be taken into
consideration is the location and abundance of commercial species in a given logging
block and to compare these values with other management units (if they exist) in the
project. If considerable skid trail construction is required to reach one individual of a
desired species, the destruction and cost necessary to harvest it may outweigh the benefit
of acquiring the log. In this case, it may be possible to take another individual of the same
species in a management unit better suited to its harvest (i.e. greater abundance of the
targeted species or closer proximity), if there are multiple landowners harvesting
simultaneously in one year as a community effort. It should not be assumed that just
because an operation is extracting at a low intensity that it is having minimal impact on
the forest. It is possible to have few commercial-sized trees injured during logging while
simultaneously incurring substantial ground and smaller stem damage (see Chapter 3).
Skid Trail Planning
One way to minimize extraction damage is to reuse old skid trails, but this method can
prove challenging for smallholders, depending on the management system being
implemented. For example, in the "Multiple-Use Community Forest Management
Project" in the extractive settlement of Porto Dias in Acre, Brazil, harvests are conducted
in 10 ha (200 m x 500 m) logging blocks within individual landholdings (approximately
300 ha), with at least five landholdings participating in a given year. The majority of the
first logging blocks (from 2000) in the participating landholdings were located near
existing roads, which greatly facilitated the removal of stems from the logged areas and
the associated construction of log decks. Logging blocks for the subsequent years were
then positioned next to these initial logged sites so that old skid trails from previous years
were reused. This is beneficial to the forest managers in that they do not need to construct
an extra length of primary skid trail (although it does require them to open up these trails
again every year), and it is favorable to the forest because the amount of superfluous skid
trails is minimized. Using this method, though, means that future tree regeneration will
never become permanently established along the length of these reused skid trails. It also
requires the constant perturbation of the previous logging blocks.
In comparison, the forest management project of Sdo Luis de Remanso (another
extractive settlement project in the state of Acre) utilizes 50 ha logging blocks, slightly
reducing the need to reenter the management area in sequential years (depending on the
number of members that remain in the project and the yearly execution of management
plans). Instead of several different landholdings simultaneously participating in the
harvesting (as is the case in Porto Dias), only one landholding (in general), or 50 ha, is
logged per year. In comparison to the Porto Dias system, this setup may be more
economically viable, as the forest managers usually harvest in just one landholding per
year, so ground efforts are not dispersed. Conversely, in Porto Dias, if five landholdings
are logged within a given year, the logging crew is compelled to maintain roads and
construct/reopen skid trails in several sites (N. Marcondes, pers. comm.), adding extra
cost and effort to the operation.
One prerequisite for RIL is the implementation of growth and yield data for
different management regimes to guide the choice of rotation cycle (Silva et al. 1995). In
the Brazilian Amazon, the rotation cycle generally ranges from 30-35 years, based on an
average diameter growth rate of 0.5 cm year-1 (Silva et al. 1995). Silva et al. (1995),
however, caution that ongoing measurement and analysis of permanent plots over the
next 30 years are needed before forest managers embrace an ironclad management
prescription. They point to the variable diameter increments in any given class of trees,
based in part on environmental conditions, growth habits, and genotypes.
Another drawback of the Brazilian harvest rotation is the application of silvicultural
recommendations generated in one region to another. By 1989, approximately 105 ha of
permanent plots had been established in the Brazilian Amazon, but most of them are
located in the state of Para (Silva et al. 1995). Meyer and Helfman (1993) point to the
need for sustainable resource management plans based on an understanding of specific
local conditions. Many of the Brazilian community timber operations are emerging in the
western state of Acre (see Kainer et al. 2003), yet these management systems tend to base
their diameter and volume increment projections on eastern Amazonian data. One
exception is the permanent plot study that has been established in the Pedro Peixoto
Settlement Project by the Brazilian Agricultural Research Corporation (EMBRAPA),
which has based the smallholder management system on locally-generated growth and
yield data. Drawing from the results generated by the project, Oliveira et al. (1998)
propose that extraction of fewer trees per hectare at shorter intervals, in concert with
silvicultural treatments, will maintain pre-intervention forest structure and biodiversity
levels by creating a gap mosaic of different ages. In other words, a short rotation cycle (in
this case, ten years whereby four hectares are harvested annually on a 40-ha stand) will
substitute for the larger impact caused by more intensive interventions.
Yet, even if site-specific growth and yield models state otherwise, a short rotation
could severely degrade the quality of the resource base. Webb (1997) warns that although
increased light levels due to removal of harvested individuals will generally promote the
rapid growth of residual trees, repeated harvests over short rotation periods will
potentially result in low levels of hardwood regeneration as well as a highly
discontinuous canopy. Oliveira (2000), though, counters that damage levels from short
interval cycles can be mitigated to a certain extent by reusing old logging roads and skid
trails. As well, he makes the observation that the Brazilian SFM system is difficult to
apply to the smallholder setting given that the required rotation cycles (20-30 years)
make the clearing of forest for agriculture and pasture a more enticing option since the
profits from those activities are much more immediate.
Proponents of RIL regard the pre-harvest cutting of lianas on crop trees essential if
canopy and future crop tree damage are to be minimized and lianas are abundant (Fox
1968, Appanah and Putz 1984, Vidal et al. 1997, Gerwing and Vidal 2002, Pereira Jr. et
al. 2002). On average, lianas connect each canopy tree to anywhere from three to nine
others in Amazonian terra firme forest (Vidal et al. 1997). During tree felling, this liana
connectivity inadvertently can bring down several other trees, creating a far larger gap
than would occur if lianas associated with harvest trees were pre-cut (Appanah and Putz
1984, Vidal et al. 1997). This suggests that directional felling (a basic tenet of most RIL
guidelines whereby harvest trees are felled in a specified direction to minimize residual
stand damage), is less effective when interconnected vines are not eliminated (Fox 1968).
Furthermore, as these uncut lianas fall with the crop tree, they typically resprout
vigorously, thus competing with new tree recruits and reducing establishment success in
gaps (Schnitzer et al. 2000, Gerwing 2001).
Yet, there is little doubt that lianas are essential components of tropical forests,
providing an important food source for wildlife species as well as supplying a canopy
pathway for some arboreal fauna (Putz et al. 2001). Lianas also contribute to canopy
closure after treefall and stabilize the microclimate of the subcanopy layer (Schnitzer and
Bongers 2002). When asked about the disadvantages of cutting lianas, local forest
managers in Acre, Brazil articulated concern for game species because many liana fruits
are available when tree fruits are not (pers. obs.), an observation that is corroborated by
several scientific studies (Heideman 1989, Morellato and Leitao-Filho 1996). These same
forest managers have also observed that lianas constitute pathways between individual
tree crowns for many animals (especially primates), something that has also been
substantiated in the literature (Schwarzkopf and Rylands 1989, Bobadilla and Ferrari
2000). Consequentially, local logging crews from some Brazilian community forests have
expressed reluctance at cutting lianas prior to the felling cycle (pers. obs.).
Implementing this particular RIL recommendation might present a few quandaries
for the community or smallholder forest manager who is looking to maintain more than
just the next cycle's timber yield, a consideration that may not manifest itself with an
industrial commercial logging operation. Yet, not cutting lianas presents an obvious
tradeoff, as they negatively influence growth, leaf production and sexual reproduction of
host trees (Parren 2003). This is an important consideration, particularly in a landholding
that is limited in size, and, consequently, in future crop tree abundance. Perhaps more
importantly, though, is the shear risk that the logging crew takes when felling a tree that
has not been freed of climbing vines (Amaral et al. 1998), a factor that is valid at both the
smallholder and industrial scales. Parren and Bongers (2001) propose that the benefits of
cutting or not cutting lianas prior to logging is heavily site dependent, so it may be that
local forest managers must experiment with this facet of the RIL paradigm before making
a final decision as to pre-harvest silvicultural prescriptions.
In this review, I have attempted to highlight some of the ecological constraints of
RIL as practiced at the community and smallholder scales, mostly drawing upon
examples from the Brazilian Amazon region. Many of the limitations outlined here can
also be applied to the industrial commercial setting, such as the importance of a pre-
harvest botanical inventory. However, a key distinction with smallholders is that these
forest managers often extract timber from undersized management units, necessarily
reducing the size and often the density of the populations from which they are harvesting.
Ghazoul et al. (1998) point out that SFM (and presumably RIL) requires the maintenance
of viable tree populations, but admit that minimum effective population sizes and
densities are rarely known. Logging, even under the RIL model, can reduce population
viability when conspecifics are isolated from each other (Ghazoul et al. 1998).
Although RIL is, in general, ecologically beneficial to the residual stand, there is
evidence that the magnitude of benefits varies, depending on which forest system is being
exploited (Fredericksen and Putz 2003, Schulze 2003). Some logging guidelines outlined
in this review are unlikely to change in the near future, simply because they are part of
Brazilian federal law, such as the requisite rotation cycle and the minimum diameter cut.
Nor perhaps should they, given the alternative options, such as reentering previously
harvested stands after a short time or taking undersized trees. Given that many of these
small scale timber operations must continue to function within the confines of their
realities, I see the following points as potential recommendations for smallholder timber
management following RIL guidelines.
Logging Block Size and Implications for Pre-harvest Inventories
A diversity of smallholder systems has been described in this review, illuminating the
need to consider details of each one (such as logging block size) when planning timber
harvests. In the state of Acre alone, several logging block sizes are used, from the four-ha
harvest compartment in Pedro Peixoto, to the 10- and 50-ha systems of Porto Dias and
Sdo Luis de Remanso, respectively. As logging operations are currently executed, I
believe that the 50-ha system is best suited to maintain timber species population viability
simply because managers implement a pre-harvest inventory in a larger unit, providing a
more accurate assessment of species distributions than with a 10-ha block (see Table 2-
1). Also at this scale, logging crews will not need to reenter the 50-ha area as frequently
as is required in the Porto Dias system, in which skidders often return annually to
previously harvested logging blocks, reopening primary skid trails. Also, by harvesting
just one 50-ha unit per year, the logging crew is arguably exerting less effort in person
hours and equipment usage, as compared to harvesting several non-contiguous 10-ha
blocks in a given year.
Even so, political, cultural and socio-economic reasons exist for choosing one system
over another. As such, although harvesting a 10-ha block in one year may not be the most
ecologically viable option, it is what certain communities have to deal with, along with
all of the associated environmental disadvantages. The 10-ha logging block system is also
likely to grow in usage, as it is reportedly being favored by new communities wishing to
manage for certified timber in Acre (N. Marcondes, pers. comm.). This reality suggests
the need for full scale inventories of the entire landholding projected for timber
management to minimize potential dangers of inbreeding and inadequate seed dispersal.
The measurement of genetic variation is impractical for forest managers, but the potential
for genetic drift can be reduced by basic knowledge of species populations generated by a
landscape scale inventory (Jennings et al. 2001). Failing implementation of a 100%
survey for those trees above at least the minimum diameter cut, rapid assessment
inventories should be considered.
Botanical identification is a critical component of logging inventories, for which local
parataxonomists are key partners. Before inventories take place, however, common
names should be agreed upon, so that the same nomenclature is applied throughout the
survey process. To ensure that these common names then correspond with their correct
scientific counterparts, care must be taken to have samples identified with known
herbarium examples, if possible. At the very least, care should be taken to not solely rely
on databases generated by other organizations, whose own tree identifiers may also be
using common names that differ from those being used by a local parataxonomist from
the timber project.
Recognizing that smallholder and community forests must establish their own
management strategies while building on RIL guidelines, I suggest that systematic post-
harvesting monitoring become integrated into logging operations, so that the benefits of
recommended practices (such as pre-harvest liana cutting) may be tested at the
appropriate scale. I recognize that smallholders in the rural tropics have limited financial
and technical resources, and that monitoring may be unfeasible for them, unless they
perceive that proposed activities add to their livelihood productivity. Accordingly,
Ghazoul (2001) concedes that given associated high costs, an adaptive management
approach, entailing collaboration between biologists and local forest managers, would
allow for sustainable management to be based on the technical skills, ecological
knowledge base, equipment, and funds available to the managers. This collaborative
approach to forestry requires considerable communication between local managers and
professional forestry technicians and an assessment of existing local ecological
knowledge and traditional management systems (Carter 1996).
To this end, collaboration between smallholders, the agencies that support them and
researchers who may be able to contribute technical expertise to the process should be
fully encouraged. I recognize that working outside of traditional field sites in community
settings obliges researchers and practitioners to approach forest management and
monitoring differently, requiring them to consider social, political, and economic factors
that affect the way that local peoples interact with the natural resource base. Beyond this
consideration, we need to also think about the dynamics of the forest managed for a suite
of products and services at the scale discussed in this review, providing communities and
smallholders with valuable feedback on their management systems and further providing
a link between biodiversity conservation and SFM.
Moving beyond RIL
It is indisputable that RIL methods are a vast improvement over the much more
destructive and haphazard conventional selective logging practices that are,
unfortunately, still the norm (Schulze 2003). Yet, forest management under RIL
guidelines alone could be at odds with the SFM model, which demands forest
management for a suite of goods and services as well as the maintenance of ecological
functions (Fredericksen and Putz 2003). Arguably, it is this broader suite of outputs that
many communities target in their forest management schemes. Rural communities
depend on wildlife, NTFPs, watersheds, and soil properties for their livelihoods,
underscoring the importance of looking beyond sustained timber production as the most
important forest management objective. Even if seed sources and dispersal mechanisms
remain intact, thereby (in theory) ensuring the viability of future timber regeneration
(Baraloto and Forget 2004), the long term logging impacts on these other livelihood and
conservation elements must not be overlooked. It seems particularly appropriate then to
move beyond an emphasis on RIL guidelines for smallholders engaged in timber
management, while still incorporating its basic tenets into practical application.
It is highly probable that the so-called sustainability of intensive timber
management may not be known until at least three cycles of harvest have been completed
(Poore et al. 1989). This observation bolsters Simberloff s (1999) suggestion that any
proposed silvicultural system designed to maintain biodiversity while producing timber
should be treated as a hypothesis, due to limited empirical evidence supporting its
viability. Until then, though, we must continue to focus on forest management research
that is not just rooted in tree reproduction biology (Guariguata and Pinard 1998), but also
consider the application of such investigations at the smallholder and community levels,
necessarily incorporating the local ecological knowledge that is often overlooked by
academic researchers yet when it is acknowledged. In this way, we might be able to avoid
the pitfalls of relying on RIL as proxy for SFM, especially in terms of the smallholder
and community settings, which necessarily require considerations beyond the biological
Table 2-1. Contrasting tree population parameters from inventories at two different spatial scales commonly used by smallholders.
Shown are the results of a 50 ha inventory of commercial species stems > 35 cm DBH from PAE Sdo Luis de Remanso,
Acre, Brazil, along with the range of values (5 and 95%) obtained by 500 simulations of randomly-placed 10 ha plots
within the larger area. Species were included only if they were represented by at least 10 trees in the plot, and their
botanical identification had been verified. Bold cases indicate that the distribution of values from the 10 ha simulations
exceeded the actual value at 50 ha, suggesting that management decisions would change depending on the scale at which
inventories are conducted.
Common Harvestable Density'
Species Came Total Density (ind ha ) (He D sity
Name (ind ha )
50 ha 10 ha 50 ha 10 ha
Aspidosperma vargasii Amareldo 0.92 0.5- 1.3 0.36 0.1 -0.5
Myroxylon balsamum Balsamo 0.31 0.2 0.6 0.15 0.1 0.4
Tetragastris panamensis vermelho 4.62 2.6 8.6 0.82 0.2 2.2
Cedrela odorata er 0.87 0.3- 1.7 0.67 0.3- 1.3
Amburana cearensis Cerejeira 0.56 0.1 0.9 0.41 0.1 0.9
Apuleia leiocarpa Cumaru 0.97 0.7- 1.7 0.79 0.5- 1.3
Barnabydendron riedelii Guaribeiro 0.82 0.3 1.3 0.67 0.2 1.0
Tabebuia serratifolia Ip6-amarelo 0.23 0.1- 0.4 0.08 0 0.2
Hymenaea parvifolia Jutai 1.21 0.7 2.1 0.64 0.2 1.2
1 refers to minimum size harvested, which varies among species from 45 60 cm DBH.
Table 2-2. An example of spatial aggregation among neotropical tree species harvested
by communities in Acre, Brazil. Shown are the minimum distance, if any, for
each species at which individuals (stems > 35 cm DBH) are significantly
clustered (Ripley's L-statistic exceeding 95% confidence interval, based on
1000 bootstrapped values at 5 m distance intervals using Besag's edge
correction conducted in Ripley software package; Marcon 2005). Data are
from a 50 ha inventory of commercial species from PAE Sdo Luis de
Remanso, Acre (Centro dos Trabalhadores da Amazonia, unpublished data).
Species were included only if they were represented by at least 10 trees in the
plot, and if their botanical identification had been verified.
Species Common Name Spatial Distribution
Aspidosperma vargasii Amareldo No
Myroxylon balsamum Balsamo No
Tetragastris panamensis Breu-vermelho Yes 6 m
Cedrela odorata Cedro-vermelho Yes 8 m
Amburana cearensis Cerejeira No
Apuleia leiocarpa Cumaru-cetim No
Barnabydendron riedelii Guaribeiro Yes 15 m
Tabebuia serratifolia Ip6-amarelo No
Hymenaea parvifolia Jutai No
EVALUATING LOGGING DAMAGE TO FUTURE CROP TREES IN ACRE,
BRAZIL: IMPLICATIONS FOR MANAGEMENT OF A CERTIFIED COMMUNITY
The guiding principle of sustainability dictates that a commercial forest should be
managed in a way that limits as much undesirable damage as possible to the ecosystem.
This is essential for not only ecological reasons, but economic ones as well (Miller
1998). For example, it is well known that unplanned (conventional) logging results in
unnecessary damage to the residual forest and reduces the chances for future timber
production (Pinard et al. 1995). Johnson and Cabarle (1993) estimate that anywhere from
26 to 75% of a given tropical stand of future crop trees (FCTs) are injured during
conventional logging operations. Damage to residual stands is often correlated with
logging intensity, which varies greatly, depending on spatial and temporal factors, access,
equipment, terrain, climate, supplemental treatments, and market acceptance of lesser-
known species (Metzger and Schultz 1984, Sist et al. 1998, Putz et al. 2000).
The most common harvesting method in the tropics is selective logging, whereby
only scattered trees of sufficient economic value are extracted from the forest. In
unplanned operations, selective logging has been observed to cause extensive canopy
cover removal (Johns et al. 1996, Webb 1997, Whitman et al. 1997, Pereira et al. 2002),
soil compaction (Whitman et al. 1997), long-term changes in tree species composition
(Thiollay 1992), and reduced faunal diversity (Johns 1991, Thiollay 1992). Accordingly,
harvesting techniques and intensities need to be controlled to reduce residual stand
damage (Webb 1997).
Controlled selective logging, or reduced-impact logging (RIL), utilizes methods
that minimize damage to the residual stand, such as pre-harvest inventory, mapping,
directional felling, vine cutting, and planning of skid trails, log decks, and roads (Uhl et
al. 1997), potentially avoiding some of the post-logging damage noted by other authors
(Pinard and Putz 1996, Johns et al. 1996). RIL techniques have been shown to reduce
damage to residual stands in mixed dipterocarp forests in Southeast Asia by as much as
30-50% (Pinard and Putz 1996, Bertault and Sist 1997, Sist et al. 1998). Furthermore,
RIL methods have demonstrated an increased efficiency of logging operations and a
reduction in the amount of timber wasted during a typical harvest (Barreto et al. 1998,
Holmes et al. 2002).
Even with implementation of RIL methods, sustainability of harvesting will
ultimately depend upon advanced regeneration to replace the larger size classes during
the next cutting cycle (Putz et al. 2000). This assumption is challenged by the damage
incurred by the residual stand during logging operations. For example, Jackson et al.
(2002) report that RIL operations in a FSC-certified timber concession in Bolivia killed
or severely damaged an average of 22 trees (> 10 cm dbh) for every one extracted. In the
same forest, Krueger (2004) found that flagging reduced damage to FCTs by 20% in
felling gaps and by 10% along skid trails receiving 2-10 skidder passes, demonstrating
that marking, or flagging, of FCTs can reduce damage to the residual stand by improving
visibility during on-the-ground operations (Dykstra and Heinrich 1996). Currently,
general Brazilian forest management practices do not employ this method.
Mechanisms underlying these residual stand impacts need to be understood in the
context of specific local conditions (Sheil and Van Heist 2000). Commercial species
distribution and human exploitation varies greatly between neotropical regions,
supporting the argument that silvicultural treatments appropriate in one location will not
necessarily be applicable in another (Putz 1996). For example, Webb (1997) points out
that difficulties encountered in Carapa nicaraguensis forest swamp in Costa Rica
required management techniques not employed in other logging operations. Given the
need to develop and evaluate site-specific management strategies that will mitigate the
negative effects of timber harvesting, it is essential to look at several variables across
One forest type on which little has been written about in the management literature
is the bamboo-dominated forest of southwestern Amazonia, which covers almost 180,000
km2 (Griscom and Ashton 2003). This forest is characterized by a mix of structurally
heterogeneous stands without bamboo and patches of trees scattered within a canopy
dominated by dense stands of Guadua species (with some species of Guadua reaching
heights of 30 m) (Griscom 2003). Similar to other mast seeding bamboo genera (Janzen
1976), Guadua spp. undergo a single synchronized reproduction event followed by
synchronized mortality, at 25-30 year intervals (Nelson et al. 2001). At least two species,
G. sarcocarpa and G. weberbaueri have been linked with reduced forest basal area as
well as tree species richness (Silveira 2001, Griscom 2003). This phenomena is strongly
linked to the species' ecology: branches of G. sarcocarpa and G. weberbaueri culms are
endowed with curved barbs which act as grappling hooks (see Fig. 3-1), enabling the
bamboo to be partially or fully supported by trees for vertical growth (Griscom 2003),
weighing them down, and increasing the likelihood of blow-downs (Griscom and Ashton
2003). In concert with the common tree blow-downs in this region, Griscom and Ashton
(2003) propose that low recruitment of trees into the larger size classes explain the
persistence of Guadua domination, rather than the catastrophic disturbances of fire and
agricultural intervention cited by other authors. If Guadua achieves dominance in
managed forests, commercial trees may become competitively excluded and a substantial
proportion of the remaining stems may become deformed, due to mass loading (Griscom
2003). The low visibility in some areas of the Guadua dominated forest also presents an
impediment to sawyers trained in directional felling (F. Correa da Cunha, pers. comm.).
Combined, these factors present some unique challenges for silvicultural management
and timber extraction in Guadua-dominated forest.
This study evaluated logging damage to FCTs in a certified community forest in
Acre, Brazil. The main objectives of this investigation were (1) to determine the extent of
area disturbed (canopy gaps and skid trails) during a small-scale certified logging
operations, (2) to determine a difference in damage incidence to FCTs between forest
with and without bamboo (Guadua), (3) to determine if likelihood of FCT damage is
increased when lianas are present, (4) to determine to what extent harvesting can be
conducted more intensely (m3ha-) without incurring greater residual stand damage in
general and damage to FCTs > 20 cm dbh in particular, and (5) to determine to what
extent marking of FCTs diminishes damage and/or mortality of FCTs (>20 cm dbh)
The Porto Dias Extractive Settlement Project (PAE) (S 1000'39, 9", W
66 46'26,4") is situated in the northeastern corer of the Brazilian state of Acre, in the
municipality of Acreldndia. It is bordered by various rubber estates, cattle ranches, and
colonization projects (see Fig. 3-2), and is separated from Bolivia by the Abuna River
(Stone 2003). The site is defined by red-yellow latisols of low fertility and relatively flat
topography, with an annual precipitation of 1890 mm yr1, most of which falls between
November and March. As of 2001, approximately 97% of the reserve was still forested
(CTA, unpublished data), but it is one of the last large contiguous pieces of forest in the
municipality (N. Marcondes, pers. comm.). As such, it is under constant pressure from
the influences of agricultural conversion within the PAE as well as illegal clearing on its
borders. Older residents recall Swietenia macrophylla (mahogany) being logged in the
region and transported via the Abuna River (Stone 2003), but mahogany is currently
extremely rare in the PAE. Three tropical humid terrafirme forest types are generally
recognized in the reserve: open canopy forest; open canopy forest mixed with arborescent
bamboo (Guadua spp.); and closed canopy forest.
Three species of Guadua are found in the study site: G. weberbaueri, G.
sarcocarpa, and one as-yet unnamed species, which occurs in riverine forest along the
Abuna River (M. Silveira, pers. comm.). At the end of 2004, when logging operations
commenced, G. sarcocarpa underwent a monocarpic dieoff that affected much of the
study area. Both G. weberbaueri, and G. sarcocarpa are thought to be synchronized on a
30 year masting cycle across large areas (100-100,000 km2) (Nelson et al. 2001). The two
species exhibit similar autecological characteristics, with an apparent lack of seed
dormancy, germination occurring in both gaps and non-gaps, and partial shade tolerance
among seedlings (Griscom 2003). In at least one study in Madre de Dios, Peru, G.
weberbaueri demonstrated a lower mean stem diameter (cm) in comparison with G.
sarcocarpa (4.04 vs. 6.67, respectively), but exhibited higher stem density (stems/100
m2) (34.2 vs. 23.8) (Griscom 2003).
The reserve encompasses approximately 22,145 ha, and is divided among
coloca5oes, or individual family landholdings constructed around pre-existing rubber
trails (approximately 300 ha each). There are around 100 registered families who live in
the PAE, in addition to an estimated 20 families who are squatting illegally in the reserve
(Santos 2000). Traditional family livelihoods in PAE Porto Dias include forest extraction,
hunting, subsistence agriculture, and small-scale animal husbandry (Stone 2003). Cash
was customarily earned from rubber and Brazil nut harvests, but in recent years rubber
tappers have been turning to logging as an alternative income source. The Center for
Amazonian Workers (CTA), a non-governmental organization based in Acre, has been
instrumental in developing the management framework of the Multiple-Use Community
Forest Management Project. The project was initiated in 1995, but the first timber harvest
was not harvested until 2000. SmartWood certified the logging operation using Forest
Stewardship Council (FSC) criteria in 2002, distinguishing it as the second FSC-certified
community timber management project in Brazil. FSC certification requires that the
project implement a special monitoring program in the bamboo-dominated forest, as there
is concern that logging damage may be greater there than in forest lacking bamboo (N.
Marcondes, pers. comm.). Additionally, Guadua-dominated stands have been identified
as important habitat for many valuable wildlife species, making them focal areas of
interest for conservation efforts (Griscom 2003).
Forest Management and Harvesting Operations
Landholdings (approximately 300 ha each) of reserve residents serve as the timber
management areas, and therefore do not favor the compartmentalization typical in
Brazilian industrial timber operations (Braz and Oliveira 1996). In compliance with
federal regulation, 10% of the landholdings may be cleared for agricultural activities and
5% must be set aside for preservation (no harvesting of trees). Thus, most residents have
at least 255 ha available for timber management, result in a 25-year logging rotation (255
ha/10 ha yr-1= 25.5 years). The original ten members of the project all harvested 10 ha per
year, but within a few years after the first timber harvest, it was decided that only five
landholdings per year would be harvested. In the latter case, project members began
harvesting every two years in their respective landholdings. However, given that some
families have since dropped out of the timber project, other members have still harvested
in sequential years (i.e. more than one or two years in a row), further confounding the
question of the cycle length.
Inventories and mapping of all tree species >35 cm dbh are carried out one year
prior to harvesting in 10 ha blocks within each landholding to be logged; as of 2005,
though, a 100% inventory of the entire property had not been executed in any of the
participating landholdings. After selecting the crop trees, the project must submit its
logging plan to the Brazilian Institute of the Environment and Renewable Resources
(IBAMA), which will then approve or disapprove the plan, based on an acceptable basal
area removed, desired residual stand species distribution and crop tree proximity to water
sources. Only trees > 45 cm dbh may be legally harvested. Logging is generally carried
out in the driest part of the year, from July to November. Twelve major timber species are
presently exploited (see Table 3-1), with a low intensity of cutting (1-3 trees ha-1 or
approximately 10 m3 ha-1). Since the timber project members began logging biannually,
the total annual harvest for the project is around 500 m3, or 100 m3 landholding-1. Sawn
board and artisan products from the community sawmill are primarily sold to the
domestic certified market in Southern Brazil. The project does not market logs.
Members of the timber project (as of August 2005 there were seven participating
families) conduct most of the logging activities. These techniques include inventories,
selection of crop trees, assessment of defects in crop trees, directional felling, and skid
trail planning, RIL skills for which they have been extensively trained over the last ten
years by the Technological Foundation of Acre (FUNTAC) and the Tropical Forest
Foundation (FFT) in Para, and for which they now train other forest managers. The only
activity that they do not directly participate in is skidding logs to the sawmill, a process
contracted to heavy equipment owners and operators from outside the community.
During logging operations, the landowner typically assists the construction of skid trails
by flagging the path that the skidder or tractor operator will use.
Future Crop Tree (FCT) Selection
Fifty-two tree species were selected for the study, based on their potential
importance as commercial timber sources (see Table 3-1; Ribeiro et al. 1999, Lorenzi
2000). Due to external market demands, however, only about 10-12 species are exploited
on a regular basis in Porto Dias, most commonly Amburana cearensis, Cedrela spp.,
Peltogyne spp., Dipteryx spp., Apuleia leiocarpa, Tabebuia spp., Aspidosperma vargasii,
and Hymenaea intermedia. For purposes of this study, the definition of FCTs was
broadened to account for all species of commercial value, both current and future (see
Table 3-1). When proper field identification was questionable, species verifications were
conducted at the Federal University of Acre herbarium in Rio Branco, Acre, Brazil, but
some trees were only identified to genus (see Table 3-1).
Experimental Design and Pre-logging Inventory
The greatest amount of damage to FCTs during logging occurs in areas closest to
felled crop trees and skid trails (Johns et al. 1996). Therefore, observations of logging
impacts were concentrated in a 50 m radius (hereafter referred to as "zone of impact")
around each designated harvest tree in each of the 10 ha (200 x 500 m) logging blocks in
four landholdings scheduled to be harvested in 2004-2005 (Barrinha 1, Barrinha 3,
Palestina, and Sao Pedro). This number was eventually reduced to three, as the rainy
season began before Sao Pedro could be logged. A full inventory of all commercial FCTs
>20 cm dbh was conducted in the zones of impact. Sample sizes in the logging blocks
(i.e. the number of zones of impact) depended on the abundance of crop trees, with a total
of 18 zones of impact and 161 FCTs in Barrinha I, 20 zones of impact and 216 FCTs in
Barrinha III, and 12 zones of impact and 109 FCTs in Palestina, including some overlap
between zones. In 2004, before logging operations commenced, the following data were
noted for FCTs: species, location (x, y coordinates), dbh (1.3 m or above the buttresses);
total estimated height; qualitative trunk and crown quality ("good", "tolerable", or
"inferior"); crown position ("dominant", "intermediate", "suppressed"); presence in
bamboo-dominated stand ("yes"/"no"), and presence of at least one liana to the canopy
All primary and secondary skid trails and treefall gaps were mapped using a
compass and meter tape. The size of treefall gaps was estimated using the "center-point"
system (modified from Runkle 1992). From a point in the approximate center of each
treefall gap, compass angles in eight different directions were measured to the edge of
intact canopy (Johns et al. 1996). This information was then entered into ArcGIS 9.0
(ESRI, Redlands, CA), to calculate area of the resulting polygon. The ground area
transformed by the total length of skid trails was determined using modified methods
from Johns et al. (1996), with skid trail width being measured at the beginning and end of
each straight segment of trail. Trail width measurements were also taken at each bend to
account for potential bulldozer maneuvering. This information was then combined with
total skid trail length (estimated using pre-harvest inventory maps) in ArcGIS 9.0.
FCT Location in Bamboo (Guadua)-dominated Forest
Zones of impact (both in areas of high and low harvest intensity) naturally fell in
locations where bamboo was present and in areas where it was absent. Location in
Guadua-dominated stands was defined as at least one culm of Guadua spp. within 2 m of
the FCT trunk.
To test the effect of presence of lianas on FCT damage probability, liana
occurrence on a FCT was denoted as at least one vine connecting to the canopy.
To detect the impact that a range of timber harvest intensities could have on FCTs
in Porto Dias, zones of impact were established in locations scheduled to receive a locally
high intensity of cutting (>3 trees ha-1), and also in areas of locally low intensity cutting
(1-2 trees ha-1). This design allowed for establishing a harvest intensity gradient in each
logging block, but did not necessarily allocate the treatments evenly across the logging
blocks, as the abundance and distribution of crop trees varied greatly between
To test the effect of marking, about half of all inventoried FCTs in each of the four
logging blocks were marked. All FCTs within the zone of impact of a tree to be harvested
were either marked or left unmarked, stratifying the treatment between regions of low
and high crop tree density (see Figs. 3-3-5). Marking entailed a single band of orange
paint 20 cm wide at approximately 1.7 m above the ground (adapted from Krueger 2004).
Initially, the random assignment of the marking treatment was considered when
designing this experiment. However, it was decided that the highly visible presence of
even one marked FCT in the vicinity of unmarked FCTs could induce the logging crew to
alter their activities in some way as to favor neighboring FCTs. This potential source of
bias suggests that statistical independence of the FCT treatments would have been
difficult to achieve even with this type of experimental design.
Assessment of FCT Damage
Inventory data were used to construct a spatial database in ArcGIS 9.0. For each
FCT, distance to nearest felled tree, nearest gap edge and nearest skid trail were
measured. Harvest intensity (trees ha1) was estimated for each FCT as the volume
exploited in the surrounding 1 ha (circumscribed by a radius of 56.4 m). Local FCT
density was estimated as all FCTs > 20 cm dbh within a 25 m radius of a given FCT. This
measurement was included to control for variation explained by local density when
testing for other factors such as marking. Seven months after logging activities were
completed, in June and July, 2005, all previously inventoried FCTs were revisited and
classified according to levels of damage or mortality (see Table 3-2).
Quantifying FCT damage was considered for this study. A logistic regression
analysis was used to test the binary response variable (damaged/not damaged) by treating
each FCT as an independent statistical unit to which a series of independent variables was
assigned: local harvest intensity (continuous); distance to nearest felled crop tree
(continuous); distance to nearest skid trail, (continuous); distance to nearest gap
(continuous); neighborhood FCT density (continuous); presence of bamboo (binary);
marking (binary); and presence of lianas (binary). A regression analysis was performed to
determine the influence of crop tree dbh on felling gap area. Differences were considered
significant at P > 0.05.
Overview of the Logging Operation
The mean tree (> 35 cm dbh) basal area was determined to have the following
values for the three landholdings: 7.7 m2 ha-1 in Barrinha I, 8.3 m2 ha-1 in Barrinha III,
and 8.8 m2 ha-1 in Palestina. All harvesting occurred between October and November
2004, and a bulldozer (Valmet 128) was used to extract logs. Two community-based
sawyers participated in the operations: one felled trees only in Palestina and the other cut
in both Barrinha I and Barrinha III. No new roads were created to facilitate logging
operations, and pre-existing skid trails in logging blocks from the project's previous
harvesting activities were reopened to access new areas. All three logging blocks were
located close to existing roads, so boles were eventually skidded to log decks adjacent to
A total of 22.3 ha of 30 ha were sampled in the three logging blocks. Forty five
trees were extracted from all of the landholdings, with 38 actually cut and 7 that fell
under natural circumstances at least one year prior to harvesting, for a total harvested
volume of 237.7 m3, or approximately 7.9 m3 ha Felled tree (n=38) dbh (cm) ranged
from 60-114, with a mean of 71.7 (s.d. 14.73). Average harvested tree density was
approximately equal between the three logging blocks (see Table 3-3), suggesting that
any differences in damage incurred by FCTs was not due to differences in logging
intensity (Johns et al. 1996). Some of the trees originally selected were not cut for various
reasons (usually they were found to be hollow prior to harvesting), so loggers harvested
several substitute trees. As a result, some of the trees designated as experimental FCTs
the previous year were not affected by logging operations, as they were far from any skid
trails or felling gaps (> 60 m from felling gap, cut stump, or skid trail). These trees (a
total of 18) were not considered for the analysis. At the same time, damaged trees that
were near the newly substituted crop tree sites were not tallied after harvesting, because
no pre-logging data were gathered for them.
A total of 70 FCTs out of 468 were damaged during logging operations, or
approximately 15% of the experimental population. The most common damage types
were slightly damaged crowns, followed by scraped boles and roots, classified as "minor"
damage (see Table 3-3). Tree felling operations damaged a greater proportion of trees
than skid trail construction and use, but the results were not statistically significant.
Marking did not significantly reduce overall FCT damage in the Porto Dias
operations (see Table 3-4). Furthermore, the incidence of FCT damage levels was not
higher in Guadua-dominated forest, nor was there a significant interaction between a
FCT's presence in a Guadua-dominated stand and the marking treatment (interaction
term in Table 3-4). Fifty-six percent of the inventoried FCTs carried lianas, but damage
probability due to liana presence on the FCT was not highly significant (P = 0.10), but
many damaged FCTs had at least one large liana attached to its crown (67%).
The closer a FCT was to a skid trail or logging gap, the greater the likelihood of it
being damaged (see Table 3-4). In contrast, distance to cut stump and local FCT
neighborhood density had no significant effect. Harvest intensity (see Table 3-4) also did
not influence the probability of FCT damage, implying that in this study, the probability
of FCT damage did not increase with the number of trees harvested in the surrounding
The percentage of total sampled area affected by harvesting was 28.9%, with the
majority in felling gaps (16.8%). Average area per felling gap amongst all three logging
blocks (mean s.d.) was estimated to be 393 m2 181. Regression analysis revealed a
significant positive effect of crop tree dbh in single tree gaps area (see Fig. 3-6). Skidding
damaged 11.8% of the total surface area.
Timber harvesting, as conducted by the Multiple-Use Community Forest
Management Project in PAE Porto Dias, had minimal impacts on residual stands, at least
on commercial individuals > 20 cm dbh. Additionally, potential damage was avoided by
not having to construct new log landings and log roads. Even so, the high values reported
for area damaged, both skid trails and felling gaps, merit analysis (see Table 3-3). The
values for felling gap area per tree cut were in some cases almost twice that of other
reported neotropical studies (see Johns et al. 1996). In many instances, increased gap
sizes can be explained by the size of the trees that were felled. For example, one of the
most valuable taxa in Porto Dias, Dipteryx spp., comprised some of the largest
individuals amongst harvested trees (up to 114 cm dbh), contributing to the strong
positive correlation between logging gap size (m2) and dbh of crop trees (see Figure 6). In
the case of Palestina, more than 60% of the harvested trees were > 70 cm dbh, explaining
in part why this particular logging block had the greatest mean gap size (433 m2) amongst
the three landholdings. When considering exceptionally large trees for harvest, forest
managers might do well to consider the tradeoffs between creating a very large gap
(thereby potentially increasing fire and windthrow risk) and the substantial profit for the
sake of the wood.
It is also worth noting that only those species valued for their merchantable timber
were inventoried for the experiment. For example, locally important NTFP species such
as Bertholletia excelsa and Hevea brasiliensis were not tallied during the original
inventories, but during the post-logging survey, damage to several trees of these species
was observed. For future studies, it would be prudent to address some of the issues of
impact of timber extraction on other livelihood activities, such as NTFP collection,
particularly in such regions as the Brazilian Amazon where many communities are
attempting to integrate a suite of income-generating activities. Commercial logging, with
potentially high financial yields, is unlikely to disappear in community forests.
Nevertheless, interest persists in the harvesting of non-timber products such as Brazil
nuts, rubber, and Copaifera and Carapa oils, species that may be adversely affected by
timber harvests. When considering these factors, can timber harvesting and other
livelihood strategies be made more compatible and if so, what are the tradeoffs (see
Salick et al. 1995, Romero 1999, Menton 2003)?
FCT Damage in Guadua-dominated Forest
Given the low visibility in Guadua-dominated forest, I expected that FCT damage
rates would be higher in these areas, but results proved otherwise. There could be a
number of explanations for this outcome: even though poor visibility in Guadua groves
has been cited as an impediment in the felling of large trees because neighboring FCTs
are not always easily seen, FCT damage might be minimal simply because there are
fewer FCTs to damage, as demonstrated by the low basal area estimates cited by other
authors (Silveira 2001), or, more specifically, by the low representation of smaller stems
in those same estimates. Guadua frequently outcompetes tree seedlings and saplings such
that they fail to recruit into larger size classes (Griscom 2003). Therefore, one could
conclude that there are comparatively more FCTs from the lower diameter classes to
damage in forests without bamboo, or at the very least, there are more trees that operators
must avoid when skidding or felling. It is also uncertain as to how the synchronized
mortality of G. sarcocarpa may have directly or indirectly affected felling operations,
although visibility may have improved in these stands because mature bamboo culms
were drying out and falling to the ground in some areas.
Even though I did not find that selective logging increased the probability of FCT
damage in Guadua-dominated stands, the potential for other types of long-term residual
stand damage warrants future investigations. We observed dramatically increased light
levels in post-logging bamboo-dominated forest, possibly attributable to the loss of the
biggest canopy trees in the block, of which there were few to begin with. There are a few
potential problems associated with this. First, by opening large areas during harvesting,
the chances of repeated small-scale disturbances, such as windthrow and fire, are
increased (Webb 1997). Given that Guadua benefits from intermediate and large canopy
disturbances (Griscom 2003), once a gap is opened, bamboo may proliferate and
competitively exclude the majority of tree stems. One example of this phenomenon is
Whitmore's (1984) observation of the reduction in value of Malayan forests that were
invaded by arborescent bamboo after unplanned logging operations. The high light
intensity characteristic of large gaps might also select for a suite of different species as
compared to small gaps (Denslow 1980), a dilemma if the regeneration of small gap-
preferring species is desired by the forest managers. In general, Guadua-dominated
stands are already characterized by a discontinuous canopy, which may amplify the
effects of openings created by logging activities (i.e. increased light levels). As such, for
future operations, forest managers may want to select crop trees based on the presence of
large neighboring trees, ensuring that some sort of basic canopy structure is kept intact
around the logging gap if smaller openings are desired.
Another concern in this forest type is the potential for high grading, or the type of
harvesting that selects for the largest and most valuable individuals. One of the
underlying assumptions of RIL is that advanced regeneration of valuable commercial
timber species exists uniformly throughout the stand so that future harvests are not
compromised by the present one (Schulze 2003). Yet in Guadua-dominated stands, the
lack of advanced regeneration potentially increases the chances of having small, poorly-
formed trees of lesser commercial value dominate the next harvest. For the operation
described in this paper, FSC certification standards, and indeed, Brazilian federal law,
require that forest managers leave at least 10% of the harvested species' populations as
seed trees. This perhaps ensures harvests for generations to come, but it does not
guarantee a sustainable timber yield for the immediate second cycle. As well, even
though Brazilian federal law imposes a minimum diameter cut of 45 cm dbh, many
species in this operation are not harvested until they are much larger (see Chapter 2),
since taxa such as Dipteryx spp. are characterized by high levels of sapwood and thus
attain commercial value at a greater girth. As a result, there is often an understandable
temptation to take the biggest and the best individuals from a site, although at least one
landowner in Porto Dias admitted hesitation in taking large Dipteryx spp. individuals, as
they are often the most prolific seed sources (F. Correa da Cunha, pers. comm.). The
complexities associated with timber harvesting in Guadua-dominated forest may require
a much different approach to silvicultural techniques than developed in other parts of
Amazonia, and underscore the need for future monitoring of harvesting activities and
residual stand dynamics as well as extreme caution in the removal of the largest trees on
Impact of Lianas on FCT Damage
Lianas connecting the crowns of trees will often cause a domino-like effect during
felling operations (Vidal et al. 1997), although few adjacent trees were completely
brought down by felled trees during this operation. Other neotropical studies (see Fox
1968, Appanah and Putz 1984, Johns et al. 1996, Vidal et al. 1997, Pereira Jr. et al. 2002,
Gerwing and Uhl 2003), have demonstrated increased damage during logging operations
when vines were present. Although I expected vines to increase FCT damage, I did not
observe any statistically significant effects at the ca=0.05 level, though given results in
other studies, it is probable that the magnitude of the outcome is consequential from a
management standpoint (Steidl and Thomas 2001). It was evident from post-logging
observations in the field that parts of crowns were pulled down by bulldozers at they
moved through the forest. Undoubtedly, if lianas associated with FCTs had been cut at
least one year prior to harvesting, some of this damage could have been avoided. Still,
Porto Dias forest managers expressed reluctance to cut lianas prior to the felling cycle
(pers. obs.), citing their importance as a food source for wildlife species. Yet, it has been
shown that lianas negatively influence the growth, leaf production and sexual
reproduction of host trees (Parren 2003), and when not cut prior to harvesting operations,
increase the potential danger for logging crews (Amaral et al. 1998). Parren and Bongers
(2001) propose that the benefits of cutting or not cutting lianas prior to logging is site
dependent, so it may be that local forest managers must experiment with this particular
RIL guideline before making a final decision as to pre-harvest silvicultural prescriptions.
Impact of Harvest Intensity on FCT Damage
Local harvest intensities in the logging operation described here removed 1-4 trees
ha-1, seemingly with no increased levels of residual stand damage in the > 20 cm dbh
class as the maximum ceiling of harvest intensity (4 trees ha-) was reached. This type of
low-intensity selective logging is attractive for conservationists espousing the
implementation of sustainable forest management as a way to limit canopy cover loss and
protect ecosystem function and biodiversity (Whitman et al. 1997). Even so, Gullison and
Hardner (1993) have pointed out that as the number of trees removed increases, fewer
new skid trails are required (although the damaged area per tree due to logging gaps
remains constant). In the case of the operation described here, even though there was no
evidence supporting increased levels of FCT damage as harvest intensity increased, the
average area (m2) damaged per tree extracted due to skid trail construction was quite high
(see Table 3-3) as compared to other neotropical studies, relative to the number of trees
extracted per hectare. For instance, Johns et al. (1996) found that unplanned logging
operations using a bulldozer to skid trees damaged 119 m2 per bole removed in relation to
skid trails (with an average harvest intensity of 5.6 trees ha-1). The high levels of area
damaged per tree extracted in this study suggest that there might be a tradeoff associated
with trying to harvest isolated trees across even a relatively small area of 10 ha, both in
terms of cost and as a function of area and trees damaged.
One way to offset this effect of low intensity harvests would be to increase
localized harvest intensity at a fixed volume. The primary benefits of this method would
be reduction in mortality of smaller size classes as well as fewer skid trails, not
necessarily reduction in mortality or damage to FCTs (Panfil and Gullison 1998).
Another consideration is that large gaps allow for more solar radiation to reach the forest
floor, drying remnant organic matter such as leaves and increasing the amounts of fuel in
post-logging sites (Johns et al. 1996). Furthermore, the release thresholds of pioneer
species, lianas and bamboo in the large gaps resulting from this scenario are unknown
(although conditions might be ideal for the regeneration of such species as S.
macrophylla) (Panfil and Gullison 1998). As such, the implementation of an increased
localized harvest intensity is risky, especially for a landowner limited to a management
unit of 300 ha.
Impact of Marking Treatment on FCT Damage
Despite positive results of marking FCTs in at least one other study (see Krueger
2004), marking did not appear to reduce FCT damage in the Porto Dias timber
operations. That FCT marking had no effect on damage levels along skid trails was
surprising, given that machine operators were from outside the community and therefore
not as familiar with the forest. Even so, the owners of the landholdings were responsible
for flagging the path of the skid trails prior to extraction, so as to avoid any damage
associated with skidding. This precaution could explain why marking had no effect on
FCT damage along skid trails in this particular study, as bulldozer operators were already
able to move without hesitation, especially when creating secondary and tertiary trails. As
well, most logs were bucked to about 5 m prior to skidding, perhaps avoiding some of the
damage that occurs when tree-length logs are skidded. Two local sawyers were also in
charge of felling operations, perhaps adding to the effectiveness of the operation, as they
are both experienced harvesters in Porto Dias, in addition to having lived for many years
in the reserve.
As more traditional forest dwellers begin to incorporate formalized timber
management into their livelihood strategies (which in the recent past have been based on
the exploitation of NTFPs), it will become increasingly important to monitor and
document these systems. Historically, most of the neotropical research and forest
legislation has focused on large-scale operations (Oliveira 2000). Smallholders will
necessarily differ from their industrial counterparts, given their low harvest intensities
and management regimes that favor a suite of forest products and services beyond timber.
Even though many general sustainable logging guidelines exist, one of the greatest
barriers to good silvicultural management is developing site-specific management
objectives and monitoring principles that recognize the heterogeneity inherent in tropical
forest systems (Putz 1996). As was shown in this study, the gap mosaic of the forest will
change dramatically in comparison to its pre-logging state, even when low intensity, RIL
guidelines are followed (Webb 1997). Accordingly, community forest managers must
establish their own management paradigm while building on the RIL model. Given the
series of poorly understood parameters outlined in this study (pre-harvest liana cutting,
post-logging stand dynamics in bamboo-dominated forest, etc.), it is suggested that
systematic post-harvesting monitoring accompany logging operations (already one of the
requirements for FSC certification), so that the benefits of recommended practices may be
evaluated by the managers at the appropriate scale. Pretty and Hine (1999) argue that
monitoring methods must be sensitive to local needs in order to work in the long term,
primarily due to limited financial and technical resources. It stands to reason then that for
monitoring purposes, local peoples' efforts may be better placed in protecting the larger
size classes of FCTs, particularly if the concessions are on a relatively short (30 years in
this case) cutting cycle. This particular study focuses on the adult and sub-adult size
classes (> 20 cm dbh) in the hopes of capturing information about the next harvest
generation. Many previous logging experiments have concentrated on seedling
regeneration, but these seedlings are unlikely to be recruited to harvestable size classes
within at least two subsequent harvest cycles (Clark and Clark 1992, Connell and Green
2000). There may be more of a solid commitment to participatory monitoring if the
individuals in question could easily project a young tree's value across thirty years (which
would be in their children's lifetime), as opposed to a seedling that may take up to 100
years to get to a harvestable size, if it survives at all.
Reducing damage is one of the most important components of a sustainable forest
management program (Johns et al. 1996). Direct damage to FCTs was shown to be slight
for this study, and indeed, it was considerably less than in other neotropical investigations.
Even so, damage could be reduced even further. The Porto Dias timber management
project is harvesting lightly (7.92 m3 ha-1), but the implementation of liana cutting,
cautious selection of both crop and seed trees and avoidance of stands heavily dominated
by Guadua could all help approach sustainable forest management, a concept which may
have the best chances of being executed by those who directly benefit from the natural
resource base on a daily basis.
Table 3-1. Commercial timber species inventoried during 2004 logging operations in the
PAE Porto Dias, Acre, Brazil. For those species harvested in 2004-2005, total
volume (m3 ha-1) is given.
Scientific Name Common Name Family harvested (m3)
cereieira Fabaceae 56.6
Astronium sp. 1
Cariniana sp. 1
Copaifera cf langsdorfii
Dipteryx sp. 1
jutai da folha
angelim da mata
Continued Table 3-1.
Scientific Name Common Name Family harvested (m3)
Martiodendrom elatum pororoca Caesalpiniaceae
Mezilaurus itauba itauba Lauraceae
Minquartia guianensis? quariquara branca Olacaceae
Myroxylon balsamum balsamo Fabaceae 2.5
Parkia pendula angico vermelho Mimosaceae
roxinho da folha
Peltogyne paniculatum r Caesalpiniaceae
Peltogyne sp. roxinho Caesalpiniaceae
Pouteria cf reticulata abiurana preta Sapotaceae
Pouteria sp. 1 abiurana roxo Sapotaceae
Pouteria sp. 2 abiurana abiu Sapotaceae
Qualea cf tesmanni catuaba Vochysiaceae 3.5
Simarouba amara marmpa preto Simaroubaceae
Tabebuia cf impetigenosa ipe roxo Bignoniaceae
Tabebuia spp. ipe amarelo Bignoniaceae 2.91
Tetragastris altissima breu vermelho Burseraceae
Vatairea spp. sucupira amarela Fabaceae
Vochysia sp. 1 cedrinho Vochysiaceae
Vochysia spp. guaruba Vochysiaceae
* volume recorded under one common name, but likely reflects more than one species of
Table 3-2. Classification of damage sustained by FCTs (future crop trees) during logging
operations (modified from Jackson et al. 2002 and Krueger 2004).
Damage type Bole Root Crown
Severe Snapped at base, Uprooted Loss > 66% crown
bent or leaning
Moderate Exposed cambium Exposed cambium Loss 33-66% crown
Minor bark scrape root scrape Loss < 33% of
Table 3-3. Summary of harvesting impacts in 2004 for the three logging blocks at PAE
Porto Dias, Acre, Brazil.
Impact variable Barrinha Barrinha Palestina
Area sampled (ha) 8.0 8.4 5.4
Number of harvested trees ha' 1.5 1.6 1.4
Percent of harvested trees > 70 cm dbh 27 50 64
Volume harvested (m3 ha-) 7.6 8.5 7.7
Number of logging gaps 14 13 11
Felling gap area (excluding bole) (m2 ha1) 545.6 433.4 476.6
Proportion of area damaged due to felling gap 5 4 5
(excluding bole) (%)
Mean gap size SD, excluding bole (m2) 389.7 361.2 433.3
S119 249 167
Skid trail area (m2 ha') 519.4 318.6 347.0
Proportion of area damaged due to skid trail (%) 5 3 4
Total area damaged (ha) 1.15 0.83 0.89
FCT damage rate
Total number of damaged FCTs 18 37 15
Basal area of damaged FCTs (m2) 0.34 0.63 0.48
Proportion of FCTs damaged (%) 12 17 15
Frequency of damage categories (%)
Minor or moderate crown, bole, or root damage 78 70 80
Severe crown damage 6 11 0
Severe bole damage 22 24 7
Severe root damage 6 3 27
Table 3-4. Results of a logistic regression for the probability of damage to future crop
trees (FCTs) at Porto Dias, Acre, Brazil during 2004 logging operations.
Shown are the estimates for five continuous covariates, including distance to
nearest stump, felling gap, and skid trail, neighborhood FCT density (number
of stems within a 25 m radius) and local felling intensity (number of trees
felled in the surrounding hectare); and for three categorical independent
variables, including liana presence, location in bamboo-dominated forest, and
marking treatment. Statistically significant variables at P < 0.05 are shown in
Variable Estimate Wald p
Log distance nearest stump -0.07 0.01 0.91
Log distance nearest felling -1.48 22.96
Log distance nearest skid 5
-2.25 28.37 _0.001
Log neighborhood FCT 0
-0.26 0.24 0.62
Local felling intensity 0.16 0.88 0.35
Liana presence (L) 0.31 2.79 0.09
Location in bamboo (G) -0.02 0.01 0.90
Marking treatment (M) -0.14 0.57 0.45
L*G 0.14 0.52 0.47
L*M 0.24 1.55 0.21
G*M -0.10 0.29 0.59
L*G*M 0.11 0.32 0.57
Figure 3-1. Photograph taken in PAE Porto Dias, Acre, Brazil, demonstrating the re-
curved thorn on a branch of a Guadua culm.
PAE Porto Dias, Acrelandia, Acre, Brasil
S Primary Road
- Secondary Road
- Creek & River
| Barrinha 3
0 2.5 5 10
Figure 3-2. PAE Porto Dias, with the four landholdings inventoried in the study, including the three (in blue) harvested in 2004.
Figure 3-3. Logging block of Barrinha 1, showing skid trails, harvested trees, and location of future crop trees (FCTs) in the zones of
Figure 3-4. Logging block of Barrinha 3, showing skid trails, harvested trees, as well as location of future crop trees (FCTs) in the
zones of impact.
Figure 3-5. Logging block of Palestina, showing skid trails, harvested trees, as well as location of future crop trees (FCTs) in the zones
70 80 90 100
DBH of harvested tree (cm)
Figure 3-6. Relationship between size of felled trees and the surface area of the
corresponding logging gap created during felling.
r = 0.26, P = 0.001
Redford and Stearman (1993) propose that despite claims that low-impact
extraction is compatible with biodiversity conservation, if a full range of genetic, species
and ecosystem diversity is to be maintained, then any significant activity by humans must
not be allowed. Advocates of SFM have countered this argument by stating that large-
scale preserves have a limited chance of succeeding in the face of restricted financial
resources, human population growth, and economic pressures (Putz et al. 2000, Pearce et
al. 2003). Who is correct? Or is there no one right answer? Instead, should areas managed
for timber resources be but one component of an interconnected model of complementary
nature, production and indigenous reserves, allowing for various degrees of protection
(Peres and Terborgh 1995)? The forest is going to be exploited, whether so-called
preservationists like it or not. The extent to which that exploitation may incur permanent
damage can be controlled in some ways by planned management activities. Even though
a strong link between biodiversity conservation and SFM may never be fully satisfied, the
global demand for tropical hardwood is not likely to subside anytime soon. This is
especially true of Brazil, which is simultaneously the world's largest producer and
consumer of tropical hardwood (May 2002).
Questions about economic and environmental sustainability still beleaguer
community-based timber projects' quest for staying power, though many see the timber
projects and forest managers as key assets in protecting forest ecosystems from
encroaching mechanized agriculture and cattle production, two of the greatest threats to
Amazon biodiversity (Nepstad et al. 2002, Kainer et al. 2003). With future studies, it will
be challenging to detect changes in the resource base given that many plant and animal
species are harvested simultaneously in tropical forests; ecological tradeoffs associated
with these multispecies systems are so complex as to warrant long-term trial-and-error
experiments (Becker and Ostrom 1995). Thus, complex community level changes may be
difficult to asses (Hall and Bawa 1993). Also, a critical factor to consider in this debate is
such ambiguous vocabulary as "sustainable," "biodiversity" and "ecosystem
management" (Ghazoul 2001). The inability of biologists to agree upon this lexicon
makes the practical application of sustainability guidelines by community forest
managers even more challenging. Consequently, there is a greater risk in resource
degradation if specific ecological problems are not readily identifiable.
The conclusions reached in this thesis suggest that forest dwellers may be able to
sustainably use common resources of the forest, but whether or not they do so will
depend on a wider setting, that of a shifting national and international politico-economic
climate (Cardoso 2002). Fortunately, particularly in the state of Acre, the government has
been extremely supportive of forest dweller communities, in addition to providing a
strong facilitative role. With the continued presence of external agencies, these
communities will stand a better chance at conserving forest resources.
One issue that is frequently mentioned in the literature but still lacks substantial
investigation is the idea that economic sustainability is not always consistent with
ecological sustainability (Hall and Bawa 1993). The links between sustainable
development and resource management must be studied further, as the rural poor do
indeed have an invested interest in sustaining the resource base, but at the same time they
must also increase and diversify their overall livelihood (Cardoso 2002). Ultimately, the
success of SFM will be contingent upon addressing the drivers of forest degradation and
deforestation. Conservation will not occur through SFM and forest certification alone;
local, regional, national, and international policy makers must also be engaged in this
process. Otherwise, SFM as a tool for biodiversity conservation has little chance of
Agrawal, A. A. and C. C. Gibson. 1999. Enchantment and disenchantment: The role of
community in natural resource conservation. World Development 27: 629-649.
Alder, D. and J. N. M. Silva. 2000. An empirical cohort model for management of terra
firme forests in the Brazilian Amazon. Forest Ecology and Management 130: 141-
Allee, W. C. 1949. Group survival value for Philodina roseola, a rotifer. Ecology 30:
Amaral, P. and M. A. Neto. 2000. Manejo florestal comunitario na Amaz6nia Brasileira:
Situacgo atual, desafios e perspectives. Brasilia: Instituto Internacional de
Educacgo do Brasil (IIEB). Electronic document, http://www.tech-
inform.de/taller regional/bibliografia/bibliografia.htm, accessed November 30,
Amaral, P. A. Verissimo, P. Barreto, and E. Vidal. 1998. Floresta para sempre: Um
manual para a produgco de madeira na Amaz6nia. World Wildlife
Fund/IMAZON/USAID, Belem, Para, Brazil.
Appanah, S. and F. E. Putz. 1984. Climber abundance in virgin Dipterocarp forest and the
effect of pre-felling climber cutting on logging damage. The Malaysian Forester
Asner, G. P., D. E. Knapp, E. N. Broadbent, P. J. C. Oliveira, M. Keller, and J. N. Silva.
2005. Selective logging in the Brazilian Amazon. Science 21: 480-482.
Awasthi, A., S. K. Uniyal, G. S. Rawat, and A. Rajvanshi. 2003. Forest resource
availability and its use by migratory villages of Uttatkashi Garwhal Himalaya
(India). Forest Ecology and Management 174: 13-24.
Banana, A. Y. and W. Gombya-Ssembajjwe. 2000. Successful forest management: The
importance of security of tenure and rule enforcement in Ugandan forests. Pages
87-98 in C. C. Gibson, M. A. McKean and E. Ostrom, eds. People and forests:
Communities, institutions, and governance. MIT Press, Cambridge, MA, USA.
Banuri, T. and F. Apffel-Marglin. 1993. Who will save the forest? Knowledge, power
and environmental destruction. Zed Books, London, UK.
Baraloto, C. and P.-M. Forget. 2004. Effects of post-logging microhabitats on
regeneration from seed: Implications for sustainable forest management. Pages
191-203 in S. Gourlet-Fleury, J.-M. Guehl, and O. Laroussinie, eds. Ecology and
management of a neotropical rainforest: Lessons drawn from Paracou, a long-term
experimental research site in French Guiana. Elsevier, Paris, France.
Barreto, P., P. Amaral, E. Vidal, and C. Uhl. 1998. Costs and benefits of forest
management for timber production in eastern Amazonia. Forest Ecology and
Management 108: 9-26.
Barreto, P., C. Uhl and G. Yared. 1993. O potential de producgo sustentavel de Madeira
em Paragominas, Para na Amaz6nia Oriental: Considerac6es ec6logicas e
econ6micas. 7 Congresso Florestal Brasileiro, Anais da Sociedade Brasileiro de
Engenheiros Florestais, pp. 387-392.
Bawa, K. S., D. R. Perry and J. H. Beach. 1985. Reproductive biology of tropical lowland
rainforest trees. I. Sexual systems and incompatibility mechanisms. American
Journal of Botany 72: 331-345.
Bawa, K. S. and R. Seidler. 1998. Natural forest management and conservation of
biodiversity in tropical forests. Conservation Biology 12: 46-55.
Becker, C. D. and R. Le6n. 2000. Indigenous forest management in the Bolivian
Amazon: Lessons from the Yuracare people. Pages 163-191 in C. C. Gibson, M. A.
McKean and E. Ostrom, editors. People and forests: Communities, institutions, and
governance. MIT Press, Cambridge, MA, USA.
Becker, C. D. and E. Ostrom. 1995. Human ecology and resource sustainability: The
importance of institutional diversity. Annual Review of Ecology and Systematics
Berkes, F. 1999. Sacred ecology: Traditional ecological knowledge and resource
management. Taylor and Francis, Philadelphia, Pennsylvania, USA.
Berkes, F. and C. Folke. 2002. Back to the future: Ecosystem dynamics and local
knowledge. Pages 121-146 in L.H. Gunderson and C.S. Holling, editors. Panarchy:
Understanding transformations in human and natural systems. Island Press,
Bertault, J.-G. and P. Sist. 1997. An experimental comparison of different harvesting
intensities with reduced-impact and conventional logging in East Kalimantan,
Indonesia. Forest Ecology and Management 94: 209-218.
Bobadilla, U. L. and S. F. Ferrari. 2000. Habitat use by Chiroptes satanas utahicki and
Syntopic platyrrhines in eastern Amazonia. American Journal of Primatology 50:
Boot, R. G. A. and R. E. Gullison. 1995. Approaches to developing sustainable extraction
systems for tropical forest products. Ecological Applications 5: 896-903.
Bowles, I., R. Rice, R. Mittermeier, and A. daFonseca. 1998. Logging and tropical
conservation. Science 280: 1899-1900.
Bray, D. B., L. Merino-Perez, P. Negreros-Castillo, G. Segura-Wamholtz, J. M. Torres-
Rojo, and H. F. M. Vester. 2003. Mexico's community-managed forests as a global
model for sustainable landscapes. Conservation Biology 17: 672-673.
Braz, E. M. and M. V. N. d'Oliveira. 1996. Planning to reduce damage. Tropical Forest
Update 6: 13-18.
Brokaw, N. V. I. 1982. The definition of treefall gap and its effects on measures of forest
dynamics. Biotropica 11: 158-160.
Bruenig, F. E. 1996. Conservation and management of tropical rainforests: An integrated
approach to sustainability. CAB International, New York, USA.
Bush, M. B. and P. A. Colinvaux. 1994. A paleoecological perspective of tropical forest
disturbance: Records from Darien, Panama. Ecology 75: 1761-1768.
Calheiros, D. F., A. F. Seidl and C. J. A. Ferreira. Participatory methods in environmental
science: Local and scientific knowledge of a limnological phenomenon in the
Pantanal wetland of Brazil. Journal of Applied Ecology 37: 684-696.
Camino, R. and M. Alfaros. 2000. Certification in Latin America: Experience to date.
Forest, Trees and People Newsletter 43: 25-27.
Cardoso, C. A. S. 2002. Extractive reserves in Brazilian Amazonia: Local resource
management and the global political economy. Ashgate Studies in Environmental
Policy and Practice, Hampshire, UK.
Carter, J. 1996. Recent approaches to participatory forest resource management. Rural
Development Forestry Network, Overseas Development Institute, London, UK.
Chapman, C. A., L. J. Chapman, R. Wangham, K. Hunt, D. Gebo, and L. Gardner. 1992.
Estimators of fruit abundance of tropical trees. Biotropica 24: 527-531.
Clark, D. A. and D. B. Clark. 1992. Life history diversity of canopy and emergent trees in
a neotropical rain forest. Ecological Monographs 62: 315-344.
Connell, J. H. and P. T. Green. 2000. Seedling dynamics over thirty-two years in a
tropical rain forest tree. Ecology 81: 568-584.
Dalling, J. W., H. C. Muller-Landau, S. J. Wright, and S. P. Hubbell. 2002. Role of
dispersal in the recruitment limitation of neotropical pioneer species. Journal of
Ecology 90: 714-727.
Daube, E., T. Fredericksen and M. Pefia. 2005. Sustainability of timber harvesting in
Bolivian tropical forests. Forest Ecology and Management 214: 294-304.
Deneven, W. M. 1992. The pristine myth: The landscape of the Americas in 1492.
Annals of the Association of American Geographers 82: 369-385.
Denslow, J. S. 1980. Gap partitioning among tropical rainforest trees. Biotropica 12: 47-
Dickinson, M. B., D. F. Whigham and S. M. Hermann. 2000. Tree regeneration in felling
and natural treefall disturbances in a semideciduous tropical forest in Mexico.
Forest Ecology and Management 134: 137-151.
Dykstra, D. P., and R. Heinrich. 1996. FAO model code of forest harvesting practice.
Food and Agricultural Organization of the United Nations, Rome, Italy.
Food and Agricultural Organization (FAO). 2005. State of the world's forests. Food and
Agricultural Organization of the United Nations, Rome, Italy.
Fearnside, P. M. 2003. Conservation policy in Brazilian Amazonia: Understanding the
dilemmas. World Development 31: 757-779.
Folke, C., F. Berkes and J. Colding 1998. Ecological practices and social mechanisms for
building resilience and sustainability. Pages 414-436 in F. Berkes and C. Folke,
editors. Linking social and ecological systems. Cambridge University Press,
Fox, J. E. D. 1968. Logging damage and the influence of climber cutting prior to logging
in the lowland Dipterocarp forest of Sabah. The Malayan Forester 31: 326-347.
Fredericksen, T. S. and F. E. Putz. 2003. Silvicultural intensification for tropical forest
conservation. Biodiversity and Conservation 12: 1445-1453.
Gerwing, J. J. 2001. Testing liana cutting and controlled burning as silvicultural
treatments for a logged forest in the eastern Amazon. Journal of Applied Ecology
Gerwing, J. J. and E. Vidal. 2002. Changes in liana abundance and species diversity eight
years after liana cutting and logging in an eastern Amazonian forest. Conservation
Biology 16: 544-548.
Ghazoul, J. 2001. Barriers to biodiversity conservation in forest conservation.
Conservation Biology 15: 315-317.
Ghazoul, J., K. A. Liston and T. J. B. Boyle. 1998. Disturbance-induced density-
dependent seed set in .\lh,mii siamensis (Dipterocarpaceae), a tropical forest tree.
Journal of Ecology 86: 462-473.
Gibson, C. C. and C. D. Becker. 2000. A lack of institutional demand: Why a strong local
community in western Ecuador fails to protect its forest. Pages 135-161 in C. C.
Gibson, M. A. McKean and E. Ostrom, editors. People and forests: Communities,
institutions, and governance. MIT Press, Cambridge, MA, USA.
Griscom, B. W. 2003. The influence of bamboo (Guadua sarcocarpa and Guadua
weberbaueri) on stand dynamics in lowland terrafirme forests of southeastern
Peru. PhD dissertation. Yale University, New Haven, CT, USA.
Griscom, B. W. and P. M. S. Ashton. 2003. Bamboo control of forest succession: Guadua
sarcocarpa in southeastern Peru. Forest Ecology and Management 175: 445-454.
Guariguata, M. R. and M. A. Pinard. 1998. Ecological knowledge of regeneration from
seed in neotropical forest trees: Implications for natural forest management. Forest
Ecology and Management 112: 87-99.
Gullison, R. E. and Hardner, J. J. 1993. The effects of road design and harvest intensity
on forest damage caused by selective logging: empirical results and a simulation
model for the Bosque Chimanes, Bolivia. Forest Ecology and Management 59: 1-
Hall, P. and K. Bawa. 1993. Methods to assess the impact of extraction of non-timber
tropical forest products on plant populations. Economic Botany 47: 234-247.
Hammond, D. S., S. Gourlet-Fleury, P. van der Hout, H. ter Steege, and V. K. Brown.
1996. A compilation of known Guianan timber trees and the significance of their
dispersal mode, seed size and taxonomic affinity to tropical rainforest management.
Forest Ecology and Management 83: 99-116.
Hardin, G. 1968. The tragedy of the commons. Science 162: 1243-1248.
Hartshorn, G. S. 1989. Application of gap theory to tropical forest management: Natural
regeneration in strip clearcuts in the Peruvian Amazon. Ecology 70: 567-569.
Heckenberger, M. J., A. Kuikuro, U. T. Kuikuro, J. C. Russell, M. Schmidt, C. Fausto,
and B. Franchetto. Amazonia 1492: Pristine forest or cultural parkland? Science
Heideman, P. D. 1989. Temporal and spatial variation in the phenology of flowering and
fruiting in a tropical rainforest. Journal of Ecology 77: 1059-1079.
Holmes, T. P., G. M. Blate, J. C. Zweede, R. Pereira, P. Barreto, F. Boltz, and R. Bauch.
2002. Financial costs and benefits of reduced-impact logging relative to
conventional logging in the eastern Amazon. Forest Ecology and Management 163:
Hubbell, S. P. 1979. Tree dispersion, abundance, and diversity in a tropical dry forest.
Science 203: 1299-1309.
Humphries, S. 2005. Forest certification for community-based forest enterprises in
Brazil's Western Amazon: Local stakeholders' perceptions of negative and positive
aspects and the certification process. MSc Thesis. University of Florida,
Gainesville, FL, USA.
Janzen, D. 1976. Why bamboos wait so long to flower. Annual Review of Ecology and
Systematics 7: 347-391.
Janzen, D. 2004. Setting up tropical biodiversity for conservation through non-damaging
use: Participation by parataxonomists. Journal of Applied Ecology 41: 181-187.
Jennings, S. B., N. D. Brown, D. H. Boshier, T. C. Whitmore, and J. do C. A. Lopes.
2001. Ecology provides a pragmatic solution to the maintenance of genetic
diversity in sustainably managed tropical rain forests. Forest Ecology and
Management 154: 1-10.
Johns, A. D. 1991. Responses of Amazonian rain forest birds to habitat modification.
Journal of Tropical Ecology 7: 417-437.
Johns, J. S., P. Barreto and C. Uhl. 1996. Logging damage during planned and unplanned
logging operations in the Eastern Amazon. Forest Ecology and Management 89:
Johnson, N. and B. Cabarle. 1993. Surviving the cut. World Resources Institute,
Washington, D.C. 71 pp.
Kainer, K. A., M. Schmink, A. C. Pinheiro Leite, and M. J. daSilva Fadell. 2003.
Experiments in forest-based development in western Amazonia. Society and
Natural resources 16: 869-886..
Krueger, W. 2004. Effects of future crop tree flagging and skid trail planning on
conventional diameter-limit logging in a Bolivian tropical forest. Forest Ecology
and Management 188: 381-393.
Lacerda, A. and G. Griffiths. 2005. Conservation and sustainability within Brazilian
Amazon forests: The role of botanical identification. Presentation given at the
Association of Tropical Biology and Conservation, Uberlandia, Brazil, July 24-28.
Lawrence, A. and F. S. Roman. 1996. The role of inventory in the communally managed
forests of Quintana Roo, Mexico. Pages 83-110 in J. Carter, ed. Recent approaches
to participatory forest resource management. Rural Development Forestry
Network, Overseas Development Institute, London, UK.
Lentini, M., A. Verissimo and L. Sobral. 2003. Forest facts in the Brazilian Amazon.
IMAZON, Belem, Brazil.
Lieberman, M. and D. Lieberman. 1994. Patterns of density and dispersion of forest trees.
Pages 106-119 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S.
Hartshorn, editors. La Selva: Ecology and natural history of a neotropical
rainforest. University of Chicago Press, Chicago, IL, USA.
Lorenzi, H. 2000. Arvores Brasileiras. Institute Plantarum de Estudos da Flora Ltda.,
Nova Odessa, Sao Paulo, Brazil.
Marcon, E. 2005. Ripley, version 3.1. http://e.marcon.free.fr/Ripley/, accessed December
Margoluis, R. and N. Salafsky. 1997. Measures of success: Designing, managing, and
monitoring conservation and development projects. Island Press, Washington, DC,
May, P. H. 2002. Forest certification in Brazil: Trade and environmental enhancement.
Consumer Choice Council, Washington, DC, USA.
Menton, M. C. 2003. Effects of logging on non-timber forest product extraction in the
Brazilian Amazon: Community perceptions of change. International Forestry
Review 5: 97-105.
Metzger, F. and J. Schultz. 1984. Understory response to 50 years of management of a
northern hardwood forest in upper Michigan. American Midland Naturalist 112:
Meyer, J. L. and G. S. Helfman. 1993. The ecological basis of sustainability. Ecological
Applications 3: 569-571.
Molnar, A. 2003. Forest certification and communities: looking forward to the next
decade. Forest Trends, Washington, DC, USA
Morellato, P. C. and H. F. Leitao-Filho. 1996. Reproductive phenology of climbers in a
southeastern Brazilian forest. Biotropica 28: 180-191.
Miller, U. 1998. Effects of refinements on the development of a neotropical rainforest
after controlled logging in Suriname. MSc thesis. Wageningen Agricultural
University, Wageningen, The Netherlands.
Nelson, B. W., A. C. A. Oliveira, G. T. Batista, P. Vidalenc, and M. Silveira. 2001.
Modeling biomass in the southwest Amazon by polar ordination of Landsat TM.
Pages 1683-1690 in Anais X SBSR, Foz do Iguacu, 21-26 April 2001, INPE.
Nepstad, D., A. Alencar, A. C. Barros, E. Lima, E. Mendoza, C. A. Ramos, and P.
Lefebvre. 2004. Governing the Amazon timber industry. Pages 388-414 in D.
Zarin, F. J. Putz, M. Schmink and J. Alavalapati, editors. Working forests in the
tropics: Conservation through sustainable management? Columbia university Press,
New York, NY, USA.
Nepstad, D., D. McGrath, A. Alencar, A. C. Barros, G. Carvalho, M. Santilli, and M. d.
C. Vera Diaz. 2002. Frontier governance in Amazonia. Science 295: 629-631.
Nyland, R. D. 2002. Silviculture concepts and applications. McGraw-Hill, New York,
Oliveira, M. V. N. d'. 2000. Sustainable forest management for small farmers in Acre
state in the Brazilian Amazon. PhD dissertation. University of Aberdeen,
Oliveira, M. V. N. d', E. M. Braz, D. F. R. P. Burslem, and M. D. Swaine. 1998. Small-
scale natural forest management: A new model for small farmers in the Brazilian
Amazon. Tropical Forest Update 8: 5-7.
Panfil, S. N. and R. E. Gullison. 1998. Short term impacts of experimental timber harvest
intensity on forest structure and composition in the Chimanes Forest, Bolivia.
Forest Ecology and Management 102: 235-243.
Parren, M. P. E. 2003. Lianas and logging in West Africa. PhD dissertation. Wageningen
University, Wageningen, The Netherlands.
Parren, M. and F. Bongers 2001. Does climber cutting reduce felling damage in southern
Cameroon? Forest Ecology and Management 141: 175-188.
Pearce, D., F. E. Putz and J. K. Vanclay. 2003. Sustainable forestry in the tropics:
Panacea or folly? Forest Ecology and Management 172:229-247.
Peres, C. A., C. Baider, P. A. Zuidema, L. H. O. Wadt, K. A. Kainer, D. A. P. Gomes-
Silva, R. P. Salomdo, L. H. Sim6es, E. R. N. Franciosi, F. C. Valverde, R. Gribel,
G. H. Shepherd, M. Kanashiro, P. Coventry, P. W. Yu, A. R. Watkinson, and R. P.
Freckleton. 2003. Demographic threats to the sustainability of Brazil nut
exploitation. Science 302: 2112-2114.
Peres, C. A. and J. Terborgh. 1995. Amazonian nature reserves: An analysis of the
defensibility status of existing conservation units and design criteria for the future.
Conservation Biology 9: 34-46.
Pereira, R. Jr., J. Zweede, G. P. Asner, and M. Keller. Forest canopy damage and
recovery in reduced-impact and conventional selective logging in eastern Para,
Brazil. Forest Ecology and Management 168: 77-89.
Pinard, M. A. and F. E. Putz. 1996. Retaining forest biomass by reducing logging
damage. Biotropica 28: 278-295.
Pinard, M. A., F. E. Putz, J. Tay, and T. E. Sullivan. 1995. Creating timber harvesting
guidelines for a reduced-impact logging project in Malaysia. Journal of Forestry