Content uploaded by Kristina Cockle
Author content
All content in this area was uploaded by Kristina Cockle
Content may be subject to copyright.
Content uploaded by Kristina Cockle
Author content
All content in this area was uploaded by Kristina Cockle
Content may be subject to copyright.
-1
Presence and abundance of birds in an Atlantic forest
reserve and adjacent plantation of shade-grown yerba
mate, in Paraguay
KRISTINA L. COCKLE
1,3,*
, MARTY L. LEONARD
1
and
A. ALEJANDRO BODRATI
2,3
1
Department of Biology, Dalhousie University, Halifax, Nova Scotia, Canada, B3H 4J1;
2
Guyra
Paraguay, Comandante Rafael Franco 381 c/ Leandro Prieto, Asuncio
´
n, Paraguay;
3
Current
Address: Fundacio
´
n de Historia Natural Fe
´
lix de Azara, Pero
´
n 2933, Buenos Aires, Argentina, 1198;
*Author for correspondence (e-mail: kcockle@dal.ca; phone/fax: +54-11-4766-5833)
Received 6 November 2003; accepted in revised form 18 May 2004
Key words: Agriculture, Agroecosystem, Atlantic forest, Birds, Ilex paraguariensis, Shade-grown,
Yerba mate
Abstract. In the Atlantic forest region, there is a need to develop economic activities that can be
carried out in buffer zones around parks, with minimal impact on forest bird species. One such
possibility is the farming of yerba mate, Ilex paraguariensis, under native trees. We compared bird
species’ presence and abundance between a forest reserve and an adjacent plantation of shade-
grown yerba mate, to determine which species might use such plantations. Of the 145 species that
were regularly recorded in the forest, 66%, including five globally threatened species, were also
regularly recorded in the plantation. Most canopy species and tree trunk insectivores showed
similar abundance in both habitats, but forest floor and understory species were absent from the
plantation. Within the plantation, higher tree density did not lead to a greater abundance of forest
birds. Yerba mate grown under native trees could be used to rehabilitate cleared land and allow
recolonization by some Atlantic forest bird species.
Introduction
The Atlantic forest of southeas tern Brazil, northeastern Argentina, and
eastern Paraguay, is one of the world’s top five biodiversity hotspots (Myers
et al. 2000) and one of South America’s highest priorities for bird conser-
vation (Stotz et al. 1996; Stattersfield et al. 1998). Agriculture, cattle-ranch-
ing, and industry have replaced more than 90% of the Atlantic forest, mostly
within the last 30 years. The diverse bird community is threatened by high-
grade logging (Aleix o 1999), hunting (BirdLife International 2000), habitat
loss, and habitat fragmentation (Marsden et al. 2001), all of which have led
to local extirpations of formerly common species (Willis 1979; Christiansen
and Pitter 1996; Ribon et al. 2003). In total, 61 of the Atlantic forest’s 199
endemic bird species are endangered, vulnerable, or extinct in the wild
(IUCN 2002; endemism from Stotz et al. 1996). Indeed , this region now
contains more critically endangered birds than any other in the neotropics
(Stotz et al. 1996).
Biodiversity and Conservation (2005) 14:3265–3288 Springer 2005
DOI 10.1007/s10531-004-0446-0
Protected areas should be the top priority for conservin g Atlantic forest birds,
but, given the region’s large population, private land-ownership , and rapid
deforestation, it is also important to find economically viable activities that do
not involve total deforestation. For example, existing parks require buffer zones
in which economic activities are limited, and, preferably, compatible with
conservation. One such activity could be the production of shade-tolerant crops
under a tree canopy. Research in Africa, Asia, an d Latin America suggests that
such shade-grown crops can conserve some species of birds and other wildlife,
especially when planted under a diverse canopy of native trees and located near
native forest (Moguel and Toledo 1999; Rice and Greenberg 2000).
One of the Atlantic forest region’s most widespread crops is the native yerba
mate, Ilex paraguariensis, whose leaves are used to make mate (hot tea) and
terere (cold tea). Although it is almost always produced as a monoculture in
full sun, yerba mate can be grown in shade under native trees (Eibl et al. 2000).
For certified organic, shade-grown yerba, farmers in Paraguay receive three
times the price of traditional, sun-grown yerba, making shade-grown yerba an
economically viable option despite slightly lower yields (A. Pryor in litt.). Thus,
shade-grown yerba mate could be planted in buffer zones or biosphere reser ves
as a compromise between bird conservation and agriculture.
Despite the potential for yerba mate to be grown under native trees, it is not
clear how such plantations would be used by birds in the region. If shade-
grown crops are to be used in buffer zones, it is important to identify which
forest birds occur in such plantations, and how plantations may be managed to
promote bird conservation.
To our knowledge, only one study has examined the potential for shade-
grown crops to conserve Atlantic forest birds. In Brazil’s coastal Atlantic
forest, Alves (1990) found that, with the exception of tinamous (Tinamidae),
cracids (Cracidae), antbirds (Formicariidae), and manakins (Pipridae), most
bird families present in the forest were also represented in shade-grown cacao
plantations.
Here we confirm and extend Alves’ (1990) findings in a species-level study of
birds in an Atlantic forest reserve and adjacent plantation of shade-g rown
yerba mate. Our fir st objective was to compare the presence and abundance of
bird species in the forest reserve and adjacent plantation, and to compare
differences in presence and abundance between broad groups of birds char-
acterized by typical habitat, strata, and diet. Our second objective was to
compare two parts of the plantation that differed in tree density and canopy
cover, to determine wheth er higher tree density in the plantation led to higher
total abundance of birds in any of the ecological groups.
Methods
Our study was conducted between 16 October and 5 December 2001 (austral
spring) and 19 February and 22 April 2002 (late summer, early autumn) at
3266
Estancia Itabo
´
, Department of Canindeyu´ , Paraguay (2430¢S5438¢W,
elevation 300 m).
Study site
The study site was located within a 5000 ha tract of Atlantic forest and an
adjacent 80 ha plantation of shade-grown yerba mate, and included approxi-
mately 50 ha of the forest and 45 ha of the planta tion (Figure 1). Fifty hectare
plots are considered necessary to avoid missing rare species in diverse tropical
forest bird communities (Terborgh et al. 1990). The size and layout of the
plantation did not allow for replication of such large plots, and trails were too
close together to be independent; thus, we compare the bird community
between the entire 50 ha of forest and the entire 45 ha of plantation.
Figure 1. A map of the study site showing the forest, the plantation, the trails surveyed in spring
and autumn, and features of the surrounding landscape. Pale gray is degraded forest and capuera
(cleared areas that are now regenerating).
3267
Overall, the forest and plantation were similar in terms of elevation, slope,
tree species, density of standing-dead trees, and abundance of epiphytes
(Cockle 2003). Both had been subject to light logging 25 years before our
study. The plantation was created by removing the forest understory and some
trees, then planting yerba mate below the tree canopy. Thus, the forest had
more lianas and vines, and greater canopy cover than the plantation (80%
canopy cover in the forest compa red to 60% in the plantation; see Cockle 2003
for more details on the study site).
Within the plantation, we also compared a 12 ha subplot with high tree
density and canopy cover (343 stems/ha; 80% cover), to a 12 ha subplot with
low tree density and canopy cover (137 stems/ha; 50% cover; Figure 1).
In the forest, we surveyed birds along three trails (525, 1260, and 1000 m, for
a total length of 2785 m) that were separated by at least 300 m at all points,
and began 75 m from the plantation. In the plantation, we surveyed birds
along trails that formed a grid of 200 m · 200 m cells. We surveyed 15 cells in
the spring sampling period and nine in the autumn. Routes in both habitats
were marked every 25 m with numbered tags that served as reference points for
recording bird locations.
Sampling techniques
Where possible, we used spot-mapping to census birds (International Bird
Census Committee 1969; Bibby et al. 2000). We surveyed the forest and
plantation on alternate days, choosing our routes in advance and using dif-
ferent starting points and directions to balance for time-of-day within and
between sites. In total, we surveyed the entire study site six times in each of
spring an d autumn. Birds were never surveyed during rain.
Beginning 30 min before first light and for the next 3–4.5 h, we spot-
mapped birds along 1–3 km of trail per day, noting every individual or group
of birds heard or seen. We identified each bird to species based on songs, calls,
and/or visual observation, then estimated its distance and measured its
compass direction from our location. We periodically checked distance esti-
mates by visually confirming the location of a bird that had been detected
aurally. We could reliably estimate distance and direction for all but two
species (Chamaeza campanisona and Grallaria varia), which we did not
attempt to spot-map.
Where possible, we recorded sex, age (adult, immature, or dependent
fledgling), behavior (singing, feeding, carrying food, carrying nesting material,
begging, fighting, etc.), and whether two records of the same species were
simultaneous. Mixed species flocks were treated as a single record, but the
number and species of birds in the flock were recorded. For single-species
groups registered aurally, we noted simply ‘group’ and later substituted the
average group size for that species based on visual observations. We later
omitted birds flying in a straight line over the study site and all birds recorded
3268
outside the plot that we were surveying (e.g. birds calling in the forest while we
were surveying the plantation).
Nocturnal species were spot-mapped in the pre-dawn period of the
morning surveys, and in surveys on five clear moonlit nights (between dusk
and 02:00). On a further 16 nights we played back recordings of 13 nocturnal
species (for details see Cockle 2003); howeve r, we failed to detect any new
species using playback, so the results of these surveys are omitted here.
Playback was most useful in establishing the year-rou nd presence and terri-
toriality of Strix huhula and Strix virgata at our study site (for details see
Cockle 2003).
Analysis
A preliminary analysis showed that, with the exception of a few migratory
species, there was little variation between seasons in the presence and abun-
dance of birds, so we pooled data over the two seasons. Based on published
literature (e.g. Sick 1993; Stotz et al. 1996) and AB’s field experience, we placed
each species into one of the following three categories (see Appendix 1): (1)
forest species, (2) edge species (includes species associated with anthropogenic
habitats), and (3) aerial species (those that spend most of their active time in
the air). Forest species were then divided into five groups based on the strata in
which they are most often found: (1) canopy species, (2) midstory species, (3)
understory spec ies, (4) forest floor species, and (5) tree trunk species, and five
groups based on the predominant food items in their diet: (1) fruit-or-grain-
eaters, (2) fruit-and-insect-eaters, (3) insectivores, (4) nectarivores, (5) carni-
vores and carrion-eaters.
Qualitative measures
Species were considered to be occasional or accidental, and therefore excluded
from the qualitative analysis, if they were encountered on few er than five
occasions and either (a) used the site only as a stop-over during migration (e.g.
Elaenia albiceps), (b) were found only within 20 m of the edge of the habitat
(e.g. Thamnophilus caerulescens in the plantation), or (c) were visiting from
outside the study site (e.g. Syndactila rufosuperciliata visiting from nearby
gallery forest).
We tested the completeness of species lists by plotting species accumulation
curves for the forest and the plantation. Since species richness is affected by
plot size (James and Rathbun 1981), we calculated species richness using only
the part of the plantation that was surveyed in both seasons (43.8 ha) and an
equivalent sized, randomly selected portion of the forest (also 43.8 ha), here-
after referred to as the ‘main plots’. We assumed that most species at our study
site could be detected up to 100 m away. In order to reduce differences in
detectability between habitats, and to be certain that all our records fell within
the habitat we were surveying, we excluded birds detected beyond 100 m from
3269
our location. We plotted the accumulated number of species against the total
number of individuals accumulated in each of the two main plots. Since both
curves reached a plateau, species richness is compared based on the total
number of species accumulated in each of the main plots.
We calculated the Sørensen coefficient (Brower and Zar 1977) to examine
qualitative community similarity between the forest and the plantation, based
on the species found in the main plots:
CC
s
¼ 2c=ð s
1
þ s
2
Þ
where s
1
is the number of species in community 1 (forest), s
2
is the number of
species in community 2 (plantation) an d c is the number of specie s shared by
both communities.
For each habitat-, strata- and diet group, we calculated the number of
species found only in the forest (forest-restricted), as a proportion of all species
in the group. For example, a high proportion of forest-restricted species in a
given diet group suggests that birds dependent on the given food source
avoided the plantation. We tested wheth er different habitat-, strata-, and diet
groups had different proportions of forest-restricted species, using a v
2
analysis
of contingency tables with v
2
= 0.05 (Zar 1999).
Quantitative measures
Since we spent about equal time in each habitat (104 and 103 morning hours in
the forest and plantation, respectivel y), we compared the total num ber of
encounters (records) of each species between habitats, for each of the 123
species that was recorded more than 10 times, to get an index of abundance.
Because we covered the same routes every week, our records presumably in-
cluded repeat observations of the same individuals, thus violating the inde-
pendence assumptions of most statistical tests. Rather than test for statistical
differences between the forest and plantation, we considered a species to be
more abundant in the forest than in the plantation if there were at least twice as
many records of that species in the forest as in the plantation. To find out
whether habitat-, strata-, and diet groups differed in the proportion of species
that were more abundant in the forest than in the plantation, we used a v
2
analysis of contingency tables with v
2
= 0.05 (Zar 1999).
For 31 territorial species, we also estimated breeding density using stan-
dard territory mapp ing (International Bird Census Committee 1969) with
some modifications to adapt the method to the tropics (after Terborgh et al.
1990; Thiollay 1994; see Cockle 2003, for details). To calculate breeding
density, we divided the number of territories by the area surveyed, and
multiplied by 100 for a density expressed as N territories/100 ha. Since trails
were at least 200 m apart, we did not assume to have surveye d all points
between trails. Rather, we plotted a detection function (number of encounters
versus distance from trail) for each species, and used the shoulder of this
curve to determine the area that was effectively mapped (see Cockle 2003, for
details). We considered species to differ in density between the two habitats if
3270
their density was twice as high in one habitat as in the other (following
Marsden et al. 2001).
Comparisons within the plantation
To determine whether variation in tree density affected the use of the planta-
tion by birds, we compared two 12 ha subplots within the plantation. For
comparison, we selected a subset of the total visits to the subplots, balancing
for time of day, time of year, and location within the subplot (i.e. corner, edge,
or middle). Night birds were excluded. This exercise reduced our sample size,
so that we did not have enough observations to detect all species in each of the
two subplots, or to compare the number of encounters by species. Instead, we
compared the two subplots based on the total number of encounters of birds in
each habitat-, strata-, and diet group (adding up all the records of all species in
each group).
Results
In 207 h of morning spot-mapping and 29 h of night-surveys and playback, we
registered 13,752 contacts with 201 species of birds in the study site.
Figure 2. Total number of species recorded in the main plots in the forest (solid line) and in the
plantation (broken line) in relation to the total number of records of any species in the same plot.
3271
Species richness
In the main plots, we registered 4627 contacts in the forest and 4024 contacts in
the plantation. We detected 165 species of birds (occasional species omitted):
145 in the forest and 116 in the plantation (Figure 2).
Qualitative community similarity
In total, 49 of the 165 species recorded in the main plots were found only in the
forest, 96 were found in both habitats, and 20 were restricted to the plantation.
Thus, 66% of the birds present in the forest were also present in the plantation,
and Sørensen’ coefficient of community similarity was 0.74.
The proportion of species restricted to the forest differed significantly among
habitat groups (v
2
=16.7, df = 2, p < 0.001) and strata groups (v
2
=48.3, df
=4,p < 0.001), but not diet groups (v
2
=9.34, df = 4, p > 0.05; Table 1).
Forest birds, especially understory-, forest floor-, and midstory species, were
those most often restricted to the forest (Table 1).
Total numbe r of encounters for each species
Overall, 40% of species were encountered at least twice as many times in the
forest as in the plantation, and were considered to be more abundant in the
forest. We found significant differences among habitat groups (v
2
=25.1,
df= 2, p < 0.001), strata groups (v
2
= 39.2, df = 4, p < 0.001), and diet
Table 1. Percent (%) of species in each habitat-, strata-, and diet group, that 1) were restricted to
the forest, and 2) were more abundant in the forest than in the plantation (twice as many records in
the forest as in the plantation).
Restricted to the forest More abundant in the forest
Total 30 40
Edge species 3 0
Aerial species 0 0
Forest species 38 52
Canopy species 14 24
Midstory species 44 81
Understory species 88 100
Forest floor species 70 80
Tree trunk species 9 12
Fruit-or-grain eaters 20 17
Fruit-and-insect eaters 30 46
Insectivores 44 60
Nectarivores 0 100
Carnivores 67 –
3272
groups (v
2
=10.6, df = 3, p < 0.025) in the proportion of species that showed
higher abundance in the forest than the plantation (Table 1).
Estimated density
Of the 31 territorial species for which we estimated density, only two were
present in similar densities in both habitats (no more than twice as many
territories in one habitat as in the other; Figure 3). Of the remaining 29 species,
20 were more than twice as common in the forest, and 9 were more than twice
as common in the plantation.
Threatened species
During this study we recorded one Endangered species (Amazona vinacea;
lUCN, 2002), one Vulnerable species (Tinamus solitarius; IUCN, 2002) and five
near-threatened species (Dryocopus galeatus, Piculus aurulentus, Phylloscartes
eximius, Phylloscartes sylviolus, and Polioptila lactea; IUCN, 2002). All seven
species were encountered in the forest, and five were also encountered in the
plantation. Amazona vinacea, Phylloscartes sylviolus,andPolioptila lactea were
recorded at least twice as often in the plantation as in the forest.
Comparisons within the plantation
In the selec ted spot-mapping visits to the high- and low tree density subplots,
we recorded 3202 contacts with 106 species of birds. Bot h forest- and edge
birds were more common in the low tree density subplot (Figure 4). Among
forest birds, canopy species and fruit-or-grain-eaters were markedly more
common in the low tree density subplot, while understory species wer e more
common in the high tree density subplot (Figure 4).
Discussion
The composition of the bird community differed between the shade-grown
yerba plantation and the forest, but 66% of the species using the forest
also used the plantation. Forest birds were both more abundant and more
diverse in the forest, while edge species were more abundant and more
diverse in the plantation. Never-the-less, five IUCN-listed forest species
were encountered in the plantation (of seven recorded in the forest).
Midstory, forest floor, and understory species, in particular, were less di-
verse and less abundant in the plantation. In contrast, canopy- and tree
3273
trunk birds showed high community similarity and similar numbers of
encounters in both habitats, suggesting that most species in these groups
are able to use the plantation habitat. We found stronger differences be-
tween strata groups than between diet groups, but, among forest birds,
fruit-or-grain-eaters were found more often in the plantation compared to
the other diet groups.
Within the plantation, both fores t- and edge species were most abundant
where tree density was low. Among forest species, canopy birds made up
the majority of the records and were most common where tree density was
low, while the few understory and midstory birds were more abundant
where tree density was high. Among forest canopy species, fruit-and-grai n-
eaters were the only diet group with a large difference in abundance be-
tween the two tree densities, with a higher abundance where tree density
was low.
There are four main sources of bias that could affect our results. (1)
The layout of survey routes differed between the plantation (grid) and the
forest (long trails), so that rare species wer e more likely to be absent from
the plantation by chance. We minimized this bias by using large plots (see
Terborgh et al. 1990). (2) We spent considerable time surveying along the
Figure 3. Number of territories per 100 ha, for 31 species, in the forest (black) and plantation
(white). 1 = Basileuterus culicivorus,2=Myiornis auricularis,3=Conopophaga lineata,
4=Synallaxis ruficapilla,5=Crypturellus obsoletus,6=Basileuterus leucoblepharus,
7=Platyrinchus mystaceus,8=Dysithamnus mentalis,9=Thamnophilus caerules-
cens,10=Hemitriccus diops,11=Trichothraupis melanops,12=Capsiempis flaveola,
13 = Otus atricapillus,14=Synallaxis cinerascens,15=Lathrotriccus euleri,16=Glaucidium
brasilianum,17=Otus choliba,18=Leptopogon amaurocephalus,19=Corythopis delalandi,
20 = Habia rubica,21=Automolus leucophthalmus, 22 = Myiodynastes maculatus,23=Mio-
nectes rufiventris,24=Contopus cinereus,25=Pitangus sulphuratus,26=Troglodytes aedon,
21 = Megarynchus pitangua,28=Colonia colonus,29=Camptostoma obsoletum,30=Myioz-
etetes similis,31=Falco sparverius.
3274
edge of the plantation, but almost no time at the edge of the forest.
This biased our results toward a lower number of encounters in the planta-
tion compared to the forest, because we ignored birds detected outside of the
habitat we were surveying. (3) Visibility (and hence detectability) was higher
in the plantation than in the forest. (4) Our study considered only one
plantation and only one forest site, so our results cannot be generalized.
Figure 4. Number of bird records in low- (white) and high- (stippled) tree density subplots within
the plantation, for each of the habitat-, strata-, and diet groups. Nectarivores are excluded because
there were no canopy nectarivores at our study site.
3275
Shade-grown crops as habitat for forest birds
With some exceptions, our species-level study was consistent with the results of
Alves’ (1990) study that examined families of Atlantic forest birds in a shade-
grown crop. Consistent with our resul ts, Alves (1990) found that, compared to
nearby Atlantic forest, plantations of shade-grown cacao supported lower
abundance of antbirds (Formicariidae and Thamnophilidae), tinamous (Ti-
namidae), and manakins (Pipridae), most of which are forest floor and
understory species. However, our study at the species-level revealed some
interesting differences that would not have been detected in the family level
study. For example, although both studies found that tyrant-flycatchers (Ty-
rannidae) as a family were at least as abundant in the shade-grown crops as in
the forest (see Appendix 1), our study found that only tyrant-flycatchers of the
forest canop y and forest edge wer e more abundant in the plantation than
the forest, while flycatchers of the forest floor and understory were absent from
the plantation. Despite differences between species for several fami lies, our
results support the general conclusions of Alves’ (1990) study that, although
some forest species are absent, birds are abundant in shade-grown cro ps with
native trees, adjacent to native Atlantic forest.
Compared to most other studies of birds in shade-grown crops (Greenberg
et al. 1997a,b; Calvo and Blake 1998), our study revealed a higher proportion of
forest birds in the plantation. In large part, this may be explained by the
proximity of native forest to our plantation. Surveys in plantations within
deforested landscapes rarely find birds associated with interior forest (Greenberg
et al. 1997a,b; Calvo and Blake 1998). In contrast, near large tracts of natural
forest, other studies have found between 25 and 62% of forest bird species in
shade-grown crops (Terborgh and Weske 1969; Thiollay 1995; Canaday 1997).
While proximity to native forest is important, several other factors also play a role
in determining the bird diversity of plantations. High canopy cover and tree density,
for instance, may allow more forest birds to use a shade-grown crop (Greenberg
et al. 1997b; Calvo and Blake 1998). Contrary to this observation, we found that
forest birds (other than a few midstory and understory birds) were more common in
the plantation subplot with lower tree density, suggesting that, at the tree densities
we studied, high canopy cover and tree density were less important than other
factors in allowing Atlantic forest birds to use the shade-grown yerba plantation.
Other authors have predicted that forest birds will be most abundant
in plantations with structural and floristic diversity and abundant edible fruit
(Moguel and Toledo 1999; Rice and Greenberg 2000), We find support
for these hypotheses when we consider the combined results of several field
studies, including ours. Plantations with structural and floristic diversity
and abundant edible fruit (incl uding plantations, like ours, where the canopy
consisted of remnant forest trees; Alves 1990; Greenberg et al. 1997b;
Calvo and Blak e 1998) tend to support a greater diversity of birds, particularly
canopy frugivores, compared to less diverse plantations without fruit (Thiollay
1995; Cal vo and Blake 1998).
3276
Management considerations
Although the shade-grown yerba plantation did not support forest understory
or midstory birds, it co ntained nearly all of the canopy- and tree trunk species
from the nearby forest, including five globally threatened and near-threatened
species. These results suggest that shade-grown yerba mate may be an
appropriate land-use for buffer zones around reserves in the Atlantic forest.
Our study considered only one plantation, and, therefore, we do not know
whether our results can be generalized. That said, the Atlantic forest requires
urgent conservation action, so we suggest an ‘adaptive management’ approach
to shade-grown yerba mate.
Shade-grown yerba mate could be especially beneficial, and less damaging to
existing forest, if used to rehabilitate some of the land that has already been
deforested, including plantations of yerba mate currently grown in the open.
Planting a wide diversity of native forest trees might return some of the
structural complexity and floristic diversity of natural forest to previously
cleared land, allowing forest birds to re-colonize areas from which they are
currently excluded. Agronomically this appears to be feasible. Eibl et al. (2000)
found promising yields for yerba mate grown alongside native tree seedlings on
abandoned agricultural land, and another study is underway to explore the
costs of reforestation with yerba mate and other native trees, on land adjacent
to Iguazu´ National Park in Argentina (S. Holz, in litt.).
In our plantation, a high tree density subplot (343 stems/ha; 80% canopy
cover) did not support more forest birds than a low tree density subplot (137
stems/ha; 50% canopy cover), suggesting that increasing canopy cover beyond
certain levels may not lead to increased abundance or diversity of forest birds.
Further studies in the Atlantic forest should aim to confirm or refute these
results, and to examine other ways to increase the bird conservation value of
shade-grown crops.
Acknowledgements
We especially thank R. Clay, M. Vela
´
zquez, and H. del Castillo, for help with
background information, project planning, and logistics, and A. Horn,
I. McLaren, B. Freedman, C. Staicer, R. Cavalcanti, and anonymous review-
ers, for helpful comments on the manuscript. R. Ribon and S. Holz provided
copies of their unpublished reports and articles in press. We thank Chololo
´
SRL and Guayakı
´
Yerba Mate for logistical support and access to the field site.
The project was financed by an Innovative Research Grant from the Canadian
International Development Agency, a Bergstrom Memorial Research Grant
from the Association of Field Ornithologists, an NSERC post graduate
scholarship with a 5NR supplement, a grant to KC from the Patrick Lett
Fund, and an NSERC Discovery grant to ML.
3277
Appendix 1. List of species detected in the plantation and forest, with habitat groups, strata groups, diet groups, special conservation status, presence/
absence in the plantation and forest, estimated density in each habitat, and total number of encounters in each habitat.
Species
A
Habitat
D
Strata
E
Diet
F
IUCN Status
G
(endemic spp.)
H
Presence and density
I
N territories/100 ha
(N territories total)
Total N birds
within 100 m of
observer
Forest Plant’n Forest Plant’n
Tinamidae
Tinamus solitarius
C
f f fi LR/nt(e) Occasional 00
Crypturellus obsoletus f f fi 23(22) 78 0
Crypturellus parvirostris e Occasional 0 0
Crypturellus tataupa fffi ++ 432
Cathartide
Coragyps atratus
B,C
e++18
Cathartes aura
B,C
e + Occasional 4 1
Sarcoramphus papa
B,C
fcc + 40
Accipitridae
Leptodon cayanensis
B,C
fcc ++ 11
Elanoides forficatus
B
a + * 23 138
Harpagus diodon
B,C
fci ++ 22
Ictinia plumbea
B
a+*3988
Geranospiza caerulescens
B,C
f c c Occasional 10
Buteo magnirostris e (1) 1 20
Falconidae
Micrastur ruficollis f m c (4) 40
Micrastur semitorquatus f c c (3) (1) 2 3
Falco sparverius
B,C
e 2(2)* 0 20
Cracidae
Penelope superciliaris
B,C
fmfi + 10
Odontophoridae
Odontophorus capueira f f fi (e) Occasional 00
3278
Rallidae
Aramides saracura f f i (e) Occasional 00
Columbidae
Columba speciosa f c fg (1) 11 0
Columba picazuro f c fg + + 22 153
Columba cayennensis f c fg + + 24 38
Zenaida auriculata e Occasional 0 2
Claravis pretiosa
C
fffg ++ 33
Leptotila verreauxi f f fg + + 7 21
Leptotila rufaxilla f f fg + 30
Geotrygon montana
B,C
f f fg + 30
Psittacidae
Aratinga leucophthalmus f c fg + * 104 76
Pyrrhura frontalis f c fg + * 212 411
Forpus xanthopterygius e +016
Brotogeris chiriri e Occasional + 2 69
Pionopsitta pileata f c fg (e) + + 12 10
Pionus maximiliani f c fg + * 66 275
Amazona vinacea f c fg EN(e) + + 13 175
Cuculidae
Coccyzus euleri f c i Occasional Occasional 0 0
Coccyzus melacoryphus
B
e Occasional 0 1
Piaya cayana fci ++ 5422
Crotophaga major f m i Occasional 40
Crotophaga ani e (1 group) 0 23
Guira guira e (2 grps) 0 26
Dromococcyx phasianellus
C
fui + 10
Dromococcyx pavoninus
C
fui + 20
Tytonidae
Tyto alba e++01
Strigidae
Otus choliba f m i 5(2) 10(5)* 16 43
Otus atricapillus
B
f m i 12(4) + 17 1
3279
Appendix 1. Continued.
Species
A
Habitat
D
Strata
E
Diet
F
IUCN Status
G
(endemic spp.)
H
Presence and density
I
N territories/100 ha
(N territories total)
Total N birds
within 100 m of
observer
Forest Plant’n Forest Plant’n
Strix hylophila f c c Occasional 00
Strix virgata f c c (2) 10
Strix huhula f c c (1) 20
Glaucidium brasilianum f m i 8(10) 12(7) 43 36
Nyctibiidae
Nyctibius aethereus
C
f c i (1) 00
Nyctibius griseus fci ++ 86
Caprimulgidae
Lurocalis semitorquatus a++4571
Nyctidromus albicollis e +04
Caprimulgus sericocaudatus f c i Occasional 00
Apodidae
Cypseloides fumigatus a Occasional 0 1
Streptoprocne zonaris a Occasional 0 1
Chaetura cinereiventris a + + 29 140
Chaetura meridionalis a +033
Trochilidae
Phaethornis eurynome
B
fmn ++ 6630
Anthracothorax nigricollis
B
e Occasional 0 1
Stephanoxis lalandi
B
f m n (e) + + 26 6
Chlorostilbon aureoventris
B
e +02
Thalurania furcata
B
fmn +22
Thalurania glaucopis
B
f m n (e) + + 14 4
Hylocharis chrysura
B
e++319
Agyrtria versicolor
B
fmn ++ 21
Trogonidae
Trogon rufus f m fi (6) + 48 3
Trogon surrucura f c fi + * 114 124
3280
Momotidae
Baryphthengus ruficapillus f m i (e) + Occasional 52 6
Momotus momota f m i (1) 60
Bucconidae
Notharchus swainsoni
B
f c i (e) + * 7 15
Nystalus chacuru e Occasional 0 4
Nonnula rubecula
B
f m i + Occasional 15 1
Ramphastidae
Pteroglossus castanotis f c fg (3 grps) (2 grps)* 78 96
Selenidera maculirostris f c fg (e) (2) (1) 35 14
Ramphastos dicolorus f c fg (4 grps) (2 grps)* 19 27
Picidae
Picumnus temminckii
B
f u i (e) + + 24 5
Melanerpes candidus e Occasional 0 0
Melanerpes flavifrons f c fi + + 23 176
Veniliornis spilogaster f m i (e) + + 18 67
Piculus aurulentus f c i LR/nt (e) + + 8 2
Colaptes melanochloros fti ++ 325
Celeus flavescens
C
fti +* 511
Dryocopus galeatus
C
f t i VU (e) + + 1 1
Dryocopus lineatus
C
fti +* 212
Campephilus robustus
C
f t i (e) + + 5 18
Furnariidae
Synallaxis ruficapilla
B
f u i (e) 24(7) 117 0
Synallaxis cinerascens f u i 11(7) 69 0
Cranioleuca obsoleta f c i (e) Occasional Occasional 1 2
Syndactila rufosuperciliata f u i Occasional Occasional 4 4
Philydor lichtensteini f m i (e) + + 228 75
Philydor rufus fci ++ 197
Philydor atricapillus f m i (e) + 30
Automolus leucophthalmus f u i (e) 2(2) 32 0
Sclerurus scansor f f i (e) (1) 50
Heliobletus contaminatus f c i (e) + Occasional 3 1
3281
Appendix 1. Continued.
Species
A
Habitat
D
Strata
E
Diet
F
IUCN Status
G
(endemic spp.)
H
Presence and density
I
N territories/100 ha
(N territories total)
Total N birds
within 100 m of
observer
Forest Plant’n Forest Plant’n
Xenops minutus fmi + 14 2
Xenops rutilans fci ++ 2012
Dendrocolaptidae
Dendrocincla turdina f t i (e) + 90
Sittasomus griseicapillus f t i + + 237 175
Xiphocolaptes albicollis fti +* 5736
Dendrocolaptes platyrostris fti +* 9079
Lepidocolaptes fuscus fti ++ 507
Lepidocolaptes falcinellus fti +01
Thamnophilidae
Mackenziaena severa f u i (e) + 37 0
Thamnophilus caerulescens f u i 17(9) Occasional 159 2
Dysithamnus mentalis f u i 19(12) 196 0
Herpsilochmus rufimarginatus f c i + + 221 6
Terenura maculata f c i (e) + + 71 4
Pyriglena leucoptera f u i (e) + 70 0
Formicariidae
Chamaeza campanisona ffi + 41 0
Grallaria varia ffi + 30
Conopophagidae
Conopophaga lineata f u i (e) 42(15) 190 0
Tyrannidae
Mionectes rufiventris f u fi (e) 1(1) 50
Leptopogon amaurocephalus f u i 5(3) 33 0
Hemitriccus diops
B
f u i (e) 16(5) 85 0
Corythopis delalandi f f i 4(2) 30 0
Phyllomyias virescens
B
f c i (e) + + 3 6
3282
Camptostoma obsoletum f c fi + 3(1)* 4 36
Capsiempis flaveola
B
f u i 15(3) 34 0
Myiopagis caniceps f c i + + 125 94
Myiopagis viridicata fci ++ 82
Elaenia flavogaster e Occasional 0 1
Elaenia albiceps
B,C
e Occasional 0 1
Phylloscartes eximius f m i LR/nt(e) + 20
Phylloscartes ventralis
B,C
fci + 20
Phylloscartes sylviolus
B,C
f c i LR/nt(e) + (1)* 3 25
Myiornis auricularis
B
f u i (e) 67(16) 7(2) 236 10
Tolmomyias sulphurescens fmi ++ 94
Platyrinchus mystaceus f u i 21(7) 141 0
Cnemotriccus fuscatus
B,C
fui + 60
Lathrotriccus euleri f u i 11(9) Occasional 109 2
Contopus cinereus f c i 1(1) 8(4)* 10 80
Colonia colonus e + 4(2)* 2 22
Sirystes sibilator fci ++ 4791
Myiarchus swainsoni e++229
Myiarchus ferox e Occasional 0 1
Ramphotrigon megacephala f m i (1) 60
Tyrannus melancholicus e Occasional 0 3
Empidonomus varius
C
e +08
Megarynchus pitangua e + 18(12)* 12 168
Conopias trivirgata fci ++ 2545
Myiodynastes maculatus e 2(1) 18(8)* 12 139
Myiozetetes similis e 2(1) 0 7
Legatus leucophaius
C
e *05
Pitangus sulphuratus e 1(1) 29(14)* 16 342
Schiffornis assemblage
Schiffornis virescens f m fi (e) (4) 48 0
Piprites chloris fmfi ++ 544
Pachyramphus viridis e++1352
Pachyramphus castaneus e+*374
3283
Appendix 1. Continued.
Species
A
Habitat
D
Strata
E
Diet
F
IUCN Status
G
(endemic spp.)
H
Presence and density
I
N territories/100 ha
(N territories total)
Total N birds
within 100 m of
observer
Forest Plant’n Forest Plant’n
Pachyramphus polychopterus e Occasional 0 1
Pachyramphus validus e +011
Tityra cayana f c fi + + 58 124
Tityra semifasciata fcfi +03
Tityra inquisitor f c fi + * 48 136
Cotingidae
Pyroderus scutatus
C
f c fi (e) Occasional 0 0
Oxyruncidae
Oxyruncus cristatus fcfi ++ 3714
Pipridae
Pipra fasciicauda
B,C
f u fi (4) 36 3
Chiroxiphia caudata f m fi (e) (2) 14 0
Vireonidae
Cyclarhis gujanensis f c i (3) + 39 3
Vireo olivaceus fci + 30
Corvidae
Cyanocorax cyanomelas e Occasional 0 0
Cyanocorax chrysops f c fi + + 170 293
Hirundinidae
Progne chalybea a 55
Progne tapera a 33
Petrochelidon pyrrhonota a 04
Troglodytidae
Troglodytes aedon e 48(20) 0 412
Polioptilidae
Polioptila lactea
B
f c i LR/nt(e) + + 8 16
3284
Turdidae
Turdus subalaris
C
f c fi (e) + 10
Turdus rufiventris e++825
Turdus leucomelas e + * 50 442
Turdus amaurochalinus e++87
Turdus albicollis fmfi ++ 41
Parulidae
Parula pitiayumi fci ++ 9170
Basileuterus culicivorus f u i 69(33) 9(3) 501 66
Basileuterus leucoblepharus f u i 21(12) 217 0
Thraupidae
Conirostrum speciosum f c i + + 185 194
Cissopis leveriana fcfi ++ 4015
Hemithraupis guira f c i + + 333 238
Tachyphonus coronatus
C
f u fi Occasional 30
Habia rubica f u fi 4(3) 45 1
Trichothraupis melanops f u fi 15(8) 133 1
Thraupis sayaca e Occasional + 2 70
Pipraeidea melanonota
B,C
fcfi ++ 21
Euphonia chlorotica fcfi ++ 1681
Euphonia violacea fcfi ++ 3456
Euphonia pectoralis f c fi (e) + + 52 41
Chlorophonia cyanea
B,C
fcfi ++ 1426
Tangara seledon
B,C
f c fi (e) + + 10 60
Dacnis cayana
B,C
fcfi ++ 2546
Tersina viridis e + + 23 139
Emberizidae
Coryphospingus cucullatus
C
e 10
Sporophila caerulescens e 01
Arremon flavirostris fffg 20
Zonotrichia capensis e 03
3285
Appendix 1. Continued.
Species
A
Habitat
D
Strata
E
Diet
F
IUCN Status
G
(endemic spp.)
H
Presence and density
I
N territories/100 ha
(N territories total)
Total N birds
within 100 m of
observer
Forest Plant’n Forest Plant’n
Icteridae
Cacicus haemorrhous f c fi + * 56 127
Icterus cayanensis e++1687
Gnorimopsar chopi e Occasional + 2 27
Molothrus bonariensis e Occasional 0 1
Molothrus oryzivorus e Occasional 0 1
Molothrus rufoaxillaris e +09
A
Taxonomy follows Mazar Barnett and Pearman (2001).
B
Species whose vocalization cannot be heard consistently 100 m away.
C
Species that vocalize infrequently during the time of year covered by our study seasons.
D
Habitat groups are: f = forest; e = edge, a = aerial.
E
Strata groups (for forest birds only) are: c = canopy, m = midstory, u = understory, f = forest floor, t = tree trunk.
F
Diet groups (for forest birds only) are: fg = fruit-or-grain-eater, fi = fruit-and-insect-eater, i = insectivore, n = nectarivore, c = carnivore or carrion-eater.
G
IUCN conservation status follows IUCN (2002). EN = endangered, VU = Vulnerable, LR/nt = lower risk / near-threatened.
H
Species endemic to the Atlantic forest are marked (e). We follow Guyra Paraguay (2004).
I
Presence (in the main plots) is indicated by +, the estimated density, the total number of territories, or* (see below). Species considered absent are marked – if
they were never recorded in the habitat, or ‘occasional’ if they were occasional visitors to the habitat. We report the total number of territories and estimated
density only for species for which these numbers could be determined in both habitats.
* Species confirmed breeding in the habitat (active nests, adults repeatedly carrying nest material or food to the same tree, or newly fledged young within a
known territory)
3286
References
Aleixo A. 1999. Effects of selective logging on a bird community in the Brazilian Atlantic forest.
Condor 101: 537–548.
Alves M.C. 1990. The Role of Cacao Plantations in the Conservation of the Atlantic Forest of
Southern Bahia, Brazil. M.A. thesis, University of Florida, Gainesville.
Bibby C.J., Burgess N.D., Hill D.A. and Mustoe S. 2000. Bird Census Techniques (Second Edi-
tion). Academic Press, London and San Diego.
BirdLife International 2000. Threatened Birds of the World. Lynx Edicions and BirdLife, Barce-
lona and Cambridge.
Brower J.E. and Zar J.H. 1977. Field and Laboratory Methods for General Ecology. W.M.C.
Brown Publishers, Dubuque, Iowa.
Calvo L. and Blake J. 1998. Bird diversity and abundance on two different shade coffee plantations
in Guatemala. Bird Conservation International 8: 297–308.
Canaday C. 1997. Loss of insectivorous birds along a gradient of human impact in Amazonia. Biol.
Conserv. 77: 63–77.
Christiansen M.B. and Pitter E. 1996. Species loss in a forest bird community near Lagoa Santa in
southeastern Brazil. Biol. Conserv. 80: 23–32.
Cockle K. 2003. The Bird Community of Shade-grown Yerba Mate and adjacent Atlantic Forest in
Canindeyu´ , Paraguay. M.Sc. thesis, Dalhousie University, Halifax.
Eibl B., Fernandez R.A., Kozarik J.C., Lupi A., Montagnini F. and Nozzi D. 2000. Agroforestry
systems with Ilex paraguariensis (American holly or yerba mate) and native timber trees on small
farms in Misiones, Argentina. Agroforest. Syst. 48: 1–8.
Greenberg R., Bichier P., Cruz Angon A. and Reitsma R. 1997a. Bird populations in shade and sun
coffee plantations in central Guatemala. Conserv. Biol. 11: 448–459.
Greenberg R., Bichier P. and Sterling J. 1997b. Bird populations in rustic and planted shade coffee
plantations of eastern Chiapas, Me
´
xico. Biotropica 29: 501–514.
Guyra Paraguay 2004. Lista Comentada de las Aves de Paraguay, Annotated checklist of the Birds
of Paraguay. Asuncio
´
n.
International Bird Census Committee 1969. Recommendations for an international standard for a
mapping method in bird census work. Bird Study 16: 248–255.
IUCN 2002. 2002 lUCN Red List of Threatened Species. www.redlist.org. Downloaded on 23
November 2002.
James F.C. and Rathbun S. 1981. Rarefaction, relative abundance, and diversity of avian com-
munities. Auk 98: 785–800.
Marsden S.J., Whiffin M. and Galetti M. 2001. Bird diversity and abundance in forest fragments and
Eucalyptus plantations around an Atlantic forest reserve, Brazil. Biodiv. Conserv. 10: 737–751.
Mazar Barnett J. and Pearman M. 2001. Lista Comentada de las Aves Argentinas Annotated
Checklist of the Birds of Argentina. Lynx Edicions, Barcelona.
Moguel P. and Toledo V.M. 1999. Biodiversity conservation in traditional coffee systems of
Mexico. Conserv. Biol. 13: 11–21.
Myers N., Mittermeier R.A., Mittermeier C.G., da Fonseca G.A.B. and Kent J. 2000. Biodiversity
hotspots for conservation priorities. Nature 403: 853–858.
Ribon R., Simon J.E.J and de Mattos G.T. 2003. Bird extinction in Atlantic forest fragments of the
Vic¸ osa region, Southeastern Brazil. Conserv. Biol. 17: 1827–1839.
Rice R.A. and Greenberg R. 2000. Cacao cultivation and the conservation of biological diversity.
Ambio 29: 167–173.
Sick H. 1993. Birds in Brazil: A Natural History. Princeton University Press, Princeton.
Stattersfield A.J., Crosby M.J., Long A.J. and Wege D.C. 1998. Endemic Bird Areas of the World.
BirdLife International, Cambridge.
Stotz D.F., Fitzpatrick J.W., Parker T.A. III. and Moskovits D.K. 1996. Neotropical Birds:
Ecology and Conservation. University of Chicago Press, Chicago and London.
3287
Terborgh J., Robinson S.K., Parker T.A. III, Munn C.A. and Pierpont N. 1990. Structure and
organization of an Amazonian forest bird community. Ecol. Monogr. 60: 213–238.
Terborgh J. and Weske J.S. 1969. Colonization of secondary habitats by Peruvian birds. Ecology
50: 765–782.
Thiollay J. 1994. Structure, density and rarity in an Amazonian rainforest bird community. J. Trop.
Ecol. 10: 449–481.
Thiollay J. 1995. The role of traditional agroforests in the conservation of rain forest bird diversity
in Sumatra. Conserv. Biol. 9: 335–353.
Willis E.O. 1979. The composition of avian communities in remanescent woodlots in southern
Brazil. Pape
´
is Avulsos de Zoologia 33: 1–25.
Zar J.H. 1999. Biostatistical Analysis (Fourth Edition). Prentice Hall, Upper Saddle River, New
Jersey.
3288
