The Condor 111(1):2135

The Cooper Ornithological Society 2009

Avian Assemblages in Altered and Natural Grasslands

in the Northern Campos of Uruguay
A drin B. A zpiroz1,2 and John G. Blake
Department of Biology, Research Building 223,University of Missouri-St. Louis, One University Boulevard,
St. Louis, MO 63121-4499

Abstract. Alteration of habitat has been highlighted as the most important factor causing declines in popula-
tions of grassland birds. In this study we characterized bird assemblages in cultivated and natural grasslands under
four different agricultural schemes in Uruguay. We surveyed birds along transects to determine species richness,
composition, and population density along this agricultural gradient. Although species richness was lower in the
most natural grassland type (24 species), community composition and abundance patterns highlighted the impor-
tance of natural grasslands for the conservation of habitat specialists. An analysis of similarity showed that bird
communities in the four grassland types were distinct (ANOSIM global R = 0.68). Several species using grassland
facultatively were characteristic of cultivated habitats, whereas the opposite was true for grassland obligates and
natural grasslands. An indicator-species analysis further supported the association of grassland-facultative
and grassland-obligate birds with cultivated and natural grassland, respectively. Population-density patterns var-
ied by species; some were more abundant in cultivated habitats, others in natural grasslands, but threatened species
attained relatively high densities in natural grasslands only. Some threatened species, such as the Ochre-breasted
Pipit (Anthus nattereri) and Pampas Meadowlark (Sturnella defilippii), were largely restricted to natural grass-
lands, so their long-term survival will depend on the conservation of suitable patches of these habitat types.

Key words: bird abundance, diversity patterns, grasslands, Northern Campos, Pampas, Uruguay

Ensambles de Aves en Pastizales Modificados y Naturales en los Campos

del Norte de Uruguay
Resumen. La alteracin de hbitat ha sido resaltada como el factor ms importante responsable por las re-
ducciones poblacionales de especies de aves de pastizal. En este estudio caracterizamos los ensambles de aves
presentes en pastizales cultivados y naturales bajo cuatro tipos diferentes de actividades agrcolas. Se utiliz la
tcnica de muestreos con distancia para determinar la riqueza de especies de aves, la composicin y la densidad
poblacional a lo largo de este gradiente agrcola. Aunque la riqueza de especies fue menor en el tipo de pastizal en
estado ms natural (24 especies), los patrones de composicin de la comunidad y abundancia resaltaron la impor-
tancia de los pastizales naturales para la conservacin de los especialistas de hbitat. Un anlisis de similaridad
mostr que las comunidades de aves en los cuatro tipos de pastizales eran distintas (ANOSIM R global = 0.68),
y varias especies facultativas de pastizal fueron caractersticas de los hbitats cultivados, mientras que lo con-
trario ocurri con las aves obligatorias de pastizal y los pastizales naturales. La asociacin entre aves facultativas
y obligatorias de pastizal y pastizales cultivados y naturales, respectivamente, fue tambin corroborada por los
resultados de un anlisis de especies indicadoras. Los patrones de densidad poblacional variaron entre las espe-
cies de aves, siendo algunas ms abundantes en los hbitats cultivados mientras que otras lo fueron en pastizales
naturales; pero las especies amenazadas slo fueron registradas en densidades relativamente altas en pastizales
naturales. Algunas especies amenazadas como Anthus nattereri y Sturnella defilippii estuvieron restringidas prin-
cipalmente a los pastizales naturales, por lo que su supervivencia en el largo plazo depender de la conservacin
de parches apropiados de este tipo de hbitat.

communities, as well as substantial declines of some species

INTRODUCTION
The ecology and conservation of grassland birds have re-
ceived substantial attention during the last decade. Numer-
ous studies, most conducted in Europe and North America,
have documented changes in the composition of grassland bird
highly dependent on grassland habitats (Zimmerman 1988,
Knopf 1994, Tucker and Heath 1994, Patterson and Best 1996,
Vickery et al. 1999, Murphy 2003). Declines of grassland birds
worldwide have been recognized as a prominent wildlife con-
servation crisis of the 21st century (Brennan and Kuvlesky

Manuscript received 15 July 2008; accepted 25 December 2008.

1
Current address: Buxareo 1311, 11300 Montevideo, Uruguay
2
E-mail: abavg5@umsl.edu

The Condor, Vol. 111, Number 1, pages 2135. ISSN 0010-5422, electronic ISSN 1938-5422. 2009 by The Cooper Ornithological Society. All rights reserved. Please direct all
requests for permission to photocopy or reproduce article content through the University of California Presss Rights and Permissions website, http://www.ucpressjournals.com/
reprintInfo.asp. DOI: 10.1525/cond.2009.080111

21

02_MS080111.indd 21 2/13/09 4:49:07 PM

 22 Adrin B. Azpiroz and John G. Blake
2005). Habitat modification, particularly that related to agri-
cultural expansion and intensification, has been identified as
one of the main causes of such declines (Askins et al. 2007).
In the Neotropical Region, the most important expanse of
grasslands is located in south-eastern South America and is
known as the Ro de la Plata grasslands (Soriano 1992). This
region (ca. 700 000 km 2), which is included within the Pam-
pas biome (sensu Stotz et al. 1996), is dominated by temperate
subhumid grasslands that extend from 28 to 38 S and cover
the plains of east-central Argentina, Uruguay, and southern
Brazil (Soriano 1992). Although the original vegetation was
a tall-grass steppe, intermingled with prairies, marshes, and
other edaphic communities, it has been heavily modified by
long-term human use (Bucher and Nores 1988). In some mesic
regions of the Pampas, more than 50% of the land is devoted
to agriculture (Vickery et al. 1999).
Presettlement information on grassland bird assemblages
in the Pampas is limited, but evidence exists that the loss of
pristine habitats in the region, as a consequence of expansion
of agriculture and livestock grazing, has affected bird popu-
lations, producing marked decreases in abundances and local
extinctions of some species (e.g., Bucher and Nores 1988, Sori-
ano 1992, Collar et al. 1992, Tubaro and Gabelli 1999). Recent
studies in the Flooding Pampa of eastern Argentina (defined
and mapped by Soriano 1992) have shown that changes in
the structure of grasslands due to fire or grazing result in the
replacement of grassland specialists, highly dependent on
tall grasslands, by others that prefer short grasslands (Com-
paratore et al. 1996, Isacch and Martnez 2001). These pat-
terns are similar to those reported in areas of North America
where declining grassland endemics have been replaced by
more widespread counterparts (Knopf 1994). In spite of these
recent efforts, patterns of avian diversity currently associated
with various land-use practices in the Pampas have not been
fully characterized, and more studies are needed to further
assess the role of landscape alteration in shaping bird assem-
blages in South American grasslands. This is particularly im-
portant in Uruguay, located within the campos subregion
of the Pampas, where agricultural lands are the most exten-
sive wildlife habitat and, therefore, represent key habitats for
biodiversity; the extent of croplands, pastures, and rangelands
greatly exceeds the areas set aside as wildlife reserves (OEA
1992, World Resource Institute 2007).
Here, we present results of the first study to examine the
effects of land-use practices on grassland bird communities
in the Northern Campos of Uruguay, a region defined and
mapped by Soriano (1992). A major goal of this study was to
determine the relative value for grassland bird conservation of
areas under different management schemes. First, we charac-
terized the distribution and seasonality of birds associated with
croplands, pasturelands, and two types of native grasslands.
Second, we compared species richness and abundance of bird
communities found in each of these four habitat types. Because
habitat loss related to agriculture has been indicated as a ma-
02_MS080111.indd 22
jor threat to many grassland specialists (BirdLife International
2000, Askins et al. 2007), we expected natural grasslands
to support more species and higher densities of grassland-
obligate birds (sensu Vickery et al. 1999) than do cultivated
lands, as well as more threatened taxa. Factors, such as habitat
loss, that reduce niche availability should represent a particu-
lar threat to species that are ecologically specialized (Owens
and Bennett 2000). If the degree of habitat specialization is a
valid surrogate for ecological specialization, the proportion of
grassland-obligate species (sensu Vickery et al. 1999) should
be inversely related to the degree of habitat loss. Thus, obligate
grassland species should be less well represented in terms of
species numbers and densities in habitats where alteration has
been more drastic (i.e., croplands and planted pastures).
METHODS
Study area
This study was conducted in the Northern Campos subregion
of the Ro de la Plata grasslands (Soriano 1992) within the
Pampas biome (sensu Stotz et al. 1996). This subregion ex-
tends through most of Rio Grande do Sul, Brazil, southeast-
ern Misiones and eastern Corrientes provinces in Argentina,
and northern Uruguay (Bilenca and Miarro 2004). Within
the Northern Campos, the study area was located in south-
ern Salto and northwestern Paysand departments (31 19 to
31 44S and 56 42 to 57 56W), and it was composed of
rolling topography in which low mesas and rocky outcrops
are interspersed. The area has mesothermic and humid fea-
tures (Soriano 1992), with a mean annual temperature of 19C
and mean annual rainfall of 1300 mm (Lezama et al. 2006).
Cattle ranching is the major activity, mainly on natural pas-
tures, but also on planted pastures, the latter found especially in
the west. Cropland, especially fields of wheat and barley, is also
widespread in the west. The eastern part of the study region is
characterized by large expanses of natural grasslands that have
never been plowed and are used for open-range livestock graz-
ing. The study was conducted on ranches and farms around
Chapicuy, in the department of Paysand and San Antonio
and Cerros de Vera in the department of Salto. Large patches
of each habitat type (500 to 1200 ha) were selected within
each of these areas.
Habitat descriptions and species
definitions and categories
The study focused on birds that inhabit grassland (sensu Vick-
ery et al. 1999) and grassland-like habitats. Habitats were clas-
sified into four categories on the basis of different cultivation
and grazing regimes.
1. Crop habitat (Crop) was represented by barley fields in-
tegrated into a system of livestock grazing in which the
crop is rotated with planted pastures that are grazed after
the crop is harvested. In fact, this habitat type consisted
2/13/09 4:49:08 PM
 Grassland Bird Diversity in Northern Uruguay 23
of three different management phases: (1) barley fields
and stubble, (2) sunflower residual crop fields and stub-
ble, and (3) planted pastures. Barley and pastures were
planted in austral winter 2004. By the time bird censuses
were initiated (September 2004), the barley was well
grown (height >1 m) and planted pastures were starting
to grow beneath the crop. Barley was harvested in late
October 2004. From December 2004 to March 2005, a
residual crop of sunflower grew on these fields. This sec-
ond crop developed from seeds that fell to the ground
during the sunflower harvest in early 2004, and it was
harvested in late March 2005; afterward, the fields were
used for cattle grazing on the planted pastures until the
end of the study (November 2005). The management
of this cropland is typical of this part of Uruguay. For
some analyses the crop phase (i.e., crop A; September
2004March 2005) and the pasture phase (i.e., crop B; Grassland bird surveys
MayNovember 2005) of this habitat type were treated
separately.
2. Planted-pasture habitat (Pasture) consisted of grasslands
seeded with exotic grasses. These pastures were part of
dairy farms and were used for cattle grazing.
3. One type of native habitat (Native 1) was represented by
native grasslands grazed by livestock. On these fields,
cattle and sheep forage on native grasses and forbs. As a
result of grazing by sheep, vegetation diversity is lower
than that in the following habitat type (Sturm 2001).
4. A second type of native habitat (Native 2) consisted of
fields of native grasslands grazed by cattle and Pampas
Deer (Ozotoceros bezoarticus). The latter are thought
to be indicative of distinct grassland habitats (Sturm
2001). Because of the lack of sheep grazing (only about
2030 sheep were maintained on these fields to supply
the ranch), this habitat supports, in general, a more com-
plex and higher vegetation structure than does Native 1
(Sturm 2001).
Cultivated habitats (Crop and Pasture) and natural grass-
lands (Native 1 and Native 2) differ with respect to two impor-
tant management activities: use of agrochemicals and plowing.
These activities were absent from natural grasslands. Addi-
tionally, the former habitat types were located within a matrix
of croplands and planted pastures. In contrast, Native 1 and
Native 2 sites were included in a matrix of activities similar to
those found in Native 1. However, because the study was con-
ducted on large patches of each habitat type, any confounding
effects related to matrix composition were minimized.
Here, with minor modifications, we have followed the eco-
logical definition of grassland birds by Vickery et al. (1999).
These authors identified two groups of grassland birds, obli-
gate and facultative species, on the basis of their dependence on
grasslands (higher and lower, respectively). A series of species
not included in their scheme we classify as grassland-facultative
birds. These include the American Golden-Plover (Pluvialis
02_MS080111.indd 23
dominica) and six species of Hirundinidae. Additionally, three
species categorized by Vickery et al. as facultative we clas-
sify as obligate: the Greater Rhea (Rhea americana), Nacunda
Nighthawk (Podager nacunda), and Chocolate-vented Tyrant
(Neoxolmis rufiventris). We believe these adjustments bet-
ter reflect these species reliance on grassland habitats in Uru-
guay (Azpiroz 2001). We assigned species to four categories
reflecting migration patterns (following Gore and Gepp 1978
and Azpiroz 2001): residents, summer residents, summer visi-
tors, and winter visitors. Birds were also classified according
to five feeding guilds that reflect the main component of their
diets: carnivores, granivores, herbivores, insectivores, and om-
nivores. Information on species migratory status and diet in
the Pampas were taken from the literature (e.g., Gore and Gepp
1978, Sick 1985, Canevari et al. 1991).
We conducted bird counts on eight 500-m variable-width
transects on each of the four habitat types every two months
from September 2004 to November 2005. Thus each transect
was sampled eight times during the whole sampling period.
Transects were located at least 400 m apart, far from fencerows,
and avoided the intersection of other nongrassland habitats (i.e.,
gallery forest). We selected transects randomly after consid-
ering these constraints. In total, approximately 120 ha of each
habitat type were sampled. All bird surveys were conducted by
Azpiroz. Transects were walked at a pace of approximately 1
km hr1 from 0 to 3.5 hours after sunrise, and all birds seen or
heard on each side, with the exception of individuals passing by
and making no use of the surveyed area, were recorded. Swal-
lows feeding on the wing within the surveyed plots were, how-
ever, counted. Coverage of transects and the direction walked
on each were rotated systematically. Distances and angles from
transects to individuals were estimated with a rangefinder and
a compass, to allow calculation of perpendicular distances (see
Buckland et al. 2001). These data were used to estimate vari-
ables (species richness, composition, and density) to describe
the avian community structure on each study site.
Statistical analysis
Species richness. We calculated species richness in each hab-
itat type (and for both Crop phases separately) as the total
number of species encountered in all transects in each habitat,
all sampling periods pooled. We also calculated an estimate
of species richness on the basis of a jackknife estimator using
observed abundance-distribution data by species and the pro-
gram SPECRICH (Hines 1996). The method, which considers
detection probabilities differing by species, was described by
Burnham and Overton (1979). To test for seasonal differences
in species richness and number of individuals among habitats,
we used repeated-measures analyses of variance (rmANOVA).
Data were tested for normality and, when necessary, standard
transformations were applied. Analyses of variance were run
in SPSS, version 15.0 (SPSS 2006). We compared the rates
2/13/09 4:49:08 PM
 24 Adrin B. Azpiroz and John G. Blake
of species accumulation across habitats through rarefaction
analyses based on Monte Carlo simulations run 1000 times in
EcoSim 7 (Gotelli and Entsminger 2006). The analyses were
based on a sample of 503 individuals, the lowest number of
birds recorded for any habitat (Crop A).
Community composition. We used species presence/ab-
sence over all sampling periods in each habitat to calculate
BrayCurtis similarity coefficients among habitat types. Data
from the similarity matrix were also used to test for differences
in species composition among habitats through an analysis of
similarity (ANOSIM; Clarke and Warwick 2001). ANOSIM
determines whether samples (i.e., transects) within each habi-
tat type are more similar to each other than to samples taken at
random from the whole sample pool (i.e., 32 transects). Thus
ANOSIM compares the level of similarity among transects of
a given habitat to that among transects of all habitats and de-
termines if the former is greater than expected by chance. Re-
sults from ANOSIM were tested for significance with a Monte
Carlo randomization procedure. ANOSIM also estimates
pairwise comparisons, so used it to determine species turn-
over at a spatial scale (comparisons among habitats) and at a
temporal scale (comparisons among sampling periods within
each habitat). Afterward, we used nonmetric multidimen-
sional scaling (MDS) to compare species composition among
transects and habitats. This analysis graphically portrays sim-
ilarities and differences among samples based on species
presence/absence; thus, points that are closer are more simi-
lar in terms of species composition. ANOSIM and MDS were
done with PRIMER version 5.2.9 (Clarke and Gorley 2002)
and PC-ORD version 4 (McCune and Mefford 1999), respec-
tively. These analyses do not account for differences in de-
tectability, but because they were used to analyze data from
habitats with similar structural characteristics (i.e., grassland
and grassland-like habitats), we believe results are still infor-
mative despite this limitation. Across-habitat differences in
the proportions of species included in categories of habitat
specialization, migration, and feeding guild were tested with
G-tests. In order to identify characteristic species of each hab-
itat type, we used an indicator-species analysis (Dufrne and
Legendre 1997). This analysis, which we did in PC-ORD ver-
sion 4 (McCune and Mefford 1999), calculates an indicator
value for species based on its relative frequency and relative
abundance in all treatment categories (i.e., habitat types). In-
dicator values can range from 0 (no indication) to 100 (per-
fect indication). A perfect indication for a given habitat means
that the species was recorded in all samples (i.e., transects)
within that habitat and was not observed in any of the samples
of other habitats. Indicator values were tested for significance
with a Monte Carlo randomization procedure which com-
pares the observed indicator values to alternative values cal-
culated from the same data and randomly assigned to habitat
type. Only species with indicator values that were significant
(P 0.01) and >25% are reported.
02_MS080111.indd 24
Population densities. We used program DISTANCE ver-
sion 5.0 (Thomas et al. 2005) to estimate densities of grassland
birds from census data obtained on transects. We constructed
detection functions for all species with at least 60 observations.
For each species, data from all habitats were pooled except
when vegetation structure was thought to influence detection
probability. For example, census data from fields of barley and
sunflower were used separately from data from other habitats
where vegetation was significantly lower (unpubl. data). Addi-
tionally, in a few cases, species similar morphologically and be-
haviorally were grouped together in order to increase sample
sizes to allow for construction of detection functions (see Buck-
land et al. 2001:302). The probability of detecting each species
as a function of perpendicular distance from the transects was
determined by means of the following robust models presented
by Buckland et al. (2001): uniform key function with cosine and
simple polynomial expansion series, the half-normal key func-
tion with cosine and hermite polynomial expansion series, and
the hazard-rate key function with cosine and simple polynomial
expansion series. We evaluated each of these models consider-
ing the complete data set (i.e., all observations) or subsets with
5 and 10% truncation of detections at largest distances to re-
duce errors incurred by outliers, as recommended by Buckland
et al. (2001). Model suitability was evaluated through Akaikes
information criterion (AIC). Density estimates are presented
with standard errors and 95% confidence intervals; estimates
for values with non-overlapping intervals were considered sig-
nificantly different.
RESULTS
Species richness
Throughout the whole 14-month sampling period, 4968 in-
dividuals of 50 grassland bird species were recorded, all
transects combined (Appendix 1). Total observed species
richness varied from 24 on Native 2 to 34 on Native 1 (Table 1,
Fig. 1), and results corrected through rarefaction analysis
yielded similar patterns (Table 1). Species richness estimated
with SPECRICH ranged from 26 2.0 for Native 2 to 43
7.8 for Crop A, indicating that more species could still be re-
corded in all habitats, especially in Crop A (Table 1).
The mean number of species per habitat varied signifi-
cantly through the sampling period (Fig. 2A) and responded to
the effects of time (rmANOVA, F7,196 = 7.46, P < 0.001), habitat
(rmANOVA, F3,28 = 10.66, P < 0.001), and time habitat in-
teraction (rmANOVA, F21,196 = 2.99, P < 0.001). Mean number
of individuals per habitat (Fig. 2B) varied as an effect of time
(rmANOVA, F7,196 = 2.09, P = 0.045) and time habitat inter-
action (rmANOVA, F21,196 = 2.09, P = 0.005); in this case, the
effect of habitat was not significant (rmANOVA, F3,28 = 1.12,
P = 0.36). During May 2005, a few large flocks (i.e., 38 indi-
viduals, the largest flocks recorded during the whole sampling
period) of the Eared Dove (Zenaida auriculata) and Shiny
2/13/09 4:49:09 PM
 Grassland Bird Diversity in Northern Uruguay 25

Phase A and Native 1 were the most different in species com-

FIGURE 1. Species-accumulation curves for each of the four habi-
tats studied, based on visual and aural detections from September
2004 to November 2005. For Crop, two curves corresponding to each
of the two phases of this habitat type are presented (Crop A = crop
phase; Crop B = pasture phase; see Methods for more details).
Cowbird (Molothrus bonariensis) were observed in Crop
(Fig. 2B). When these flocks were excluded from the anal-
ysis, the effect of time (rmANOVA, F7,196 = 1.75, P = 0.099)
was not significant and the effect of time habitat interaction
(rmANOVA, F21,196 = 1.58, P = 0.058) was only marginally so.
Species composition
The mean number of habitat types used per species was 2.4
0.2. Eleven species were recorded in all four habitats, and 14
were seen in only a single habitat (Appendix 1). In terms of
species composition, the analysis of similarity indicated that
overall differences in species composition by habitat were sig-
nificant (global R = 0.68, P = 0.001), and such differences are
highlighted in the MDS graphical representation (Fig. 3). In
terms of spatial turnover, the two phases of Crop were the least
different in species composition (R = 0.41, P = 0.001), followed
by Native 1 and Native 2 (R = 0.42, P = 0.001), whereas Crop
position (R = 0.99, P = 0.001) (Table 2). In the assessment of
temporal turnover within each habitat, the results of ANOSIM
indicated that in all habitats differences in species composition
by sampling period were small (Fig. 3). Crop was the habitat
most different (global R = 0.43, P = 0.001) from the other three
habitats (Pasture global R = 0.30, P = 0.001; Native 1 global R =
0.15, P = 0.001; Native 2 global R = 0.34, P = 0.001).
The indicator-species analysis identified 11 species with
significant indicator values; four were associated with Crop,
three with Native 1, and four with Native 2 (Table 3). It iden-
tified no indicator species for Pasture. Except for the Great
Pampa-Finch (Embernagra platensis), Crop indicator species
were all facultative grassland birds. Conversely, all Native 2
indicator species were obligate grassland birds. Notably, all
three species of pipit showed indicator signals.
Species differed in terms of habitat specialization, mi-
gration status, and diet (Appendix 1). The total 50-species
pool included 23 obligate and 27 facultative grassland spe-
cies (sensu Vickery et al. 1999), respectively. Crop harbored
the highest proportion of facultative species and the lowest
of obligate species; the opposite pattern prevailed in Native
2. Bird assemblages were dominated by resident species,
which represented between 71% and 85% of all species in the
four habitat types. Among migrants, only summer residents
(present mainly from September to March) were well repre-
sented, accounting for 12 to 14% of all species found in each
habitat. In terms of feeding guilds, insectivores were the best
represented in all habitats, accounting for 42 to 66% of spe-
cies recorded. Proportion of granivores was highest in Crop
(30%) and lowest in Native 2 (8%). Despite these trends,
overall proportions of obligate/facultative species (G = 3.20,
df = 3, 0.50 > P > 0.25) and of species in different migration
(G = 6.76, df = 9, 0.75 > P > 0.50) and feeding-guild (G = 9.23,
df = 12, 0.75 > P > 0.50) categories did not differ significantly
by habitat.
Species of conservation concern were recorded in all hab-
itat types. The Greater Rhea was the only one found in all

TABLE 1. Observed numbers of individuals and species and expected number of species in each of four
habitat types in the Northern Campos of Uruguay (RSR = rarefied species richness; ESR = estimated species
richness). Values for Crop distinguish between the crop phase (Crop A) and the pasture phase (Crop B;
see Methods for more details).

Crop

Crop A Crop B Pasture Native 1 Native 2

Total number transects 8 8 8 8 8
Total survey periods 4 4 8 8 8
Individuals observed 503 1052 1150 1225 1038
Observed species richness 26 28 30 34 24
RSR (95% CI) 26.0 25.9 26.4 28.8 24.0
(26.0) (24.028.0) (23.029.0) (25.032.0) (20.024.0)
ESR ( SE) 43 7.8 30 2.0 33 2.5 38 2.8 26 2.0

02_MS080111.indd 25 2/13/09 4:49:11 PM

 26 Adrin B. Azpiroz and John G. Blake

FIGURE 2. Temporal variation in mean species richness (A) and mean number of individuals (B) for each habitat type from September
2004 to November 2005 in the Northern Campos of Uruguay.

02_MS080111.indd 26 2/13/09 4:49:19 PM

 Grassland Bird Diversity in Northern Uruguay 27

TABLE 3. Indicator values (as percentage of perfect indication) and

values of Monte Carlo test of significance of observed maximum indi-
cator values. Only values significant at least at P < 0.01 are shown.

Indicator values based on habitats

Species Crop Pasture Native 1 Native 2 P
Gallinago paraguaiae 0 6 9 61 0.002
Columba maculosa 63 0 0 0 0.003
Zenaida auriculata 87 13 0 0 0.001
Geositta cunicularia 0 0 63 0 0.001
Xolmis cinerea 68 0 1 0 0.001
Alopochelidon fucata 1 0 56 0 0.002
Anthus furcatus 0 19 47 30 0.003
Anthus hellmayri 2 0 19 55 0.006
Anthus nattereri 0 0 1 83 0.001
Sicalis luteola 3 2 10 75 0.007
FIGURE 3. Nonmetric multidimensional scaling based on spe- Embernagra platensis 54 2 0 0 0.003
cies presence/absence per transect for each sampling period. Sym-
bols represent individual transects during a given sampling period:
Crop transects during crop phase (open circles) and pasture each habitat, the Spotted Nothura and Grassland Yellow-Finch
phase (filled circles); Pasture (squares), Native 1 (triangles), and Na- (Sicalis luteola) were the only ones shared by all. Additionally,
tive 2 (diamonds) habitats. The closer the symbols the more similar
the Southern Lapwing (Vanellus chilensis), White-rumped
they are in terms of species composition.
Swallow (Tachycineta leucorrhoa), and Short-billed Pipit were
included in the top ranks of all habitats except Crop. Unlike
four habitats; the Ochre-breasted Pipit (Anthus nattereri) and most species, density estimates of the White-rumped Swallow
Pampas Meadowlark (Sturnella defilippii) were recorded in and Brown-chested Martin (Progne tapera) were similar for
both Native 1 and Native 2, whereas the Dark-throated Seed- all four habitats (Table 4). In terms of individuals, the five most
eater (Sporophila ruficollis) was found in Crop and Pas- abundant species accounted for 67.6% of all birds recorded in
ture. The Buff-breasted Sandpiper (Tryngites subruficollis), Crop, 78% of those in Pasture, 62.5% of those in Native 1, and
Bearded Tachuri (Polystictus pectoralis) and Black-and-white 73.0% of those in Native 2 (Appendix 1).
Monjita (Heteroxolmis dominicana) were recorded only in
Native 1, Pasture, and Crop, respectively. DISCUSSION

Population densities Species richness

Patterns of species densities differed by habitat type. Of species The highest number of species was recorded in Native 1, the
with enough observations for detection functions to be built for lowest in Native 2; the former has a lower vegetation struc-
density estimation (Table 4, Appendix 2), five had higher den- ture that reflects the effects of sheep grazing (Sturm 2001, un-
sities in Native 1, four species had higher densities in Crop, publ. data). In the West Pampa (region defined and mapped by
and three species had higher densities in Native 2 and Pasture, Soriano 1992), Isacch et al. (2003) found that ungrazed natu-
respectively (Table 4). Species with the highest densities in- ral grasslands supported fewer species than did those under
cluded the White-browed Blackbird (Sturnella superciliaris) grazing regimes. They hypothesized that the greater struc-
and Eared Dove in Crop, Spotted Nothura (Nothura maculosa) tural heterogeneity promoted by grazing was responsible for
and White-browed Blackbird in Pasture, and Short-billed Pipit differences in species richness (Isacch et al. 2003). In con-
(Anthus furcatus) and Spotted Nothura in both Native 1 and trast, in his study of five North American grassland habitats,
Native 2. Among the 10 species with the highest densities for Wiens (1974) found it especially intriguing that habitats with
higher structural development did not support more species
TABLE 2. Pairwise comparisons in species composition than those with less.
among habitats based on presence/absence. Values are pairwise Undoubtedly, the relatively high number of species we
R from ANOSIM based on BrayCurtis similarity. All the com- found in Crop is explained, at least in part, by this habitat
parisons were significant (P < 0.05). categorys including several management phases with differ-
ing vegetation structures. Just as spatial habitat heterogeneity
Crop A Crop B Pasture Native 1
in agricultural landscapes is known to facilitate the co-oc-
Crop B 0.41 currence of species with diverse habitat requirements (e.g.,
Pasture 0.56 0.51 Verhulst et al. 2004), temporal habitat heterogeneity in Crop
Native 1 0.99 0.80 0.67 probably has a similar effect by allowing species with differ-
Native 2 0.95 0.81 0.70 0.42
ing ecological needs to exploit these areas whenever suitable

02_MS080111.indd 27 2/13/09 4:49:22 PM

 28 Adrin B. Azpiroz and John G. Blake

TABLE 4. Density (individuals per 100 ha), SE, and 95% CI calculated with program
DISTANCE for species with at least 60 observations. The highest density value for each
species is shown in bold.

Habitat
Species Crop Pasture Native 1 Native 2
Rhea americana
Mean 5.5 0.3 10.7 7.0
SE 2.1 0.3 3.1 2.3
95% CI 2.611.6 <0.11.6 6.218.7 3.713.2
Nothura maculosa
Mean 20.7 126.7 98.3 93.1
SE 9.6 41.2 32.5 29.4
95% CI 8.649.5 67.8236.8 52.0185.7 50.6171.2
Vanellus chilensis
Mean 5.6 28.1 49.3 31.7
SE 2.1 6.9 5.9 4.2
95% CI 2.711.5 17.345.6 39.062.4 24.441.1
Pluvialis dominica
Mean 0.0 0.0 4.8 1.1
SE 1.9 0.8
95% CI 2.210.4 0.34.2
Bartramia longicauda
Mean 1.3 4.0 3.0 1.3
SE 1.0 1.7 1.2 0.8
95% CI 0.35.3 1.79.1 1.46.3 0.44.1
Zenaida auriculataa
Mean 134.4
SE 43.7
95% CI 70.8255.3 N/A N/A N/A
Zenaida auriculatab
Mean 25.4 66.1 0.0 0.0
SE 15.0 29.0
95% CI 8.575.6 29.0151.2
Tachycineta leucorrhoa
Mean 18.4 14.1 11.7 4.9
SE 5.3 3.8 3.4 1.9
95% CI 10.432.4 8.323.9 6.620.7 2.310.4
Progne tapera
Mean 7.7 5.1 2.6 3.1
SE 2.1 2.1 1.7 1.2
95% CI 4.513.2 2.311.5 0.88.6 1.46.7
Anthus furcatus
Mean 0.0 94.3 321.2 219.2
SE 32.7 78.5 56.9
95% CI 48.5183.3 200.0515.6 132.7362.0
Anthus hellmayri
Mean 1.5 0.0 5.8 11.8
SE 0.7 1.7 3.2
95% CI 0.73.5 3.210.4 6.920.1
Anthus nattereri
Mean 0.0 0.0 1.1 21.1
SE 0.8 4.2
95% CI 0.34.3 14.231.1
Ammodramus humeralis
Mean 20.2 27.3 9.7 2.6
SE 6.9 14.0 4.3 2.1
95% CI 10.539.0 10.471.1 4.222.3 0.710.5
(continued)

02_MS080111.indd 28 2/13/09 4:49:23 PM

 Grassland Bird Diversity in Northern Uruguay 29

TABLE 4. (Continued)

Habitat
Species Crop Pasture Native 1 Native 2
Sicalis luteola
Mean 3.2 3.2 5.0 47.9
SE 1.5 2.0 1.7 10.4
95% CI 1.37.7 1.09.9 2.69.6 31.273.5
Sturnella defilippii
Mean 0.0 0.0 13.6 1.5
SE 6.7 1.1
95% CI 5.434.3 0.45.8
Sturnella superciliarisa
Mean 242.4 N/A N/A N/A
SE 83.7
95% CI 124.8470.8
Sturnella superciliaris b
Mean 91.7 95.1 11.3
SE 27.0 21.4 4.1
95% CI 51.3164.1 61.2147.7 0.0 5.622.6
a
Detection function built exclusively with data from September 2004 to March 2005, corre-
sponding to the crop phase (barley and sunflower fields) of Crop habitat.
b
Detection function built with data from May to November 2005, corresponding to the crop
phase (barley and sunflower fields) of crop habitat, as well as with data from other habitats for
the total sampling period.

conditions become available (i.e., feeding and/or breeding
opportunities).
In general, more species were present during the austral
spring and summer than during fall and winter. Similar tem-
poral patterns have been reported for the West and Flood-
ing Pampas (Isacch and Martnez 2001, Isacch et al. 2003),
where seasonal variation in species numbers stems, at least
in part, from more migratory birds reaching the study area
during the summer than during the winter (Isacch and Mar-
tnez 2001). This same pattern also applied to our study area
in the Northern Campos, where the number of summer mi-
grants (11) was substantially higher than that of their winter
counterparts (2).
Species composition
A relatively low proportion (22%) of all species in the study
area was present in all habitat types; most species were re-
stricted to certain subsets of habitats (50%) or to single
habitats (28%). The absence of several tyrant-flycatcher spe-
cies from Native 2 is probably related to the lack of suitable
perches, a key feature of these species foraging strategies.
Species absent from Crop suggest that this habitat type pro-
vides limited opportunities for ground-nesting birdsnot
surprising since operations in agricultural fields, especially
crop harvesting, can have detrimental effects on nesting suc-
cess (see Mller et al. 2005). Alternatively, many species
absent from Native 1, such as the Bearded Tachuri, Black-
and-white Monjita, Long-tailed Reed-finch (Donacospiza
albifrons), Great Pampa-Finch, and Dark-throated Seed-
eater, place their nests in tall grass or shrubs, which also
suggests that nest-site availability plays an important role in
determining bird communities in this habitat type. This fac-
tor is thought to be key in shaping bird assemblages in agri-
cultural areas in general (Sderstrm et al. 2003).
Cultivated and natural grasslands sites differed not only
in terms of agricultural management (local characteristics)
but also in terms of the composition of the matrix in which
the different types of grasslands were embedded (regional
characteristics). Cultivated grasslands were located within a
similar matrix of crops and planted pastures, whereas natu-
ral grasslands were located within a matrix characterized
by conditions resembling Native 1 habitat. This pattern re-
flected a logistic constraint, since there is currently no region
in Uruguay that contains habitat patches with the characteris-
tics of those studied here within a single type of matrix. Be-
cause of this mismatch in matrix characteristics, differences
in the bird community between habitats in cultivated and
natural grasslands may be influenced by this variable. But a
substantial proportion of species being restricted to a single
habitat within each matrix type indicates that management
practices (i.e., local conditions) do play an important role in
determining species richness in the region. Information from
the southern Pampas supports this conclusion; the presence
of breeding Pampas Meadowlarks could be predicted by local
variables (e.g., local habitat-specific characteristics) but not
by landscape variables (Fernndez et al. 2003).

02_MS080111.indd 29 2/13/09 4:49:24 PM

 30 Adrin B. Azpiroz and John G. Blake
Population densities
In general, species with higher densities in cultivated habitat
types are common and widespread throughout the Pampas,
while all species of conservation concern for which density es-
timates are available occurred exclusively or in higher densi-
ties in natural grasslands (Greater Rhea, Ochre-breasted Pipit,
Pampas Meadowlark). Although detailed information on the
diet of grassland birds in the Pampas is limited, data from the
five most abundant species in each habitat type suggest that bird
assemblages in cultivated grasslands are dominated by species
that rely heavily on seeds, while those in natural grasslands are
dominated largely by species that regularly take insects.
The Short-billed Pipit was the most abundant species in
natural grasslands. Additionally, differences in pipit densities
between cultivated and natural grasslands were very marked.
Pipits are also important elements in natural grasslands of
other subregions of the Pampas (Comparatore et al. 1996,
Isacch et al. 2003) and also in some North American grass-
lands (Owens and Myers 1973).
In the case of the Greater Rhea, density estimates were
significantly higher for Native 1 than for Pasture. In contrast,
in Crdoba province, Argentina, Bellis et al. (2004) reported
the rhea to prefer planted pastures over natural grasslands.
This difference may reflect the effects of poaching, which af-
fects natural populations of the Greater Rhea severely (Bellis
et al. 2004 and references therein). We noted poaching in our
Pasture sites, whereas it was infrequent in the Argentine study
area (Bellis et al. 2004).
Density estimates of the Ochre-breasted Pipit for Native 2
were significantly higher than those for Native 1, suggest-
ing that the species population trends will be affected by the
availability of areas with characteristics similar to those found
in Native 2. In other parts of its range, the Ochre-breasted
Pipit has been reported to inhabit burnt areas with regener-
ating short grass and lightly grazed grasslands (Tyler 2004).
It is worth noting that a few individuals were recorded in
Pasture from early October to early November 2005 (i.e., be-
tween sampling periods). Attempts at breeding were not suc-
cessful (nests were depredated or abandoned; unpubl. data).
On the basis of available information, which suggests
high sensitivity to habitat change (Fernndez et al. 2003), a
tight association of the Pampas Meadowlark with Native 2
was expected, but this was not the case. It is intriguing that the
Pampas Meadowlark was largely confined to a specific area
of Native 1 and mostly absent from Native 2, even though, in
terms of grazing pressure and vegetation cover, the latter hab-
itat better resembles the species preferences in southern Bue-
nos Aires province, where the largest remaining population
is located (Fernndez et al. 2003, R. Snchez pers. comm.,
Azpiroz unpubl. data). Interestingly, in Buenos Aires many
fields with adequate vegetation characteristics remained
unused, and it has been suggested that factors such as lim-
ited food resources and presence of predators may deter the
02_MS080111.indd 30
species from occupying what seem to be adequate nesting
sites (Fernndez et al. 2003).
CONCLUDING REMARKS
Contrary to expectations, natural grasslands did not support
more grassland-obligate species than did cultivated grass-
lands, although all but one of the species restricted to natu-
ral grasslands were grassland obligates. The latter, however,
did attain higher densities in less modified habitats. The value
of natural grasslands for grassland obligates is further high-
lighted by the fact that most facultative indicator species were
tied to Crop whereas most obligate indicator species were as-
sociated with either Native 1 or Native 2. Because of the fairly
pristine condition of Native 2, bird assemblages there probably
resemble those typical of pre-settlement times in the Northern
Campos of Uruguay more closely. On the basis of our results,
other alternative land-use practices seem to have caused both
an increase in species richness and substantial shifts in spe-
cies composition in the Northern Campos.
During the last two decades (especially since 2000) land
use in Uruguay has shifted toward an expansion of agriculture
and planted pastures, reducing native grasslands traditionally
used for cattle ranching (Martino and Methol 2008). If this
trend continues it could trigger changes in bird distribution and
abundance patterns. The patterns we identify suggest that an
expansion of croplands will probably benefit a few common
species (e.g., the Eared Dove and White-browed Blackbird) but
will be detrimental to several other common and threatened
birds. More land devoted to planted pastures will benefit several
common species (e.g., the Southern Lapwing and Short-billed
Pipit) that do not attain high densities in cropland but will nega-
tively affect birds that are restricted to native grasslands (e.g.,
the Pampas Meadowlark). Although a substantial increase of
land devoted to the Pampas Deer (where sheep are excluded) is
unlikely, at a local scale an expansion of this habitat type would
also benefit both common (e.g., the Grassland Yellow-finch)
and threatened (Ochre-breasted Pipit) birds.
Ours is the first characterization of grassland bird com-
munities inhabiting an agricultural landscape in the Northern
Campos of Uruguay. We found that cultivated grasslands
seem to provide suitable habitat for a substantial proportion
of grassland birds, including some of the species of conserva-
tion concern in the Pampas. Threatened species, however, at-
tained relatively high densities only in natural grasslands. The
availability of opportunities for feeding and breeding seem to
have been an important factor shaping the bird communitys
structure in this region, something that has also been reported
for avian communities in other farmland ecosystems (e.g.,
Sderstrm et al. 2003, Whittingham et al. 2006). Because
these patterns are based solely on presence/absence data and
density estimates by species, the question of whether each
habitat supports viable populations of the species recorded
in it has not been fully addressed. To do so, information on
2/13/09 4:49:24 PM
 Grassland Bird Diversity in Northern Uruguay 31
demographic parameters, such as survival rates and breeding
success in each habitat, need to be considered.
ACKNOWLEDGMENTS
We are indebted to J. P. Castro and family, G. Constantin, M. Berretta,
A. Menndez, E. Muguerza, J. Oviague, and L. Baranov for grant-
ing permission to work on their lands. We thank field assistants and
volunteers for their time and effort in data collection: G. Corts, A.
Ocampo, E. Mndez, A. P. Blazina, C. A. Balbusso, K. Alexander,
R. Snchez, G. Barrenechea, L. Liguori, C. Abud, A. Lafranconi, P.
Rocca, J. P. Bouvet, J. Aldabe, D. Caballero, N. Zalda, G. Arnedillo,
V. Trucco, P. Ducommun, I. Lpez Garca, P. Ferrer, A. Vega, M. Br-
mida, F. Encinas, G. Veiga, F. Garca Olaso, and P. Arbiza. For logis-
tic support we are grateful to G. and E. Battocletti, M. Echenique, L.
Bisio, M. Toucon, E. Mena, N. Rodrguez, R. Nuez, Familia San
Martn, and P. Larrosa. Financial support for this research was pro-
vided by the Wildlife Conservation Society, Rufford Maurice Laing
Foundation, Cleveland Metroparks Zoo and Cleveland Zoological
Society, Neotropical Grassland Conservancy, and Sigma XiThe
Scientific Research Society. Idea Wild and Neotropical Grassland
Conservancy provided additional grants for field equipment. The
University of Missouri-St. Louis Institutional Animal Care and
Use Committee approved all survey methods, and the Direccin de
Recursos Naturales, Ministerio de Agricultura, Ganadera y Pesca
(Uruguay), issued the necessary permits. The manuscript bene-
fited greatly from constructive comments and suggestions from A.
Rodrguez-Ferraro, P. D. Vickery, R. E. Ricklefs, B. Loiselle, and two
anonymous reviewers.
LITERATURE CITED
Askins, R. A., F. Chvez-R amrez, B. C. Dale, C. A. H aas, J. R.
H erkert, F. L. K nopf, and P. D. Vickery. 2007. Conservation
of grassland birds in North America: understanding ecologi-
cal processes in different regions. Ornithological Monographs
64:146.
A zpiroz, A. B. 2001. Aves del Uruguay. Lista e introduccin a su
biologa y conservacin. Aves Uruguay-GUPECA, Montevideo,
Uruguay.
Bellis, L. M., M. B. M artella, and J. L. Navarro. 2004. Habitat
use by wild and captive reared Greater Rheas Rhea americana in
agricultural landscapes in Argentina. Oryx 308:304310.
Bilenca, D., and F. Miarro. 2004. Identificacin de reas valiosas
de pastizal en las pampas y campos de Argentina, Uruguay y sur
de Brasil. Fundacin Vida Silvestre Argentina, Buenos Aires.
BirdLife International. 2000. Threatened birds of the world. Lynx
Edicions, Barcelona.
Brennan, L. A., and W. P. Kuvlesky, Jr. 2005. North American
grassland birds: an unfolding conservation crisis? Journal of
Wildlife Management 69:113.
Bucher, E. H., and M. Nores. 1988. Present status of birds in
steppes and savannas of northern and central Argentina, p. 71
79. In P. D. Goriup [Ed.], Ecology and Conservation of Grassland
Birds. ICBP Technical Publication 7. International Council for
Bird Preservation, Cambridge, UK.
Buckland, S. T., D. R. A nderson, K. P. Burnham, J. L. Laake,
D. L. Borchers, and L. Thomas. 2001. Introduction to distance
sampling. Estimating abundance of biological populations.
Oxford University Press, New York.
Burnham, K. P., and W. S. Overton. 1979. Robust estimation of
population size when capture probabilities vary among animals.
Ecology 60:927936.
02_MS080111.indd 31
Canevari, M., P. Canevari, G. R. Carrizo, G. Harris, J. Rodrguez
M ata, and R. J. Straneck. 1991. Nueva gua de las aves argenti-
nas, vol. 2. Fundacin Acindar, Buenos Aires.
Clarke, K. R., and R. N. Gorley. 2002. PRIMER 5 for Windows,
version 5.2.9. PRIMER-E, Ltd., Plymouth, UK.
Clarke, K. R., and R. M. Warwick. 2001. Change in marine com-
munities: an approach to statistical analysis and interpretation.
2nd ed. PRIMER-E, Ltd., Plymouth, UK.
Collar, N. J., L. P. Gonzaga, N. K rabbe, A. M adroo Nieto, L. G.
Naranjo, T. A. Parker III, and D. C. Wege. 1992. Threatened
birds of the Americas: the ICBP/IUCN Red Data Book. Interna-
tional Council for Bird Preservation, Cambridge, UK.
Comparatore, V. M., M. M. M artnez, A. I. Vasallo, M. Barg,
and J. P. Isacch. 1996. Abundancia y relaciones con el habitat de
aves y mamferos en pastizales de Paspalum quadrifarium (paja
colorada) manejados con fuego (Prov. de Buenos Aires, Argen-
tina). Interciencia 21:228237.
Dufrne, M., and P. Legendre. 1997. Species assemblages and
indicator species: the need for a flexible asymmetrical approach.
Ecological Monographs 67:345366.
Fernndez, G. J., G. Posse, V. Ferretti, and F. M. Gabelli. 2003.
Birdhabitat relationship for the declining Pampas Meadowlark
populations in the southern Pampas grasslands. Biological Con-
servation 115:139148.
Gore, M. E. J., and A. R. M. Gepp. 1978. Las aves del Uruguay.
Mosca Hnos. S.A., Montevideo, Uruguay.
Gotelli, N. J., and G. L. Entsminger [online]. 2006. EcoSim: null
models software for ecology. Version 7. Acquired Intelligence,
Inc., and Kesey-Bear, Jericho, VT. <http://garyentsminger.com/
ecosim.htm> (15 September 2006).
Hines, J. E. [online]. 1996. SPECRICH Software to compute spe-
cies abundance from emperical species abundance distribution
data. USGS-PWRC. <http://www.mbrpwrc. usgs.gov/soft-ware/
specrich.html> (28 October 2006).
Isacch, J. P., M. S. Bo, N. O. M acira, M. R. Demara, and S. Peluc.
2003. Composition and seasonal changes of the bird community
in the west pampa grasslands of Argentina. Journal of Field Orni-
thology 74:5965.
Isacch, J. P., and M. M. M artnez. 2001. Estacionalidad y rela-
ciones con la estructura del habitat de la comunidad de aves de
pastizales de paja colorada (Paspalum quadrifarium) maneja-
dos con fuego en la provincia de Buenos Aires, Argentina. Orni-
tologa Neotropical 12:345354.
K nopf, F. L. 1994. Avian assemblages in altered grasslands. Studies
in Avian Biology 15:247257.
Lezama, F., A. A ltesor, R. J. Len, and J. M. Paruelo. 2006. Het-
erogeneidad de la vegetacin en pastizales naturales de la regin
basltica de Uruguay. Ecologa Austral 16:167182.
M artino, D., and M. Methol. 2008. Cambios en el uso de la tierra,
p. 56117. In PNUMA, GEO Uruguay 2008. Grfica Mosca,
Montevideo, Uruguay.
McCune, B., and M. J. Mefford. 1999. PC-ORD. Multivariate anal-
ysis of ecological data, version 4.28. MjM Software, Gleneden
Beach, Oregon.
Mller, M., R. Spaar, L. Schifferli, and L. Jenni. 2005. Effects of
changes in farming of subalpine meadows on a grassland bird, the
Whinchat (Saxicola rubetra). Journal of Ornithology 146:1423.
Murphy, M. T. 2003. Avian population trends within the evolving
agricultural landscape of eastern and central United States. Auk
120:2034.
OEA. 1992. Estudio ambiental del Uruguay. Organizacin de Esta-
dos Americanos, Washington D. C.
Owens, I. P. F., and P. M. Bennett. 2000. Ecological basis of extinc-
tion risk in birds: habitat loss versus human persecution and
2/13/09 4:49:25 PM
 32 Adrin B. Azpiroz and John G. Blake
introduced predators. Proceedings of the National Academy of
Sciences 97:1214412148.
Owens, R. A., and M. T. Myres. 1973. Effects of agriculture upon
populations of native passerine birds of an Alberta fescue grass-
land. Canadian Journal of Zoology 51:697713.
Patterson, M. P., and L. B. Best. 1996. Bird abundance and nest-
ing success in Iowa CRP fields: the importance of vegetation
structure and composition. American Midland Naturalist 135:
153167.
Sick, H. 1985. Ornitologia Brasileira, uma Introduo. Editora Uni-
versidade de Brasilia, Brasilia, Brazil.
Sderstrm, B., S. K iema, and R. S. R eid. 2003. Intensified agri-
cultural land-use and bird conservation in Burkina Faso. Agri-
culture, Ecosystems and Environment 99:113124.
Soriano, A. 1992. Ro de la Plata Grasslands, p. 367407. In R. T.
Coupland [ed.], Natural grasslands: introduction and western
hemisphere. Ecosystems of the World, Vol. 8A. Elsevier, Amster-
dam, the Netherlands.
SPSS. 2006. SPSS for Windows, Version 15.0. SPSS Inc., Chicago.
Stotz, D. F., J. W. Fitzpatrick, T. A. Parker III, and D. K. Mosko-
vits. 1996. Neotropical birds: ecology and conservation. Univer-
sity of Chicago Press, Chicago.
Sturm, M. 2001. Pampas Deer (Ozotoceros bezoarticus) habitat
vegetation analysis and deer habitat utilization. Salto, Uruguay.
M.Sc. thesis, State University of New York, Syracuse, NY.
Thomas, L., J. L. Laake, S. Strindberg, F. F. C. M arques, S. T.
Buckland, D. L. Borchers, D. R. A nderson, K. P. Burnham,
S. L. Hedley, J. H. Pollard, J. R. B. Bishop, and T. A. M arques.
[online]. 2005. Distance 5.0. Research Unit for Wildlife
Population Assessment, University of St. Andrews, UK. <http://
www.ruwpa.st-and.ac.uk/distance> (24 January 2006).
02_MS080111.indd 32
Tubaro, P. L., and F. M. Gabelli. 1999. The decline of the Pampas
Meadowlark: difficulties of applying the IUCN criteria to neo
tropical grassland birds. Studies in Avian Biology 19:250257.
Tucker, G. M., and M. F. Heath. 1994. Birds in Europe: their con-
servation status. Conservation Series 3, BirdLife International,
Cambridge, UK.
Tyler, S. 2004. Family Motacillidae (Pipits and wagtails), p. 686
786. In J. del Hoyo, A. Elliott, and D. A. Christie [eds.], Hand-
book of the Birds of the World, Vol. 9. Cotingas to pipits and
wagtails. Lynx Edicions, Barcelona.
Verhulst, J., A. Bldi, and D. K leijn. 2004. Relationships between
land-use intensity and species richness and abundance of birds in
Hungary. Agriculture, Ecosystems and Environment 104:465
473.
Vickery, P. D., P. L. Tubaro, J. M. Cardoso Da Silva, B. G. Peter-
john, J. R. H erkert, and R. B. Cavalcanti. 1999. Conservation
of grassland birds in the western hemisphere. Studies in Avian
Biology 19:226.
Whittingham, M. J., C. L. Devereux, A. D. Evans, and R. B.
Bradbury. 2006. Altering perceived predation risk and food
availability: management prescriptions to benefit farmland birds
on stubble fields. Journal of Applied Ecology 43:640650.
Wiens, J. A. 1974. Climatic instability and the ecological satura-
tion of bird communities in North American grasslands. Condor
76:385400.
World R esources Institute [online]. 2007. EarthTrends: envi-
ronmental information. World Resources Institute, Washington,
DC. <http://earthtrends.wri.org> (12 July 2007).
Zimmerman, J. L. 1988. Breeding season habitat selection by the
Henslows Sparrow (Ammodramus heslowii) in Kansas. Wilson
Bulletin 100:1724.
2/13/09 4:49:25 PM
 Grassland Bird Diversity in Northern Uruguay 33

APPENDIX 1. Number of individuals and species found in four grassland habitats, Crop (CR), Pastures (PA), Native 1 (N1), and Native
2 (N2). For the Crop column superscripts denote species recorded during one management phase only. Each species is also classified (+) by
habitat specialization, migration status, and feeding guild. Habitat specialization: obligate, OB; facultative, FA, sensu Vickery et al. (1999).
Migration status: year-round resident, RE; summer resident, SR; nearctic migrant, NM; winter migrant, WM. Feeding guild: carnivore, CA;
granivore, GR; herbivore, HE; insectivore, IN; omnivore, OM.

Habitat
Habitat specialization Migration Feeding guild
Family/species CR PA N1 N2 OB FA RE SR NM WM CA GR HE IN OM
Rheidae
Greater Rhea Rhea americana 41 4 69 50 + + +
Tinamidae
Red-winged Tinamou Rhynchotus 15 9 0 0 + + +
rufescens
Spotted Nothura Nothura maculosa 38 143 142 116 + + +
Ardeidae
Whistling Heron Syrigma sibilatrix 2c 4 1 0 + + +
Ciconiidae
Maguari Stork Ciconia maguari 0 0 0 3 + + +
Accipitridae
White-tailed Kite Elanus leucurus 1 0 0 0 + + +
Cinereous Harrier Circus cinereus 0 1 0 2 + + + +
Charadriidae
Southern Lapwing Vanellus chilensis 29c 146 278 186 + + +
American Golden-Plover Pluvialis 0 0 130 16 + + +
dominica
Tawny-throated Dotterel 0 0 1 0 + + +
Oreopholus ruficollis
Scolopacidae
Upland Sandpiper Bartramia 6c 20 27 15 + + +
longicauda
South American Snipe Gallinago 0 6 9 34 + + +
paraguaiae
Buff-breasted Sandpiper Tryngites 0 0 4 0 + + +
subruficollis
Columbidae
Spot-winged Pigeon Columba 19 0 0 0 + + +
maculosa
Picazuro Pigeon Columba picazuro 23c 10 2 0 + + +
Eared Dove Zenaida auriculata 481 74 0 0 + + +
Psittacidae
Monk Parakeet Myiopsitta monachus 61c 12 0 0 + + +
Strigidae
Burrowing Owl Athene cunicularia 0 5 3 4 + + +
Caprimulgidae
Nacunda Nighthawk Podager 0 0 2 0 + + +
nacunda
Picidae
Field Flicker Colaptes campestris 1c 25 20 1 + + +
Furnariidae
Common Miner Geositta cunicularia 0 0 18 0 + + +
Rufous Hornero Furnarius rufus 2 0 0 0 + + +
Firewood-Gatherer Anumbius 0 5 5 4 + + +
annumbi
Tyrannidae
Bearded Tachuri Polystictus 0 3 0 0 + + +
pectoralis
Grey Monjita Xolmis cinerea 29 1 2 0 + + +
White Monjita Xolmis irupero 3c 2 1 0 + + +
Black-and-white Monjita 20 0 0 0 + + +
Heteroxolmis dominicana

(continued)

02_MS080111.indd 33 2/13/09 4:49:26 PM

 34 Adrin B. Azpiroz and John G. Blake

APPENDIX 1. (Continued)

Habitat
Habitat specialization Migration Feeding guild
Family/species CR PA N1 N2 OB FA RE SR NM WM CA GR HE IN OM
Chocolate-vented Tyrant Neoxolmis 0 0 2 1 + + +
rufiventris
Cattle Tyrant Machetornis rixosus 4 1 5 0 + + +
Fork-tailed Flycatcher Tyrannus 14 12 5 0 + + +
savana
Hirundinidae
White-rumped Swallow Tachycineta 43 45 21 13 + + +
leucorrhoa
Gray-breasted Martin Progne 4a 3 0 6 + + +
chalybea
Brown-chested Martin Progne 22 22 7 6 + + +
tapera
Blue-and-white Swallow 0 0 2 0 + + +
Notiochelidon cyanoleuca
Tawny-headed Swallow 1a 0 8 0 + + +
Alopochelidon fucata
Rough-winged Swallow 0 0 2 0 + + +
Stelgidopteryx ruficollis
Motacillidae
Short-billed Pipit Anthus furcatus 0 118 292 216 + + +
Hellmayrs Pipit Anthus hellmayri 5 0 23 48 + + +
Ochre-breasted Pipit Anthus 0 0 5 92 + + +
nattereri
Emberizidae
Rufous-collared Sparrow Zonotrichia 6b 0 1 0 + + +
capensis
Grassland Sparrow Ammodramus 29 43 32 13 + + +
humeralis
Long-tailed Reed-Finch 1b 0 0 0 + + +
Donacospiza albifrons
Grassland Yellow-Finch Sicalis 15 9 26 148 + + +
luteola
Great Pampa-Finch Embernagra 12 2 0 0 + + +
platensis
Dark-throated Seedeater Sporophila 3 4 0 0 + + +
ruficollis
Icteridae
Pampas Meadowlark Sturnella 0 0 33 4 + + +
defilippii
White-browed Blackbird Sturnella 460 416 0 55 + + +
superciliaris
Brown-and-Yellow Marshbird 23 2 43 3 + + +
Pseudoleistes virescens
Bay-winged Cowbird Agelaioides 15 0 0 0 + + +
badius
Shiny Cowbird Molothrus 127c 3 4 2 + + +
bonariensis
Total individuals 1555 1150 1225 1038
Total species 33 30 34 24 23 27 37 8 3 2 5 10 1 28 6
a
Recorded only when field grown to barley.
b
Recorded only when field grown to sunflower.
c
Recorded only when field used as pasture.

02_MS080111.indd 34 2/13/09 4:49:27 PM

 Grassland Bird Diversity in Northern Uruguay 35

APPENDIX 2. Number of observations and model selection of

detection functions of grassland birds in four habitats in northern
Uruguay. Pluvialis dominica, Bartramia longicauda, Progne chaly-
bea, and Sturnella defilippii had fewer than 60 observations each,
so detection functions for these species included additional obser-
vations of species of similar characteristics and behavior (Vanellus
chilensis for P. dominica and B. longicauda, Tachycineta leucorrhoa
for P. tapera, and Sturnella superciliaris for S. defilippii).

Species na Model selectedb mc

Rhea americana 81 HR + cosine 2
Nothura maculosa 131 HR + cosine 2
Vanellus chilensis 351 HN + cosine 3
Pluvialis dominica 367 HN + cosine 3
Bartramia longicauda 377 HN + cosine 3
Zenaida auriculatad 71 HN + cosine 2
Zenaida auriculatae 62 HR + cosine 2
Tachycineta leucorrhoa 80 HN + cosine 1
Progne tapera 116 HN + cosine 1
Anthus furcatus 498 HR + cosine 5
Anthus hellmayri 63 UN + polynomial 1
Anthus nattereri 83 HN + cosine 1
Ammodramus humeralis 68 HR + cosine 3
Sicalis luteola 131 UN + cosine 1
Sturnella defilippii 266 HR + polynomial 5
Sturnella superciliarisd 89 HR + cosine 4
Sturnella superciliarise 264 UN + cosine 4
a
Number of observations.
b
HR, hazard-rate base function; HN, half-normal base function;
UN, uniform base.
c
Number of parameters in detection function.
d
Detection function built exclusively with data collected from
September 2004 to March 2005, corresponding to the crop phase
(barley and sunflower fields) of crop habitat.
e
Detection function built with data collected from May to November
2005, corresponding to the pasture phase of crop habitat, as well
as with data from other habitats for the total sampling period.

02_MS080111.indd 35 2/13/09 4:49:27 PM