Managing Tropical Agricultural Landscapes to Enhance Biodiversity
Organized by:
 Robin L. Chazdon, Department of Ecology and Evolutionary Biology, University
of Connecticut
Traditionally, conservation of biodiversity in the tropics (and biodiversity studies in
general) have focused almost exclusively on reserves and protected areas, rather than
considering the broader agricultural landscape surrounding protected areas. Management
of land use, agricultural systems, and spatial configuration of forest cover within
agricultural landscapes can potentially mitigate some of the negative effects of forest
fragmentation and reduction of prime forested habitat. Speakers will focus on
applications of research to enhance biodiversity within agricultural landscapes in the
tropics. Many of these speakers were participants in the NCEAS Working Group on
Biodiversity and Conservation Value in Agricultural Landscapes of Mesoamerica,
organized by Robin Chazdon and Daniel Griffith (2004-2006); however, we have
attempted to broaden the scope of the symposium beyond Mesoamerica.
Maya agriculture and the forest matrix: intentional succession in the
Lacandon rain forest
Ronald Nigh
CIESAS Sureste
Email: rbnigh@gmail.com
Understanding agriculture in tropical secondary forest landscapes is inherently
interdisciplinary, focusing the interest of the ecological, physical and social sciences.
Emphasis is usually directed towards natural successional processes after human
disturbance, yet intentional human management of the forest matrix was central to the
history and genesis of Maya agriculture. To the present day, Mayan people farm in the
forest and their traditional methods seek to shape and conserve that environment.
Intervention in secondary succession after agriculture resulted in forest enrichment in
ethnobotanical species, accelerated the recovery of soil fertility, increased biodiversity,
helped control pests and enhanced landscape resilience. The Maya have viewed forests in
similar ways to ecologists classifying trees according to their longevity and abundance in
different successional stages. Our studies suggest that such Maya classes correspond to
functional groups of woody species. We consider cases of managed succession in the
eastern Chiapas and their potential of Maya practices to contribute to restoration in
degraded areas.
Keywords: Secondary succession, Maya agriculture, management, forest matrix
Traditional agricultural practices enhance agro-biodiversity and pest
management
Helda Morales
El Colegio de la Frontera Sur
Email: hmorales@sclc.ecosur.mx
Agriculture is often perceived as a threat to conservation. There is evidence that
agriculture based on monocrops, with heavy inputs of pesticides and fertilizers put
biodiversity and ecosystems services at risk. Nevertheless, here I show that traditional
agriculture practiced by small corn farmers from Chiapas and Guatemala may not only
enhance agro-biodiversity and pest management, but also may contribute to biological
conservation at the landscape level. Based on controlled field experiments and lab
bioassays, I explore the role of traditional agricultural practices, involving soil fertility
management, intercropping, weed managemen, and pest repellents, in the natural
regulation of potential pests and in conservation. The data show that traditional corn plots
maintain high diversity of plants, birds and insects, enhancing natural biological pest
control and preventing pest damage. Many of these traditional practices could be useful
in modern agriculture and could be promoted for conservation purposes. Agroecologists
can help bridge the communication gap between traditional farmers on the one hand and
conservationists and policy makers on the other.
Keywords: Agroecosystems, pest prevention, cultural practices, biological control,
biodiversity
Biodiversity, ecological, and economic services in coffee agroecosystems
Stacy Philpott1 and Ivette Perfecto2
1
University of Toledo
2
University of Michigan
Email: stacy.philpott@utoledo.edu
Coffee agroecosystems are extremely important for biodiversity conservation and
contribute to the economic well being of millions of producers. Dozens of studies have
highlighted the importance of maintaining diverse coffee agroecosystems (including high
tree diversity and density, canopy cover, epiphytes, etc.) for protecting biodiversity.
Maintenance of diverse coffee systems may further promote ecosystem services such as
pest control, protection from disease outbreaks, enhanced pollination, increased
connectivity between forest fragments, and buffering against the effects of global
environmental change. Coffee management choices also strongly influence coffee yields
and the alternative products provided from the shade trees which can contribute to farmer
incomes. We present here results of a quantitative synthesis of more than 40 studies
examining the relationship between biodiversity of ants, birds, and plants and coffee
management intensification. We also present a meta-analysis of the specific vegetation
and site characteristics of coffee management systems that strongly correlate with species
richness, especially that of forest species. We also briefly summarize the impacts of
coffee management changes on different functional groups of ants, trees, and birds and
discuss the implications for ecosystem services and economic outputs from coffee
agroecosystems. The results from these studies can provide researchers, conservationists,
and managers with information about which management factors most strongly affect
biodiversity, and how management for biodiversity may influence the availability of
alternative products, the prevalence of pests and disease, and the variation of coffee
yields. Based on these results, we discuss management recommendations for protecting
both biodiversity and economic and ecological value of coffee agroecosystems.
Keywords: biodiversity, ecosystem services, agroecology, conservation, sustainability
Quantification of avian use of a shaded coffee plantation as a biological
corridor
Oliver Komar
Salva NATURA Conservation Science Program
Email: okomar@salvanatura.org
Shaded agroecosystems, such as coffee plantations, are often assumed to function as a
biological corridor for forest birds, although this assumption has not been tested. I
recognize two types of avian users of such putative corridors: strict dispersers (forest
specialists that only occupy the corridor matrix while dispersing to new forest habitat)
and secondary corridor users (forest generalists that can breed within the corridor matrix).
Corridor use by strict dispersers may be rare, difficult to detect in nature. Quantifying
gene flow through corridors is methodologically challenging. I propose that long-term
avian monitoring stations, using mark and recapture methods, can provide data useful for
quantification of relative avian use of corridors. I provide an example from El Salvador
with 30 months of monitoring data (7145 net-hours) collected at a shaded coffee
plantation in the Apaneca biological corridor, which spans 30 km between El Imposible
and Los Volcanes national parks. Of 59 resident bird species captured at the station, at
least 12 (such as Cyanocompsa parellina) were strict dispersers from forests, being rarely
observed, and captured in <10% of sampling months. The 18 individuals of these species
represented 2.2% of resident bird captures, and were captured at a rate of 1 individual per
400 net-hours. Another 29 species, totalling 78% of resident bird captures and 37 birds
per 400 net-hours, were secondary corridor users, breeding in local forests as well as in
the corridor matrix. Probability of gene flow among forest patches is greatly increased for
these species because of breeding within coffee plantations. Relative use of corridors can
be compared across habitats for both types of avian corridor users, provided that
differences in capture probabilities due to habitat structure are accounted for.
Keywords: biological corridor; agroecology; coffee; monitoring data; birds
Farmers, tree cover and biodiversity conservation in agricultural
landscapes of Central America
Celia A. Harvey, Joel C. Saenz, Cristobal Villanueva, Rene Gomez, Marlon Lopez,
Muhammad Ibrahim and Fergus L. Sinclair
Conservation International
Email: c.harvey@conservation.org
By actively managing the forest and tree resources that occur on their land, farmers can
have an important impact on the ability of agricultural landscapes to conserve
biodiversity. Farmer decisions determine the availability and distribution of forest
patches, riparian forests and fallow areas; the densities, composition and spatial
arrangements of dispersed trees and live fences on farms; and the ways in which forests
and other tree resources are managed (such as the use of fire and the harvesting of
firewood or timber). Collectively these decisions shape the structure and composition of
agricultural landscapes, with potentially profound impacts on the resident biodiversity.
For example, when farmers select which naturally regenerating trees to maintain in their
pastures, they generally reduce the overall tree diversity and tree density, and skew the
species composition towards species that provide products to farmers (typically timber or
fodder species), thereby affecting both the quantity and type of habitat and resources
available to wildlife. Similarly, the planting of extensive networks of live fences can
greatly enhance the structural connectivity of landscapes, facilitating the movement of
some (but not all) animal species. Using interdisciplinary studies on the relationships
between farmer decisions, land management and biodiversity from 4 cattle landscapes in
Costa Rica and Nicaragua, we characterize the abundance and distribution of tree cover
within agricultural landscapes, highlight the ways in which farmer decisions shape the
forest and tree cover present in agricultural landscapes, and explore the impacts of these
changes for the plant and animal diversity present. Our results highlight the urgent need
to actively integrate farm management into landscape –level conservation planning and to
create new alliances with farmers that permit the long-term management of agricultural
landscapes for both conservation and production goals.
Rainforest trees outside forest fragments as seed sources and
regeneration nuclei for forest recovery
Javier Laborde, Sergio Guevara and Graciela Sanchez-Rios
Instituto de Ecología A.C.
Email: javier.laborde@inecol.edu.mx
Forest recovery in tropical pastures is limited by seed dispersal, mainly because the seed
dispersers of woody plants avoid deforested areas and also because the potential seed
sources of woody plants are too far away. In the fragmented landscape of Los Tuxtlas,
Mexico we fenced in five fig trees isolated in pastures. We monitored seed deposition
under their canopies over a year, and sampled soil seed content as well as the vegetation
that had established after three years of cattle exclusion. Dispersal distances were
estimated for captured seeds and established plants. Seventy tree and shrub species were
captured in seed rain with a cumulative density of 833 seeds/m2/yr. Most of the captured
seeds (>90%) were small (seed width <3mm). Soil seed content was dominated by
ruderal herbs and was particularly poor in woody plants. After three years of exclusion,
77 species of trees and shrubs had established with a density of 4.0 plants/m2. Seeds <7
mm in diameter were frequently dispersed more than 75 m across the pasture. Larger
seeds were dispersed over shorter distances and in much lower numbers, but once they
arrived at the fig trees, germination and establishment success was higher than for small
seeds. Rainforest trees left within pastures have an enormous potential for accelerating
forest recovery by acting as regeneration nuclei when cattle is excluded. Because of their
location, these trees are also extremely important seed sources for secondary succession,
provided that they are visited by pollinators and seed dispersers. Current management
practices can easily be modified to promote the preservation of rainforest trees within
active pastures, particularly in the vicinity of protected areas. The inclusion of rainforest
trees within agricultural areas can mitigate and even counteract the detrimental effects of
forest fragmentation, and should be part of landscape restoration and landscape design
initiatives.
Keywords: Dispersal limitation, landscape matrix, grasslands
Sustaining the genetic diversity of trees in managed landscapes
David Boshier, Tonya Lander and Paul Rymer
Oxford Forestry Institute, Department of Plant Sciences, University of Oxford
Email: david.boshier@plants.ox.ac.uk
Forest conversion to agriculture and other uses has dramatically reduced tropical forest
cover and left much forest as highly fragmented patches, with consequent impacts on
population sizes. A high proportion of tropical tree species appears to be naturally
outcrossing, with associated risks from inbreeding and small populations. Thus
maintenance of genetic diversity is vital for long term viability and adaptability of
populations. Fragmentation studies are dominated by neutral matrix and isolation by
distance models. Landscapes are viewed as islands, containing the only functional
habitat, surrounded by a homogenous matrix or hostile sea of non-habitat. The matrix is
only a distance to be crossed, with decreasing gene flow between island populations as
distance increases. A number of studies suggest that for many populations and
individuals of tropical trees, gene flow may be high across managed landscapes with little
forest cover. Trees in a range of agroecosystems may therefore play an important role in
the long term genetic viability of many tree species through; facilitation of gene flow
between existing reserves, conservation of particular genotypes not found in reserves
and/or the maintenance of minimum viable population numbers, provision of habitat that
facilitates gene flow in other tree species. Managed landscapes may, however, vary in the
degree to which they present a barrier to, or facilitate, gene flow, depending on the
ecological requirements of the organism and the ecological attributes of the matrix. This
may have significant implications for the way habitat corridors and biological reserves
are conceived of and designed as well as our ability to simultaneously manage landscapes
for conservation, economic and social value.
Keywords: fragmentation, landscape matrix, gene flow: corridors
Predicting the forest regeneration capacity of post-agricultural tropical
landscapes
Bruce G. Ferguson1, Douglas H. Boucher2, Miguel Martinez-Ramos3, Karen D.
Holl4, Daniel M. Griffith5, Alexandre Bonesso Sampaio6 and Deborah Lawrence7
1
El Colegio de la Frontera Sur
2
Hood Collage
3
Centro de Investigaciones en Ecosistemas, Universidad Nacional Autónoma de México
4
University of California, Santa Cruz
5
Saint Louis Zoo
6
University of Brasilia
7
University of Virginia
Email: bruhel@hotmail.com
Post-agricultural secondary forests are an increasingly important part of tropical
landscapes, and their rates of regeneration are quite variable. We develop a simple model
to predict how the rate of regeneration should vary with previous land use. The model is
based on two major independent variables, the availability of propagules and site quality,
which interact to determine the rate of regeneration of woody plant biomass. Comparing
among land use types, the model predicts lowest rates of regeneration for pastures and
highest for agroforestry and shifting cultivation. The model also predicts how rates of
biomass regeneration should vary within a land use type, depending on variables such as
the temporal and spatial extent of agricultural use. A review of the data available
indicates that the model is quite successful in predicting both between- and within-landuse-type differences in biomass regeneration rates. It can be extended to predict other
important aspects of regeneration, such as recovery of biodiversity and rate of reestablishment of late-successional species.
Keywords: biodiversity, ecological restoration, ecological succession, seed dispersal,
shifting cultivation
Spatial patterns of tropical forest recovery in an agricultural landscape
Karen D. Holl1, Rebecca J. Cole1, Rakan A. Zahawi2, Catherine A. Lindell3 and
Rebecca D. Fink4
1
University of California, Santa Cruz
2
Organization for Tropical Studies
3
Michigan State University
4
Duke University.
Email: kholl@ucsc.edu
Increasingly, large areas of post-agricultural land are becoming available for restoration
in the tropics. A better understanding of spatial patterns of tropical forest recovery is
needed in order to allocate limited restoration resource. We initiated a large-scale
restoration study: 1. to test planting tree islands of different sizes (applied nucleation) as
compared to planting the entire plot with trees (plantation), or control no planting; and 2:
to evaluate the relative importance of the amount of forest cover in the surrounding
landscape on forest recovery. During 2004-2006 we established 16 sites across a >100
km2 in southern Costa Rica. Each site has three 0.25 ha plots with one of the three
treatments and are surrounded by a forest cover range <5-50%. Thus far, we have
recorded tree seedling survival rates of 75-95% for the four species, and growth rates of
0.5-1.5 m per year. Only two years after treatment establishment we have recorded strong
effects on the number and types of birds using the restoration plots. During November
2006, 0.7, 1.2, and 1.7 individuals and 0.5, 0.7, and 1.1 species were captured per net
hour in the controls, islands, and plantations in 2-yr old sites, respectively. Birds were
more likely to visit larger tree islands. We have started detailed measurements of seed
rain and seedling establishment, and will monitor vegetation and dynamics of bird use in
these plots over the long term in order to inform tropical forest restoration efforts.
Keywords: restoration, Costa Rica, succession, landscape ecology
Restoring tropical forest biodiversity using late-successional tree species
Cristina Martinez-Garza1 and Henry F. Howe2
1
Universidad Autónoma del Estado de Morelos
2
University of Illinois at Chicago
Email: cmarti22@uic.edu
Fragmentation of tropical forest is accelerating at the same time that already cleared land
reverts to secondary growth. Early-successional environments are low-diversity stands of
the few species that are able to arrive (i.e., pioneer trees). Planting non-pioneer species
may help to bypass this low-diversity stage by overcoming dispersal limitation. Nonpioneer species will also provide a variety of food resources for animals that will
reestablish normal processes of seed dispersal into and out of the fragments. To develop
criteria for selection of non-pioneer species that perform best in early-successional
environments, we followed 24 non-pioneer species growing in an experimental plantation
near the Los Tuxtlas Biological Station in Veracruz, southeast Mexico. We measured the
growth and survival of these species together with their morphological and demographic
leaf traits under the different microhabitat of early successional environments: dark
understory of the secondary forest, edges and open pastures. We found that the best
criteria for selection of species are plasticity in leaf mass per unit area, leaf survivorship,
and leaf production. High plasticity in leaf mass per unit area is a powerful tool for
selecting mid-canopy species with potentially higher growth rates and survival across the
microhabitats of early-successional environments. High plasticity in leaf survivorship in
conjunction with homeostatic leaf production also led to higher survival and growth rates
in early-successional environments. Use of variability in morphological and demographic
leaf traits and other indices (e.g., maximum mature height) that are related to
performance may alleviate the need to individually screen large numbers of non-pioneer
species for restoration projects. Use of such easily assessed measures would free time and
resources for evaluation of other criteria, such as economic value or dispersal attributes
that influence animal populations. Enrichment of early successional environments that
includes as many species as possible will maximize diversity and complexity of
regenerating forests.
Keywords: Functional leaf traits, late-successional trees, Los Tuxtlas, Restoration
Ecology, SLM
The fate of dry Afromontane forests in northern Ethiopia
Frans Bongers1, Alemayehu Wassie1, Raf Aerts2, Tefera Mengistu3, and Frank
Sterck1
1
Wageningen University, Centre for Ecosystem Studies
2
Katholieke Universiteit Leuven, Division Forest, Nature and Landscape
3
Debub University
Email: frans.bongers@wur.nl
Dry afromontane forest in Northern Ethiopia are heavily deforested and fragmented, and
currently only small, isolated patches of forests remain in a matrix landscape of mainly
agricultural land. For long-term preservation of these forests effective protection is highly
needed, and restoration activities are necessary to improve quality and possibly extension
of the forested area into the agricultural matrix. In this paper we analyse the deforestation
patterns, the current situation of the remaining forest patches and a number of possible
restoration measures. The data come from a number of detailed studies in the area.
Current levels of forested area is between 1 and 5% and deforestation rates are high,
sometimes 30% in 30 years. Most forest patches remain in areas around churches. These
forests typically are small, in most cases smaller than 10 ha. The forests harbour
relatively large numbers of species but population sizes generally are small. Also, species
composition is rather different from forest to forest. General threats to these forests are
agricultural intrusion into the forest area, selective cutting of trees and extensive grazing.
Restoration efforts in and around these forests, in the form of enrichment planting (seeds
or seedlings), and soil management show that restoration is possible. Germination,
growth and survival of selected species is good. Shrubs can act as nursery plants for tree
seedlings. However, it is crucial that grazing animals are kept out of experimental
restoration plots. This is also shown by larger scale enclosures where animals are not
allowed to enter the enclosed area. Vegetation cover generally establishes relatively
quickly. Enclosures as such provide good opportunities. Attention will be given also to
peoples attitudes towards restoration measures. We will present and discuss the results of
these studies and their possible application for larger scale restoration
efforts in these afromontane forest areas.
Keywords: afromontane forests, restoration, forest islands, forest conservation, Ethiopia
Structural and functional dynamics of second-growth forests under
contrasting previous land use regimes: a long-term permanent sample
plot study in Northeastern Costa Rica
Robin L. Chazdon1 and Bryan Finegan2
1
University of Connecticut
2
Tropical Agricultural Centre for Research and Higher Education (CATIE)
Email: robin.chazdon@uconn.edu
We describe temporal dynamics in tree community structure and functional composition
in eight permanent monitoring plots (PMP) in second-growth rain forests of NE Costa
Rica. Four plots that were cleared and used for pasture for several years (PP) were
monitored annually for 10 yr, whereas four plots that were cleared, planted briefly, and
then abandoned (Light use plots, LUP) were followed at irregular intervals for 19 yr.
Seven plots were initially 12-25 yr post-abandonment (the older plots); one LUP was
initially 1 yr old. We evaluated functional composition using a classification of five plant
functional types (PFTs) developed by cluster analysis of tree species described by two
traits, height range (H) reached by adult trees and absolute annual diameter growth rate
(AGR, five categories, from long-term PMP data in logged mature forests of the study
area). Stand density ≥10 cm DBH (N) showed highly variable spatial patterns and
temporal dynamics within and among plots. Stand basal area (G), in contrast, was closely
related to initial stand age (R2 = 0.79), and increased consistently over time in all plots.
Previous land use did not affect dynamics in N or G. Stand density for PFT-1 (understory
tree species) and PFT-2 (slow or very slow-growing mid-canopy species) was generally
lower in PP than LUP plots. PFT-5 (canopy and emergent tree species with moderate to
very fast growth) was the most abundant type, largely accounting for dynamics in stand
G and G increment. PFT-4 (slow or very slowgrowing subcanopy and canopy species)
increased in G over time. PFT-5 was over-represented in both N and G in secondary
stands compared to mature forest, whereas other PFTs were underrepresented. Our
analysis suggests that second growth forests may differ more from mature forest in
functional than taxonomic composition of trees.
Keywords: agricultural landscapes, land use, plant functional types, secondary forest,
vegetation dynamics