See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/242731281

Biodiversity, profitability and vegetation

structure in coffee agroecosystems of central
Veracruz, Mexico

Article in Agriculture Ecosystems & Environment · January 2007

Impact Factor: 3.4 · DOI: 10.1016/j.agee.2006.05.023

CITATIONS READS

72 107

4 authors, including:

Robert H. Manson Jeffrey O. Sundberg

Institute of Ecology INECOL Lake Forest College
54 PUBLICATIONS 1,452 CITATIONS 7 PUBLICATIONS 227 CITATIONS

SEE PROFILE SEE PROFILE

Andrea Cruz-Angón
Comisión Nacional para el Conocimiento y …
23 PUBLICATIONS 639 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Andrea Cruz-Angón
letting you access and read them immediately. Retrieved on: 28 May 2016
 This article was originally published in a journal published by
Elsevier, and the attached copy is provided by Elsevier for the
author’s benefit and for the benefit of the author’s institution, for
non-commercial research and educational use including without
limitation use in instruction at your institution, sending it to specific
colleagues that you know, and providing a copy to your institution’s
administrator.
All other uses, reproduction and distribution, including without
limitation commercial reprints, selling or licensing copies or access,
or posting on open internet sites, your personal or institution’s
website or repository, are prohibited. For exceptions, permission
may be sought for such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial
 Agriculture, Ecosystems and Environment 118 (2007) 256–266
www.elsevier.com/locate/agee

Biodiversity, profitability, and vegetation structure in a

py
Mexican coffee agroecosystem
Caleb Gordon a,*, Robert Manson b, Jeffrey Sundberg c, Andrea Cruz-Angón a

co
a
Departamento de Ecologı́a y Comportamiento Animal, Instituto de Ecologı́a, A.C.,
Km 2.5 Carretera Antigua a Coatepec #351, Congregación El Haya, A.P. 63, Xalapa 91070, Veracruz, Mexico
b
Departamento de Ecologı́a Funcional, Instituto de Ecologı́a, A.C., Km 2.5 Carretera Antigua a Coatepec #351,
Congregación El Haya, A.P. 63, Xalapa 91070, Veracruz, Mexico
c
Department of Economics and Business, Lake Forest College, 555 N. Sheridan Rd. Lake Forest, IL 60045, USA
Received 5 July 2005; received in revised form 17 May 2006; accepted 22 May 2006
Available online 5 July 2006

Abstract
al
on
We studied the relationships of bird and small mammal species richness, composition, and abundance to vegetation structure and economic
profitability across a coffee intensification gradient in central Veracruz, Mexico. We conducted 2 years of point count censuses for summer
resident birds, 2 years of Sherman live trapping for small mammals, and gathered vegetation structure data at 147 sampling points distributed
rs

over 16 sites spanning a cultivation intensification gradient. We calculated net annual revenue per hectare as an index of profitability from
economic and management data collected during interviews with plantation owners/managers. Both the species richness and abundance of
forest-affiliated birds were significantly greater in floristically and structurally diverse ‘bajo monte’ coffee and forest compared with
commercial polyculture coffee, which was, in turn, significantly richer than statistically indistinguishable specialized shade and sun coffee.
pe

Mammal capture rates were extremely low at all but two sites. Forest bird species richness and abundance were explained by multiple linear
regression models that included statistically significant effects of shade cover, percent of trees with epiphytes, and canopy height. We found no
clear relationship between profitability and biodiversity, with biodiverse bajo monte coffee plantations ranking among the most profitable
under all price scenarios. The high profitability of biodiverse bajo monte coffee systems was not dependent on the inclusion of long-term
environmental costs or premium pricing systems. Our results demonstrate that high-biodiversity coffee cultivation can be compatible with
high profitability, and has significant potential for conserving biodiversity in coffee-growing regions, but only as a substitute for low-
biodiversity coffee cultivation, not forest.
r's

# 2006 Elsevier B.V. All rights reserved.

Keywords: Agroecology; Birds; Mammals; Coffee; Economics; Mexico

o

1. Introduction significance of coffee as a commodity, second only to oil

th

in terms of value of international trade for most of the post-

Increasingly over the past two decades, shade coffee World War II period (Ponte, 2002), and its biological
agroecosystems have drawn the attention of the conservation significance, as it grows in tropical lower montane forested
Au

biological community because of the seeming potential to
conserve biodiversity while maintaining economic produc-
tivity. This potential is underscored by the economic
* Corresponding author. Present address: Biology Department, Lake
Forest College, 555 N. Sheridan Rd. Lake Forest, IL 60045, USA.
Tel.: +1 847 735 6051; fax: +1 847 735 5194.
E-mail address: Gordon@lfc.edu (C. Gordon).
regions rich in endemic biodiversity. One of the most
significant features of coffee agroecosystems from the
standpoint of biological conservation is their tremendous
variation in vegetation structure and diversity. A generation
of field studies (recently reviewed by Donald, 2004) has
produced a general consensus that highly intensified sun
coffee systems have little biodiversity conservation value,
but there is still a great deal of controversy over the

0167-8809/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.agee.2006.05.023
 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
conservation value of shade coffee systems which may vary
widely in vegetation structure and diversity (Philpott and
Dietsch, 2003; Rappole et al., 2003a,b; O’Brien and
Kinnaird, 2003, 2004; Dietsch et al., 2004).
The sociopolitical and economic dimensions of coffee
cultivation and its alternatives lie at the center of this
controversy, but have received little study to date (but see
Nestel, 1995; Rice, 1999; Gobbi, 2000; Perfecto et al.,
2005). Because profit maximization is one of the primary
objectives of most coffee farmers, management recommen-
dations based solely upon their biological implications are
less likely to have an impact than those whose economic
implications can be explicitly understood. Coffee farmers
and policy makers need to know not only how much
biodiversity you can conserve in coffee and how to do it, but
how much it will cost.
That is, assuming it costs anything. The relationship
between the biodiversity and profitability of ecosystems is
usually assumed to be a trade-off: if you want to save
biodiversity, you have to pay for it, and conversely, that the
highest profits are achieved in low-biodiversity ecosystems
such as biologically simplified ‘‘conventional’’ agroecosys-
on
tems. However, this is not necessarily the case. Examples of
synergistic interactions between biodiversity and profit-
ability have been demonstrated in many agroecosystems,
and include the control of pest populations by a diverse and
rs
robust community of predatory, or ‘‘beneficial’’, arthropods,
increased pollination services to crop plants by native
pollinating insects, and increased soil nitrogen availability
pe
created by the N-fixing microbes of leguminous plants
(Vandermeer, 1995; Buchmann and Nabhan, 1997; Daily,
1997; Pimentel et al., 1997; Costanza et al., 1997; Perfecto
et al., 2004).
The explicit examination of the profit–biodiversity
relationship in coffee is still in its infancy, and it is essential
that this relationship continue to be explored. Perfecto et al.
r's
(1996) drew attention to the increased production costs of
intensified, or technified coffee, suggesting that in some
cases and depending on the price of coffee, these costs may
more than offset the revenue generated by the increased
o
yields of such systems. Gobbi (2000) found that conversion
of El Salvadorian coffee plantations to meet ‘‘bird-friendly’’
th
criteria would be financially viable for all plantation types
examined, yet this study did not examine biodiversity
explicitly, and this conclusion was dependent on farmers
Au
receiving a price premium for selling certified bird-friendly
coffee. In northern Chiapas, Mexico, Soto-Pinto et al. (2000)
found that shade cover was positively correlated with coffee
yield when shade cover was between 23 and 38%, with
coffee yield only dropping off when shade cover exceeded
50%. In the same system, Soto-Pinto et al. (2002)
documented reduced incidence of coffee leaf rust with
more shade strata, and reduced spontaneous herb growth
with increasing shade density, both of which would
contribute to greater net revenue in high shade systems.
Additional revenue provided by non-coffee products in these
257
systems has also been shown to improve net revenue (Liza
et al., 2003). Further west in Chiapas, Romero-Alvarado
et al. (2002) found no significant difference in coffee yields
between diverse and Inga-dominated shade plantations,
though the latter required one fewer annual round of weed
removal. In Indonesian coffee, Klein et al. (2003a,b)
documented a positive correlation between the species
diversity of pollinating bees and coffee pollination and fruit
py
set. Ricketts (2004) and De Marco and Monteiro Coelho
(2004) demonstrated similar enhancements of coffee
pollination in coffee plantations located in close proximity
to forest patches in Costa Rica and Brasil, respectively.
co
These examples provide some preliminary evidence that
the relationship between biodiversity and profitability in
coffee agroecosystems is complex, and may include
synergies as well as trade-offs. To the extent that there
are trade-offs, a rigorous understanding of profit–biodiver-
sity relationships in coffee will help target conservation
dollars most efficiently and effectively. To the extent that
al
there are synergies, cultivation techniques and policies can
be designed that could potentially generate significant gains
for both coffee producers and tropical biodiversity
conservation.
Herein, we present the results of a study designed to
address the following questions:
(1) How do the species richness, composition, and abundance
of native forest birds and small mammals vary across a
coffee intensification gradient in central Veracruz,
Mexico?
(2) What specific aspects of the vegetation structure of
coffee plantations are most closely correlated with
patterns of bird and small mammal diversity and
abundance across this gradient?
(3) How do management techniques, vegetation structure,
cost structure, coffee yield, and prices interact to shape
the relationship between biodiversity and profitability in
coffee agroecosystems?
2. Methods
2.1. Study system
We conducted this study in the central Veracruz coffee-
growing region, between 1000 and 1400 m above sea level
on the eastern slope of the central plateau of Mexico. Total
annual precipitation in this region varies between 1350 and
2200 mm, while the mean annual temperature is between 12
and 19.5 8C (Manson et al., 2004). There are three well-
defined seasons: the relatively dry cool season lasting from
October–November to March, a dry warm season during
April and May, and the wet warm season from June to
September–October (Williams-Linera, 2002). Tropical
montane cloud forest (Grubb, 1977; Rzedowski, 1996) is
the dominant land cover type in the region, covering 28% of
258 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266

the region of Veracruz above 800 m, but 12.7% of this
coverage (18,687 ha) was lost between 1984 and 2000 due to
conversion to crops, cattle pastures, and urban expansion
(Manson et al., 2004).
The state of Veracruz is the second most important
coffee-producing region in Mexico (State of Veracruz,
2004). The central Veracruz region contributed 42.25% of
the state total in coffee production in the 2000–2001 harvest,
coffee plantations included in this study, one (MF)
containing tall, mature forest in the transition zone between
cloud forest and lower montane forest, and the other (DF)
containing 15-year-old secondary regrowth forest.
2.3. Bajo monte sites
These two large (>100 ha) plantations are the most

and currently has 21,089 coffee producers covering a total of
58,712 ha, or 7.33% of the land surface dedicated nationally
to this crop (Mario Hernández Córdova, personal commu-
nication).
py
forest-like of the coffee plantations in this study. The shade
stratum at these sites consists of a diversity of forest remnant
trees. In this sense, they are similar to the ‘‘rustic’’ coffee
included in many previous studies of coffee agroecology.

During the spring of 2001, we selected and obtained
permission to use 16 study sites, spanning the range of
vegetation structural variation along the coffee intensifica-
tion gradient within this region (Table 1). Each site consisted
of a separate patch of forest, coffee plantation, or coffee
cultivation technique within a plantation, described in detail
below.
co
The notion of rustic coffee in the literature simultaneously
connotes floristic diversity and low socioeconomic status of
the producers. The bajo monte sites included in our study are
floristically diverse but are both located on large, capital-
intensive farms whose owners have access to, and employ
high-input agronomic techniques such as fertilization with
commercial agrochemicals, hence we feel it is significant to

2.2. Forest sites
on
Our three forest sites were all located in small fragments
(< 20 ha) of forest. The Xolostla site (XF) is larger than the
others but is long and thin, with a width not exceeding 1 km.
The other two forest sites consist of small (<10 ha) forest
rs
al
distinguish them from rustic coffee systems (Moguel and
Toledo, 1999). Coffee cultivation at the El Mirador site
(MB) is certified organic. This is the only plantation that
employs sheep grazing both as a means to control weed
growth, and as a source of alternative revenue. At the other
bajo monte site (DB), more conventional cultivation
techniques are employed, including modern varieties of

fragments contained within two of the large (>100 ha) coffee bushes planted at high density, and the application of

Table 1
pe

Biodiversity, management, and economic profiles of 16 study sites representing a coffee intensification gradient in central Veracruz, Mexico
Sitea Site Nc Bird Small Size Shade Epiphyte Synthetic Synthetic Fertilizer Labor Coffee
typeb spp. mammalsd (ha) e treef removal fertilizer herbicide cost g cost g yieldh
MF FO 2 20 1 (1) 5 Div. NA NA NA NA NA 0
DF FO 8 34 3 (11) 10 Div. NA NA NA NA NA 0
XF FO 10 30 9 (26) 20 Div. NA NA NA NA NA 0
DB BM 10 47 0 300 Div. No Yes No 1400 1100 4900
MB BM 10 43 3 (5) 74 Div. No No No 0 1492 5743
r's

ZP CP 10 32 0 100 Ent. Yes Yes Yes 562 4758 3648

EP CP 10 32 1 (3) 200 Inga No Yes Yes 250 3219 2940
NP CP 10 21 0 200 Inga Yes Yes Yes 250 3219 2940
OP CP 10 26 1 (1) 10 Inga No No No 1750 3500 5250
o

RP CP 10 30 4 (8) 21 Ent. No No Yes 0 2762 2548

DP CP 10 24 4 (12) 300 Inga No Yes No 1400 1100 4900
PP CP 10 26 0 5 Inga No No Part 325 2608 3065
th

AP CP 5 23 3 (3) 2 Inga No No No NA NA NA
MS SS 10 7 9 (67) 100 Inga Yes Yes Yes 1376 6515 NA
DS SS 12 30 3 (3) 300 Inga No Yes No 1400 1100 6370
TU SU 10 15 1 (1) 132 NA NA Yes Yes 1515 8282 7350
Au

a
Two-letter abbreviations for each study site are listed in Section 2.
b
Site type abbreviations are as follows—FO: forest; BM: bajo monte coffee; CP: commercial polyculture; SS: specialized shade coffee; SU: sun coffee.
c
N indicates the number of circles used to sample birds and mammals at each study site.
d
Small mammal data are the total number of species captured followed by total number of individuals captured in parentheses (see text for details of bird and
mammal sampling).
e
Size column refers to the size of forest patches for forest sites, and to the size of total coffee holdings of the plantation owner for all other sites except PP,
whose 5 ha includes the total coffee holdings of three separate owners, only one of whom reported using synthetic herbicides.
f
Shade tree column refers to the dominant genus of shade tree if one existed (div.: diverse, no clear dominant; Ent.: Enterolobium).
g
Fertilizer and labor costs are expressed in terms of Mexican pesos per hectare per year for the year leading up to the harvest season of October 2000–March
2001. Labor cost includes all labor up to, but not including harvest.
h
Coffee yield data are expressed in terms of kilogram of coffee cherries harvested per hectare, and are either multi-year annual averages (where such data
were available), or correspond to the harvest season of October 2000–March 2001.
 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
synthetic fertilizers. Both bajo monte coffee sites had been
planted to coffee for at least 10 years prior to the start of our
study.
2.4. Commercial polyculture sites
This is the most common type of plantation in central
Veracruz. All shade trees are planted, and most are
leguminous trees in the genus, Inga. The commercial
polyculture sites included in this study represent a variety of
different plantation sizes, shade tree species, densities, and
management techniques (Table 1). None are certified
organic, though three sites (OP, AP, and 2/3 of PP) use
no synthetic chemicals.
2.5. Specialized shade sites
This category corresponds to Moguel and Toledo’s
(1999) shaded monoculture. These two sites (MS and DS)
are characterized by very low diversity and density of shade
trees and high density of coffee bushes.
on
2.6. Sun coffee site
Our study contained a single farm representing this
cultivation system (TU), which is widespread in many
rs
countries and accounts for approximately half of the world’s
coffee production (Perfecto et al., 1996), but is relatively
rare in central Veracruz and in Mexico in general. It is the
pe
most ecologically simplified coffee agroecosystem, contain-
ing no shade stratum.
No plantation owners in this study reported using
synthetic insecticides, and none reported significant
production losses from any pests, though various owners
expressed concern about the insect pests coffee berry borer
(Hypothenemus hampei Ferr., Coleoptera: Curculionidae:
r's
Scolytinae) and ‘‘gallina ciega’’ (Phyllophaga spp.,
Coleoptera: Scarabaeidae), as well as the fungal pests
coffee leaf rust Hemileia vastatrix Berk and Br. and ‘‘ojo de
gallo’’ Mycena citricolor Berk and Curt.
o
At each study site, a series of permanent, non-over-
lapping 25-m radius sampling circles was established and
th
marked in the field with a fluorescent painted iron stake,
paint marks on trees (if available) and flagging tape. Ten
such circles were established at each site except where space
Au
did not allow (Table 1), for a total of 147 sampling circles
over the 16 study sites combined. We used this system of
circles as the basic sampling unit for all bird, mammal, and
vegetation data collection.
2.7. Bird censuses
We used 10-min, 25-m fixed radius point count censuses
(described in Greenberg et al., 1997a) as the basis for all
analyses of bird biodiversity patterns. This technique is
efficient for measuring the abundance of diurnal birds during
259
periods of high vocal activity, and has been used extensively
in previous studies of bird communities in coffee plantations
(Greenberg et al., 1997a,b; Tejeda-Cruz and Sutherland,
2004). In 2001 and 2002, two such point counts were
conducted on separate days between June 1 and August 17,
between 06:30 and 11:00 h at each of the 147 sampling
circles. As our index of abundance, we used the maximum
number of individuals of each species observed in any one
py
census at each circle in each year. This avoids double-
counting individual birds while reducing the impact of daily
variability due to weather or other factors.
co
2.8. Small mammal trapping
We sampled small mammals using Sherman live-traps
baited with vanilla-scented crushed oats. We conducted one
round of small mammal trapping in July–August of 2001,
and a second round in November 2002–January 2003. In
each round, we conducted three nights of trapping at each
al
study site. Four Sherman traps were placed 10 m from the
center of each sampling circle; one in each of the four
cardinal compass directions. In half of the sampling circles,
a fifth trap was placed directly in the center of the circle. This
yielded a total of 3969 trap-nights (2 rounds  147 sampling
points  3 nights  4.5 traps per sampling point). All
trapped animals were retrieved from the traps the following
morning and were then weighed, sexed, identified to species,
and then live-released at the point of capture.
2.9. Vegetation measurement
During the summer of 2001, we visually estimated the
following vegetation characteristics during a single visit to
each sampling circle: (1) number of tree species with at least
one trunk (minimum 10 cm diameter at breast height) rooted
inside the circle; (2) number of trees (10 cm dbh minimum)
rooted inside the circle; (3) average height of the canopy; (4)
percent shade from woody vegetation other than coffee
plants, ranging from 0 to 100%; (5) percent of non-coffee
trees (10 cm dbh minimum) containing at least 2% vascular
epiphyte cover on their main trunk and branches. If no trees
were recorded in the circle, a zero value was recorded.
2.10. Economic surveys
Between October and December 2001, C. Gordon
conducted in-person interviews with farm owners/adminis-
trators to gather all the economic data necessary to calculate
net revenue. Multi-year averages were used if available.
Where multi-year data were not available, we used data from
the year corresponding to the fall, 2000 spring, and 2001
harvest. In addition, we gathered data about basic coffee
cultivation management such as herbaceous vegetation
control technique and frequency, whether or not epiphytes
were routinely removed from shade trees, and whether or not
any synthetic fertilizers or pesticides were used.
260 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
2.11. Analyses
We used both years of bird point count census data in
conjunction with the vegetation data to perform multiple
linear regression analysis to describe the relationships
between bird and vegetation variables. We used information
on the general habitat affinities of each species compiled
from Howell and Webb (1995) and Stotz et al. (1996) to
classify all bird species recorded in the study as forest
species, generalist (forest or open habitat) species, or open
habitat species. In order to focus on the component of the
avifauna most significant for conservation, we performed
bird–vegetation regression analyses using forest bird species
only. Separate analyses were conducted using forest bird
abundance and forest bird species richness as dependent
variables.
To examine biodiversity–profitability relationships we
conducted simple linear least squares regression analyses
between profitability and biodiversity variables, with
ANOVA to test for the statistical significance of these
relationships. We used net revenue (NR), expressed in terms
of Mexican pesos per hectare per year, as our index of
on
profitability. NR is a simplified version of profitability that
only includes the most significant annual operating
expenditures and revenues. We calculated NR by subtracting
variable costs (VC) from total revenue (TR). VC was
rs
calculated as the sum of fertilizer costs plus labor costs. We
only included labor and fertilizer costs because, taken
together, they accounted for over 95% of the annual
pe
operating costs reported by plantation owners. TR was
calculated by adding total coffee revenue and non-coffee
revenue generated by the plantation. Significant non-coffee
revenue was only reported in monetary terms for two
plantations (MB and PP), though several others reported an
unquantified additional value of fruit grown on trees within
the plantations that is consumed by workers. Coffee revenue
r's
was calculated by multiplying the production (in kg/ha) of
o
th
Au
Fig. 1. Patterns of bird species richness across the 16 study sites, reflected by the total number of bird species recorded at each study site after two summer
seasons of sampling. Differently shaded sections of the bars represent bird species of different habitat affiliations (see legend and text). Each bar represents a
single study site, with the sites grouped by type, and types ordered left to right along an intensification gradient from most to least forest-like habitat (see x-axis
labels). The pattern is very similar if bird abundance is substituted for species richness. The correlation coefficient between forest bird abundance and forest bird
species richness is 0.78.
unprocessed coffee cherries, by the price under four different
scenarios listed below. Each of these price scenarios was
based on recent market fluctuations and actual market prices
that producers receive in this region.
Scenario 1: $1.4 p/kg, corresponding to the price in
January 2002 at the lowest point in the recent coffee price
crisis.
py
Scenario 2: $2.2 p/kg, corresponding to the price of
cherries in Veracruz in January 2003.
Scenario 3: $3.5 p/kg, representing a ‘‘normal’’ pre-crisis
price (Müller Grohmann, personal communication).
co
Scenario 4: $2.2 p/kg (the January 2003 price) was
assigned to most plantations, but a higher price of $3.5 p/
kg was assigned to plantations producing organic coffee
(MB, PP, OP, and AP). This represents the actual organic
premium price expected in 2004 by the owner of MB, the
only certified organic plantation in this study (Müller
Grohmann, personal communication).
al
All coffee-producing sites were included in these
analyses except for two (AP and MS) for which it was
not possible to obtain complete, quantitative economic data.
We performed separate biodiversity–profitability analyses
using forest birds (habitat affinity classification described
above), native mammals, and combined. We also performed
separate analyses using abundance and species richness
data.
3. Results
3.1. General avifaunistic patterns over the
intensification gradient
Our point count censuses yielded 2877 total observations
of 101 species of birds, including 54 forest species, 33
 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266 261

habitat generalist species, and 14 open habitat species. The Table 3

variation in bird species richness across study sites in this Multiple linear regression models of relationships between vegetation
variables and the species richness (top) and abundance (bottom) of for-
data set is depicted in Fig. 1. Both species richness (Fig. 1) est-affiliated birds (see text)
and bird abundance exhibited significant among-site
Coefficient t p(t)
variation across the coffee intensification gradient (one- 2
Forest bird species richness model (adjusted r = 0.47, p(F) < 0.001)
way ANOVA, p < 0.0001). Some coffee plantations
Epiphytes 0.016 3.05 0.003
exhibited forest bird species richness (Fig. 1) and abundance Percent shade 0.046 5.68 <0.001
equal to, or even greater than some forested sites. For forest Canopy height 0.096 2.26 0.025

py
birds, bajo monte coffee plantations were statistically Forest bird abundance model (adjusted r2 = 0.41, p(F) < 0.001)
indistinguishable from the forested sites, both of which Epiphytes 0.48 6.50 <0.001
were significantly richer than the commercial polycultures, Percent shade 0.48 4.64 <0.001
which were, in turn, significantly richer than the statistically

indistinguishable sun and specialized shade plantations.
This division of site types into three distinct clusters was
significant at the p < 0.05 level whether using species
richness or abundance of forest bird species (Tukey–Kramer
adjusted Student’s t-tests).
3.2. Small mammal patterns
co
epiphytes, and canopy height (Table 3). All three of these
variables exhibited statistically significant effects on species
richness of forest-affiliated birds (Table 3). Using forest bird
abundance as the dependent variable, only percent shade and
percent of trees with epiphytes exhibited significant effects
(Table 3). The overall explanatory power and significance of

A total of 3969 trap-nights yielded a total of 142 captures
of 13 species of small mammals, including 11 native and 2
on
introduced species. Captures were distributed highly
unevenly among study sites, with 47% of all captures
occurring at a single specialized shade site (MS), and an
additional 18% at one of the forest sites (XF) (Table 1). The
rs
al
the models were similar whether using forest bird species
richness (adjusted r2 = 0.47, ANOVA, p < 0.0001) or forest
bird abundance (adjusted r2 = 0.41, ANOVA, p < 0.0001) as
the dependent variable. Percent shade cover exhibited the
most significant effect on forest bird species richness, while
the percent of trees with epiphytes had the most significant
effect on forest bird abundance (Table 3). In both of these

low number and irregular pattern of captures did not permit models, the estimated coefficients and level of statistical
detailed statistical comparisons of species richness or significance were robust to the inclusion of other vegetation
abundance as a function of vegetation variables or across variables.
pe

sites or cultivation types. Inclusion of the mammal data

exhibited little influence on the biodiversity–vegetation or
biodiversity–profit analyses. Consequently, our discussion
of these analyses focuses primarily on the bird data.
3.3. Bird–vegetation relationships
r's
3.4. Biodiversity–profitability relationships
Linear least-squares regression analysis revealed that the
species richness of forest-affiliated birds was not correlated
in a simple, linear fashion to the profitability of coffee
plantations in our study system (Fig. 2). Linear relationships

The five vegetation variables included in our analysis of
bird–vegetation relationships exhibited a complex pattern of
intercorrelation (Table 2). To determine which vegetation
variables exerted the most significant effects on birds, we
o
between net revenue per unit area (NR) and biodiversity
variables were positive but insignificant for all four pricing
scenarios, whether using forest birds alone or combined with
mammals, and whether using forest bird species richness or

performed multiple linear regression analysis, generating abundance. Excluding the two bajo monte sites resulted in a
models that included only those vegetation variables with suggestion of a negative linear relationship between net
th

statistically significant effects ( p(t) < 0.05). Our models revenue and forest bird species richness, but this relationship
indicate that forest bird species responded positively to three was not statistically significant (ANOVA, p(F) = 0.090).
vegetation variables: percent of shade cover (all woody The relationship between forest bird species richness and
Au

vegetation taller than coffee plants), percent of trees with coffee yield, expressed as kg of cherries per hectare, was

Table 2
Correlation coefficients between and among bird and vegetation variables (see text for descriptions of variables)
Tree species Number of trees Canopy height Percent shade Epiphytes Forest bird spp. richness
Number of trees 0.68
Canopy height 0.62 0.38
Percent shade 0.52 0.55 0.59
Epiphytes 0.61 0.41 0.52 0.40
Forest bird spp. richness 0.44 0.35 0.55 0.63 0.49
Forest bird abundance 0.42 0.26 0.51 0.50 0.58 0.78
262 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
Fig. 2. Relationship between coffee plantation profitability and the species
richness of forest-affiliated birds. Net revenue calculation assumes price of
$2.2 p/kg of coffee cherries. Closed circles represent bajo monte sites, plus
signs represent commercial polyculture sites, open circle represents specia-
lized shade site, and open square represents sun coffee site. Substituting forest
bird abundance for species richness yields a virtually identical pattern.
negative but insignificant (Fig. 3, ANOVA, p(F) = 0.78).
Fig. 3 suggests a U-shaped relationship between coffee yield
and forest bird species richness, confirmed by quadratic
regression (ANOVA, p(F) = 0.019). This result is entirely
dependent on the inclusion of the two bajo monte sites. If
rs
these sites are excluded, the linear relationship between
yield and forest bird species richness is negative and
statistically significant (ANOVA, p(F) = 0.012).
pe
The profitability of coffee farms was significantly affected
by coffee price in our analyses (Table 4). Costs are assumed to
be independent of coffee prices, so that price fluctuations
affect the value of the output but not its production cost. As a
r's
o
th
Au
Fig. 3. Relationship between coffee yield and the species richness of forest-
affiliated birds. Coffee yield is expressed as kilogram of coffee cherries
produced per hectare. Closed circles represent bajo monte sites, plus signs
represent commercial polyculture sites, open circle represents specialized
shade site, and open square represents sun coffee site. Substituting forest
bird abundance for species richness yields a virtually identical pattern.
Table 4
Net revenue per unit area as a function of coffee price for four types of
coffee production systems in central Veracruz, Mexico
Production system Price 1 Price 2 Price 3 Price 4
Bajo monte (n = 2) 6448 10705 17623 14438
Specialized shade (n = 1) 6418 11514 19795 11514
Commercial polyculture (n = 7) 1505 4485 9328 5817
Sun coffee (n = 1) 493 6373 15928 6373
py
Data are expressed in terms of net Mexican pesos generated per hectare per
year. The number of study sites for each production system is indicated in
parentheses. Prices represent realistic market price scenarios, expressed in
terms of Mexican pesos per kilogram of unprocessed coffee cherries
(cereza) as follows (and see text): price 1 = $1.4 p/kg; price 2 = $2.2 p/
co
kg; price 3 = $3.5 p/kg; price 4 = $2.2 p/kg for regular coffee, $3.5 p/kg for
organic coffee.
result, those plantations with higher costs and higher yields
were much more sensitive to changes in coffee prices. In fact,
the highest yielding plantation (sun coffee) was the least
profitable at the lowest price, but almost as profitable as bajo
al
monte coffee at the highest price. The specialized shade site
was the most profitable plantation under price scenarios 2 and
3 (intermediate and high price), whereas bajo monte was the
on
most profitable in scenarios 1 (low price) and 4 (intermediate
price plus organic premium). Lower price scenarios also
rendered plantations with non-coffee income sources (MB
and PP) relatively more profitable.
4. Discussion
4.1. Contrasts among cultivation types
Our data suggest that most of the important variation in
biodiversity across the coffee intensification gradient lies not
in the sun versus shade dichotomy, but within the broad
category of shade coffee. The two specialized shade sites in
our study were highly depauperate in birds, statistically
indistinguishable from sun coffee with respect to the
abundance and diversity of forest-affiliated birds. The species
composition of these sites was similar to that of sun coffee,
dominated by open-habitat bird species such as Volatinia
jacarina and Aimophila rufescens. This result appears in
general agreement with the study of Greenberg et al. (1997b)
in central Guatemala, who found similarly depauperate
avifaunas in large, relatively homogeneous Inga-dominated
shade coffee plantations. However, it contrasts sharply with
the results of Greenberg et al. (1997a) and Tejeda-Cruz and
Sutherland (2004) in Chiapas, both of whom documented
much richer avifaunas in Inga-dominated shade plantations.
In both of these cases, the species richness and abundance of
birds in the Inga-dominated shade coffee was very similar to
that of rustic coffee.
The wide variation between studies in the avifauna of Inga-
dominated specialized shade coffee plantations has yet to be
resolved, but regional variation in landscape characteristics is
a likely explanation. Recent studies have noted that the effects
 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
of landscape variables on biodiversity in coffee plantations are
strong, particularly for mobile taxa such as birds and
Lepidoptera (Daily et al., 2001; Ricketts et al., 2001; Luck
and Daily, 2003; Perfecto et al., 2003; Tscharntke and Brandl,
2004; Tscharntke et al., 2005). In contrast to Chiapas, where
rich avifaunas have been found in Inga-dominated specialized
shade coffee, the central Veracruz landscape is much more
heavily deforested (Palacio-Prieto et al., 2000; Manson et al.,
2004). Central Veracruz may be a region essentially lacking
significant source populations of forest specialized birds to
drift into coffee plantations. Thus, the significantly richer
avifauna of commercial polyculture coffee plantations
relative to specialized shade coffee (both Inga-dominated)
recorded in our study is most likely explained as an actual
difference in the value of the former relative to the latter as a
habitat for forest bird populations.
Our data also suggest that bajo monte plantations have
significantly more value as bird habitats than do commercial
polycultures, the former being statistically indistinguishable
from small forest patches in terms of forest bird species
richness and abundance in our study. However, it is
important to note that the conservation value of bajo monte
on
coffee is not equivalent to that of forest for several reasons.
First, a significant component of cloud forest bird diversity
was entirely missing from all of our study sites. Many bird
species expected to occur in forests within this region were
rs
not recorded at all in our study. These species include rare
species such as the Bearded Wood-partridge (Dendrortyx
barbatus), and a variety of other species affiliated with
pe
forested habitats. The missing species also include 31 of 44
bird species with biogeographic distributions restricted
entirely (2 species) or with geographically disjunct
populations (42 species) restricted to the northeastern
Mexican cloud forest region (biogeographic data compiled
from Howell and Webb, 1995). Biogeographic restriction
may be a better indicator of ecological restriction than local
r's
habitat use patterns for birds (Gordon and Ornelas, 2000).
Many of these missing species may be forest-dependent
area-sensitive species, absent from small habitat fragments
and coffee plantations alike, not because of variation in
o
local-scale attributes of the habitat, but because of the
widespread deforestation of the landscape (Robinson et al.,
th
1995; Mitchell et al., 2001; Fahrig et al., 2002).
Second, compared to the biodiversity expected in forest,
bajo monte coffee is relatively depauperate in ground- and
Au
understory-dwelling taxa. Of the 17 forest bird species
recorded in forest but not in bajo monte coffee in this study, 9
are species that frequent the understory. This relative
underrepresentation of birds associated with the forest
understory is consistent with the results of Tejeda-Cruz and
Sutherland (2004). The abundance and diversity of ground-
dwelling small mammals was extremely low in all but one
coffee plantation in this study, and bajo monte appeared no
better than any other cultivation type in this regard. This is
consistent with the results of Daily et al. (2003) who found the
species richness of small mammals in coffee plantations in
263
Costa Rica to be similar to that of pasture, and much lower
than that of forest. These patterns suggest that while the
canopy of bajo monte coffee may harbor a significant
component of forest biodiversity, the understory is a more
disturbed, less forest-like microhabitat, supporting less forest
biodiversity. These results also highlight the importance of
multitaxon studies of biodiversity in coffee for making
broader inferences about the biodiversity conservation value
py
of various coffee agroecosystems (Perfecto et al., 2003).
4.2. Bird–vegetation relationships
co
Multiple linear regression models suggest that shade
cover, epiphyte abundance, and canopy height are positively
correlated to the species richness and abundance of forest
birds. These results are consistent with the findings of
Greenberg et al. (1997b) for the relationship of shade cover
and canopy height to bird abundance, and with Perfecto and
Snelling (1995), Perfecto and Vandermeer (1996), and
al
Armbrecht and Perfecto (2003) for the relationship of shade
cover to ant species richness. Our results also suggest that
epiphytes significantly enhance the bird conservation value
of coffee plantations, a pattern verified experimentally by
Cruz-Angón and Greenberg (2005).
4.3. Biodiversity–profitability relationships
Our results provide the first direct, quantitative estimation
of a profitability–biodiversity relationship in coffee. We do
not find any support for a trade-off between biodiversity and
profitability. In other words, biodiverse coffee agroecosys-
tems are not necessarily less profitable than low-biodiversity
cultivation systems (Fig. 2). Even factoring production costs
out completely and looking at the relationship between
coffee yield and forest bird species richness, the relationship
is still complex, and not well represented by a simple trade-
off (Fig. 3).
This conclusion has several important caveats. First, our
economic data are subject to a certain amount of error, owing
to the variation in the precision of record-keeping and
reporting among the plantation owners/managers included
in this study. However, there is no reason to suspect that this
variability should contain a systematic bias correlated either
with profitability or biodiversity, thus the qualitative nature
of the relationship, and specifically, the lack of a strong
biodiversity–profit trade-off is likely to be robust.
Second, the owners/managers of coffee plantations in this
study report yearly variation in coffee yields of well over
100%, indicating the need for multi-year studies of profit–
biodiversity relationships. Using multi-year average eco-
nomic data when such figures or estimates were available
(TU, DB, DS, and DP sites) reduced the impact of this
variability in the current study. However, the economic data
for the 2000–2001 harvest year that we used for the other
sites may represent specific conditions unique to this
particular year. For example, the sun coffee plantation
264 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
reported its highest ever yield in the 2000–2001 harvest,
almost double the average annual yield of this plantation for
the period from 1988 to 2000. Substituting this high-yield
figure into our analyses renders the sun coffee plantation the
most profitable under all price scenarios, but does not change
the qualitative nature of the overall profitability–biodiversity
relationship reflected by the profit–biodiversity regressions.
Based on these considerations of the strengths and
limitations of our economic data, we restrict our inference
to the conclusion that there is not a simple trade-off between
biodiversity and profitability.
5. Conclusion
The two bajo monte coffee plantations in our study system
stand as examples of highly profitable yet biodiverse coffee
agroecosystems. The most likely explanation for this result is
that high yields obtained through the use of modern
cultivation techniques do not necessarily reduce bird
biodiversity if the proper canopy vegetation structure is
maintained. Interestingly, petrochemical fertilizers are used at
on
one of the bajo monte sites in our study, but strictly organic
fertilization is practiced at the other. Both bajo monte sites
plant modern varieties of coffee at high densities typical of
modernized cultivation systems in the region. Policies and
rs
incentives targeted at helping farmers overcome the costs of
conversion from low-biodiversity systems to more biodiverse
systems modeled after bajo monte coffee cultivation may,
pe
therefore, generate simultaneous increases in the biodiversity
and profitability of coffee agroecosystems.
A simple management switch that would simultaneously
improve the profitability and biodiversity of coffee cultiva-
tion would be the elimination of the practice of ‘‘destencho’’
or epiphyte removal from shade trees. This practice,
common in this, and other regions of Latin America (Jones
r's
et al., 2000; Cruz-Angón, personal observation) incurs
significant labor costs, reduces biodiversity (Table 3; Cruz-
Angón and Greenberg, 2005), and is done at least in some
cases, according to the manager of one of the plantations
o
included in our study, based on the erroneous belief that
epiphytes parasitize the shade trees.
th
We would note that the economic benefits of both of these
recommended management changes are not dependent on
considerations of long-term degradation of soil or water
Au
quality. Such factors were not included in the current study, but
would confer additional economic advantages to cultivation
systems that sustain soil and water resource quality in the long
term by virtue of increased vegetation cover.
We also note that the economic strength of biodiverse
coffee cultivation systems in our study is not dependent on
premium pricing systems. Consumer demand for biodiver-
sity friendly coffee is growing but is still, and may always
remain, a small fraction of the coffee market (Messer et al.,
2000). Therefore, the potential of certification schemes to
promote high-biodiversity coffee management by capturing
demand for biodiversity among coffee consumers is
inherently limited on the demand side. On the supply side,
the costs and practical difficulties of implementing such
schemes in coffee-growing regions further limits their
potential to generate widespread improvement in the
biodiversity friendliness of coffee production. Such systems
generate a certain amount of incentive for increasing
biodiversity within coffee (Perfecto et al., 2005). However,
py
we suggest that there may be more to be gained both for
conservation and for coffee growers by dispelling the notion
that high-input, low-biodiversity sun and specialized shade
coffee cultivation systems are the most economically
co
sensible ways to grow coffee.
Despite the high coffee yields of intensified coffee
cultivation systems, they are economically risky for owners
because while their high costs are certain, high revenues are
uncertain, and contingent upon a high coffee price. The
recent coffee price crash on the international market caused
the net revenue of the sun coffee plantation in our study
al
system to drop by more than 30-fold relative to the pre-crisis
price, whereas the net revenue of the bajo monte plantations
fell by an average of 2.9-fold, and that of all other
plantations by an average of 6.7-fold. This crash has
triggered a major economic upheaval in coffee-growing
regions, including the abandonment of farms by over
300,000 coffee growers in Mexico (LaFranchi, 2001). While
this recent price dip was immediately caused by singular
events including the rapid growth of coffee production in
Vietnam (Stein and Burke, 2002), such wide price swings in
the international coffee market are nothing new. The history
of coffee price fluctuations on the international market is
pocked with price crises, generally attributed to the 4–5-year
time lag between coffee bush planting and berry production,
which makes it difficult for supply to track demand
(Pendergrast, 2000). More biodiverse, lower input systems,
particularly those with non-coffee revenue generators, incur
less economic risk, cushioning coffee growers against coffee
price dips. This economic advantage provides an additional
important conservation advantage. Low-risk coffee farms
are less likely to be converted to low-biodiversity land uses
such as cattle pasture, sugar cane, or housing development
when the price of coffee falls, as has happened in central
Veracruz in recent years (personal observation).
Finally, we caution that any policies promoting biodi-
verse coffee should not come as substitutes for forest
conservation policies or incentives, and should be carefully
crafted to discourage further deforestation in coffee-growing
regions. The extent of forest biodiversity that can be
sustained in biodiverse coffee agroecosystems is still not
well-known, but it is certainly less than that of intact forest,
particularly for area-sensitive species, and for ground- and
understory-dwelling species. For this reason, the potential
for enhancing biodiversity conservation in biodiverse coffee
lies entirely in the potential to substitute biodiverse coffee
cultivation techniques for low-biodiversity agroecosystems
such as sun and specialized shade coffee.
 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
Acknowledgments
Mammal field work was conducted by J. Eduardo
Martinez Leyva. Vegetation measurements and some bird
censuses were conducted by Julio C. Gallardo del Angel. We
wish to thank all of the coffee plantation owners and
managers for permitting us to include their plantations in our
study and for providing economic and management data.
Comments from I. Perfecto and three anonymous reviewers
on earlier drafts helped improve this manuscript. This study
was supported by NSF international postdoctoral fellowship
INT-0076201 awarded to CEG, and by a grant from
Mexico’s Environmental Ministry (SEMARNAT-CONA-
CyT 2002-C01-0194) to RHM. This study was also
supported by the personnel and facilities of CONACyT’s
Instituto de Ecologı́a, A.C., in Xalapa, Veracruz, Mexico.
References
Armbrecht, I., Perfecto, I., 2003. Litter-twig dwelling ant species richness
and predation potential within a forest fragment and neighboring coffee
plantations of contrasting habitat quality in Mexico. Agric. Ecosyst.
Environ. 97, 107–115.
Buchmann, S.L., Nabhan, G.P., 1997. The Forgotten Pollinators. Island
Press, Washington, DC.
Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B.,
Limburg, K., Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton,
rs
P., van den Belt, M., 1997. The value of the world’s ecosystem services
and natural capital. Nature 387, 253–260.
Cruz-Angón, A., Greenberg, R., 2005. Are epiphytes important for birds in
coffee plantations? An experimental assessment. J. Appl. Ecol. 42, 150–
pe
159.
Daily, G.C. (Ed.), 1997. Nature’s Services: Societal Dependence on Natural
Ecosystems. Island Press, Washington, DC.
Daily, G.C., Ehrlich, P.R., Sánchez-Azofeifa, G.A., 2001. Countryside
biogeography: use of human-dominated habitats by the avifauna of
southern Costa Rica. Ecol. Appl. 11, 1–13.
Daily, G.C., Ceballos, G., Pacheco, J., Suzán, G., Sánchez-Azofeifa, G.A.,
2003. Countryside biogeography of Neotropical mammals: conserva-
r's
tion opportunities in agricultural landscapes of Costa Rica. Conserv.
Biol. 17, 1814–1826.
De Marco, P., Monteiro Coelho, F., 2004. Services performed by the
ecosystem: forest remnants influence agricultural cultures’ pollination
and production. Biodivers. Conserv. 13, 1245–1255.
o
Dietsch, T.V., Philpott, S.M., Rice, R.A., Greenberg, R., Bichier, P., 2004.
Conservation policy in coffee landscapes. Science 303, 625.
th
Donald, P.F., 2004. Biodiversity impacts of some agricultural commodity
production systems. Conserv. Biol. 18, 17–37.
Fahrig, L.M., Freemark, K., Currie, D.J., 2002. Importance of patch scale vs.
landscape scale on selected forest birds. Oikos 96, 110–118.
Au
Gobbi, J., 2000. Is biodiversity-friendly coffee financially viable? An
analysis of five different coffee production systems in Western El
Salvador. Ecol. Econ. 33, 267–281.
Gordon, C.E., Ornelas, J.F., 2000. Comparing endemism and habitat
restriction in Mesoamerican tropical deciduous birds: implications
for biodiversity conservation planning. Bird Conserv. Int. 10, 289–
303.
Greenberg, R., Bichier, P., Sterling, J., 1997a. Bird populations in rustic and
planted shade coffee plantations of eastern Chiapas México. Biotropica
29, 501–514.
Greenberg, R., Bichier, P., Cruz-Angón, A., Reitsma, R., 1997b. Bird
populations in shade and sun coffee plantations in central Guatemala.
Conserv. Biol. 11, 448–459.
265
Grubb, P.J., 1977. Control of forest growth and distribution on wet tropical
mountains, with special reference to mineral nutrition. Ann. Rev. Ecol.
System. 8, 83–107.
Howell, S.N.G., Webb, S., 1995. A Guide to the Birds of Mexico and
Northern Central America. Oxford University Press, Oxford, UK.
Jones, J., Ramoni-Perazzi, P., Carruthers, E.H., Robertson, R.J., 2000.
Sociality and foraging behavior of the Cerulean Warbler in Venezuelan
shade-coffee plantations. Condor 102, 958–962.
Klein, A.M., Steffan-Dewenter, I., Tscharntke, T., 2003a. Fruit set of
highland coffee increases with the diversity of pollinating bees. Proc.
py
R. Soc. Lond., Ser. B 270, 955–961.
Klein, A.M., Steffan-Dewenter, I., Tscharntke, T., 2003b. Pollination of
Coffea canephora in relation to local and regional agroforestry manage-
ment. J. Appl. Ecol. 40, 837–845.
LaFranchi, H., 2001. Economic upheaval over coffee-rising production and
co
falling demand bankrupt farms from Indonesia to Mexico. Christian
Science Monitor August 15.
Liza, Y.K., Peeters, L., Soto-Pinto, L., Perales, H., Montoya, G., Ishiki, M.,
2003. Coffee production, timber, and firewood in traditional and Inga-
shaded plantations in southern Mexico. Agric. Ecosyst. Environ. 95,
481–493.
Luck, G.W., Daily, G.C., 2003. Tropical countryside bird assemblages:
richness, composition, and foraging differ by landscape context. Ecol.
al
Appl. 13, 235–247.
Manson, R.H., Williams-Linera, G., Monroy, R., 2004. El bosque de niebla.
Pronatura 6, 32–33.
on
Messer, K.D., Kotchen, M.J., Moore, M.R., 2000. Can shade-grown coffee
help conserve tropical biodiversity? A market perspective. Endangered
Species Update 17, 125–131.
Mitchell, M.S., Lancia, R.A., Gerwin, J.A., 2001. Using landscape-level
data to predict the distribution of birds on a managed forest: effects of
scale. Ecol. Appl. 11, 1692–1708.
Moguel, P., Toledo, V.M., 1999. Biodiversity conservation in traditional
coffee systems of Mexico. Conserv. Biol. 13, 11–21.
Nestel, D., 1995. Coffee in Mexico: international market, agricultural
landscape and ecology. Ecol. Econ. 15, 165–178.
O’Brien, T.G., Kinnaird, M.F., 2003. Caffeine and conservation. Science
300, 587.
O’Brien, T.G., Kinnaird, M.F., 2004. Conservation policy in coffee land-
scapes—response. Science 303, 625–626.
Palacio-Prieto, J.L., Bocco, G., Velázquez, A., 2000. Nota técnica. La
condición actual de los recursos forestales en México: resultados del
Inventario Forestal Nacional 2002. Investigaciones Geográficas, Boletı́n
del Instituto de Geografı́a, UNAM 43, 183–203.
Pendergrast, M., 2000. Uncommon Grounds: The History of Coffee and
How it Transformed Our World. Perseus Books, Boulder, CO.
Perfecto, I., Snelling, R., 1995. Biodiversity and the transformation of a
tropical agroecosystem: ants in coffee plantations. Ecol. Appl. 5, 1084–
1097.
Perfecto, I., Vandermeer, J., 1996. Microclimatic changes and the indirect
loss of ant diversity in a tropical agroecosystem. Oecologia 108, 577–
582.
Perfecto, I., Mas, A., Dietsch, T., Vandermeer, J., 2003. Conservation of
biodiversity in coffee agroecosystems: a tri-taxa comparison in southern
Mexico. Biodivers. Conserv. 12, 1239–1252.
Perfecto, I., Rice, R.A., Greenberg, R., Van der Voort, M.E., 1996. Shade
coffee: a disappearing refuge for biodiversity. Bioscience 46, 598–608.
Perfecto, I., Vandermeer, J., López Bautista, G., Ibarra Nuñez, G., Greenberg,
R., Bichier, P., Langridge, S., 2004. Greater predation in shaded coffee
farms: the role of resident neotropical birds. Ecology 85, 2677–2681.
Perfecto, I., Vandermeer, J., Mas, A., Soto-Pinto, L., 2005. Biodiversity,
yield, and shade coffee certification. Ecol. Econ. 54, 435–446.
Philpott, S.M., Dietsch, T., 2003. Coffee and conservation: a global context
and the value of farmer involvement. Conserv. Biol. 17, 1844–1846.
Pimentel, D., Wilson, C., McCullum, C.M., Huang, R., Dwen, P., Flack, J.,
Tran, Q., Saltman, T., Cliff, B., 1997. Economic and environmental
benefits of biodiversity. Bioscience 47, 747–757.
266 C. Gordon et al. / Agriculture, Ecosystems and Environment 118 (2007) 256–266
Ponte, S., 2002. The ‘latte revolution’? Regulation, markets and consump-
tion in the global coffee chain. World Dev. 30, 1099–1122.
Rappole, J.H., King, D.I., Vega, J.H., 2003a. Coffee and conservation.
Conserv. Biol. 17, 334–336.
Rappole, J.H., King, D.I., Vega, J.H., 2003b. Coffee and conservation III:
reply to Philpott and Dietsch. Conserv. Biol. 17, 1847–1849.
Rice, R.A., 1999. A place unbecoming: the coffee farm of northern Latin
America. Geogr. Rev. 89, 554–580.
Ricketts, T.H., 2004. Tropical forest fragments enhance pollinator activity
in nearby coffee crops. Conserv. Biol. 18, 1262–1271.
Ricketts, T.H., Daily, G.C., Ehrlich, P.R., Fay, J.P., 2001. Countryside
biogeography of moths in a fragmented landscape: biodiversity in
native and agricultural habitats. Conserv. Biol. 15, 378–388.
Robinson, S.K., Thompson III, F.R., Donovan, T.M., Whitehead, D.R.,
Faaborg, J., 1995. Regional forest fragmentation and the nesting success
of migratory birds. Science 267, 1987–1990.
Romero-Alvarado, Y., Soto-Pinto, L., Garcı́a-Barrios, L., Barrera-Gaytán,
J.F., 2002. Coffee yields and soil nutrients under the shades of Inga sp.
vs. multiple species in Chiapas, Mexico. Agrofor. Syst. 54, 215–224.
Rzedowski, J., 1996. Análisis preliminar de la flora vascular de los bosques
mesófilos de montaña de México. Acta Botánica Mexicana 35, 25–44.
Soto-Pinto, L., Perfecto, I., Castillo-Hernández, J., Caballero-Nieto, J.,
2000. Shade effect on coffee production at the northern Tzeltal zone
rs
pe
o r's
th
Au
of the state of Chiapas, Mexico. Agric. Ecosyst. Environ. 80,
61–69.
Soto-Pinto, L., Perfecto, I., Caballero-Nieto, J., 2002. Shade over coffee: its
effects on berry borer, leaf rust and spontaneous herbs in Chiapas,
Mexico. Agrofor. Syst. 55, 37–45.
State of Veracruz, 2004. http://www.veracruz-cafe.com/indice.php3
(accessed July 29, 2004).
Stein, N., Burke, D., 2002. Crisis in a coffee cup. Fortune 146 (12).
Stotz, D.F., Fitzpatrick, J.W., Parker III, T.A., Moskovits, D.K., 1996.
Neotropical Birds: Ecology and Conservation. University of Chicago
py
Press, Chicago, IL.
Tejeda-Cruz, C., Sutherland, W.J., 2004. Bird responses to shade coffee
production. Anim. Conserv. 7, 169–179.
Tscharntke, T., Brandl, R., 2004. Plant-insect interactions in fragmented
landscapes. Ann. Rev. Entomol. 49, 405–430.
co
Tscharntke, T., Klein, A.M., Kruess, A., Steffan-Dewenter, I., Thies, C.,
2005. Landscape perspectives on agricultural intensification and bio-
diversity-ecosystem service management. Ecol. Lett. 8, 857–874.
Vandermeer, J., 1995. The ecological basis of alternative agriculture. Ann.
Rev. Ecol. System. 26, 201–224.
Williams-Linera, G., 2002. Tree species richness complementarity, distur-
bance and fragmentation in a Mexican tropical montane cloud forest.
Biodiv. and Conserv. 11, 1825–1843.
al
on