Perspectives in Ecology and Conservation Perspectives in Ecology and Conservation
Research Letters
Bird-grassland associations in protected and non-protected areas in southern Brazil
Lucilene Inês Jacoboskia,, , Raquel Klein Paulsenb, Sandra Maria Hartza
a Laboratório de Ecologia de Populações e Comunidades (LEPeC), Programa de Pós-Graduação em Ecologia, Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Rio Grande do Sul, Brazil
b Laboratório de Ecologia de Populações e Comunidades (LEPeC), Graduação em Ciências Biológicas, Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Rio Grande do Sul, Brazil
Received 07 February 2017, Accepted 03 May 2017

Conversion of grasslands in other land uses is the main threat for grassland birds. We investigated habitat use by grassland birds in Permanent Preservation Areas surrounded by Eucalyptus plantation stands and non-protected grasslands (grazed native grasslands). As there is no evidence whether Permanent Preservation Areas are effective habitats for grassland avifauna, we compared such areas with grazed native grasslands, regarding richness, abundance and composition of grassland birds. Short and tall grassland bird specialists were recorded, with some species related to the non-protected areas and others to the protected areas. Thus, both areas are fundamental habitats for conservation of grassland bird species. We highlight the importance of grassland mosaics to maintain grassland bird species. Furthermore, we highlight the necessity of maintaining Permanent Preservation Areas in ranching and cropland areas, in order to connect grassland remnants in an extremely converted landscape and to conserve especially bird species that are more sensitive to disturbances.


  • We tested the importance of protected and non-protected area for conservation of grassland birds.

  • Maintaining grassland mosaics is essential for tall and short grassland birds.

  • It is important to maintain PPAs in grazing areas.

Grassland birds, Pampa, Tree plantation, Grazed grasslands
This article is only available in PDF

Grassland ecosystems are among the most threatened at global scale due to the great disparity between habitat loss and the low percentage of protected area (Hoekstra et al., 2005). Such ecosystems are submitted to strong anthropic pressure owing mainly to changes in land-use caused by monocultures (Azpiroz et al., 2012; Develey et al., 2008). In Brazil, grasslands of Pampa biome have a high degree of degradation, in the state of Rio Grande do Sul more than 50% of native grasslands has been converted mainly for agriculture and forest plantations (Bencke, 2009; Fontana et al., 2016).

Afforestation and agriculture with nonnative species have expanded over grazed native grasslands in extensive areas of Pampa biome (Azpiroz et al., 2012). The traditional cattle production is a compatible activity with conservation biodiversity if adequately managed (Develey et al., 2008; Isacch and Cardoni, 2011). The moderate grazing does not cause the suppression of native vegetation, maintaining the main characteristics of grassland ecosystems (Overbeck et al., 2007). On the other hand, tree monocultures totally replace the native local vegetation structure, affecting distribution of bird populations by reducing availability of resources that are important for nesting and feeding (Codesido et al., 2008). Tree plantations are known to alter the structure of bird communities in forest ecosystems, benefiting generalist and edge bird species (Jacoboski et al., 2016). However, in grassland landscapes, tree plantations can have more detrimental effects in birds than in forest ecosystems (Filloy et al., 2010).

The legal protection can effectively avoid conversion of native grasslands into other uses, and thus prevent complete loss of grassland biodiversity (Overbeck et al., 2007). Legal protection measures proposed by the Brazilian Forest Code determine the areas that must be preserved and which regions are allowed for cultivation, law n° 12.651/12 (CFB, 2012). Between the vegetation to be protected, is the marginal vegetation surrounding aquatic ecosystems. This vegetation, denominated Permanent Preservation Area (PPA), must be maintained by the delimitation of a buffer zone for both margins, which varies according to the width of a water course. A PPA consists of a protected area, covered or not by native vegetation, with the environmental function of preserving water resources, landscape, geological stability, and biodiversity, as well as of facilitating genetic flow of fauna and flora and protecting the soil. In tree plantations, the regulation of law is performed, however, in cattle ranching farms or soybean, such regulation is often not maintained, and when maintained, the tendency is to consider forest vegetation more than non-forest vegetation, even in grassland biomes (Develey et al., 2008; Overbeck et al., 2015).

Our goal was to investigate whether protected native grasslands (PPAs within Eucalyptus plantation stands) are viable to grassland birds. Furthermore, we aimed to identify whether these species are different from those recorded in non-protected areas (grazed native grasslands). Toward this aim we compared richness, abundance and composition of bird species between PPAs and grazed native grasslands. We hypothesized that (1) PPAs are viable for grassland species use because Eucalyptus stands are located in a matrix of grassland vegetation and possess connections among native grassland areas; (2) species composition should differ between protected and non-protected grasslands areas due to their differences in vegetation structure caused by grazing.

Material and methodsStudy area

We conducted the study in two types of grassland vegetation: eight non-grazed PPAs located within Eucalyptus plantation stands (surrounded by eucalyptus), and eight areas of grazed native grasslands, for a total of 16 sample sites. These sites are located in the South Region of Brazil, in the municipalities of São Gabriel (30°, 20′11″S, 54°, 19′12″W) and Rosário do Sul (30°, 15′30″S, 54°, 54′51″W). The study area is located in the Pampa biome. Each Eucalyptus stand comprises a distinct silviculture area for cellulose production. The eucalyptus plantations had ages of cultivation between four and six years, which superior height to 15m. For each of the eight Eucalyptus stands, adjacent areas of grazed grassland were sampled (Fig. 1).

Fig. 1.

Location of the study area in the state of Rio Grande do Sul. Circles represents sampled grazed native grasslands and triangle Permanent Preservation Area.

The sampled sites were first selected using Google Earth (2014) images and subsequently checked in the field. The main characteristics taken into account for site selection were that sites for both PPAs and grassland areas should possess native vegetation and, specifically for PPAs, they should have predominantly grassland vegetation and be 100m wide minimum. The PPAs composed of mostly forest vegetation were not included. The selected PPAs are located within Eucalyptus stands, protect margins of small water courses and have a minimum width of 100m and varied length that depends on the forest stand size. No type of management is allowed within PPAs (e.g. ranching, fire). For the grazed native grassland, we included the sites that experienced low to intermediate grazing intensity and all pasture sites also had water courses, though without PPA delimitation.

Bird sampling

We carried out bird sampling once at each one of the 16 sites during the austral spring, between 2014 and 2015, including two reproductive seasons. Bird sampling was undertaken by applying the point count method (Bibby et al., 1992). We distributed the count points according to site size, ranging from three to nine points equally separated from each other by a distance of 200m. We recorded all bird species seen or heard during a period of 10min within a 50m radius around each one of the points. We sampled 50 count points in each type of grassland, PPA or grazed native grassland (n=100). Observations were limited to a 50m radius in order to maximize detectability and to minimize the potential for observer error in identifying cryptic grassland species over long distances (Hovick et al., 2015). Sampling always started 10min after sunrise and lasted for a maximum of 3h. We performed all samplings during days without wind or rainfall. Birds in flight were not considered. Only grassland bird species were considered in the analyses, following the classification of Vickery et al. (1999) and Azpiroz et al. (2012). We also categorized the birds in relation to the height of the used vegetation, short or tall. Species restricted to short grass (<20cm) or tall grass (>40cm) and those that use both short and tall grasslands (sensuAzpiroz et al., 2012), besides on our own personal observations. We followed the Red Lists of IUCN (IUCN, 2015) and Rio Grande do Sul (DOE, 2014) to verify whether recorded species found were of conservation concern.

Vegetation sampling

To characterize vegetation of each sample site, we measured plant height and density using a stick subdivided into centimeters (Dias and Scarano, 2007). This method consists of positioning the stick vertically over the soil to record: (1) the number of times the stick was touched by vegetation (density); (2) height at which vegetation touches the stick. We performed four measurements of height and density at each bird count point (n=400). Each measure was undertaken toward one of the four cardinal points (east, west, north and south), starting from the center count point in the distances of 5, 15, 30 and 50m. Distance and direction to place the vegetation sampling points were randomly selected. To carry out the statistical analyses, we used mean values of plant height and density for each count point.

Statistical analyses

Initially, to check whether there was spatial autocorrelation in the sampled sites, we performed a Mantel correlation test between a matrix generated from taxonomic composition data (Bray–Curtis dissimilarity) and a matrix generated from geographical coordinate data of each count point (Euclidean distance).

To evaluate differences in richness and abundance of bird species, as well as in plant height and density between sample sites, we performed Analysis of Variance (ANOVA). To test differences in the species composition, we carried out a Multivariate Analysis of Variance (MANOVA), using Euclidean distance as the dissimilarity measure, both with 999 permutations. To test the influence of vegetation structure on taxonomic composition of grassland bird species, we employed a Canonical Correspondence Analysis (CCA). MANOVA and CCA were performed with logarithmized species abundance data (logx+1). We tested CCA significance with an ANOVA. Analyses were conducted with vegan package (Oksanen et al., 2015) on R 3.2.2 software (R Development Core Team, 2015).


We recorded 36 grassland bird species (Table 1), 24 in the PPAs and 27 in the grazed native grasslands. Most of the recorded species uses both tall and short grasslands. However, only 15 species were shared between areas. The most abundant species were Zonotrichia capensis (n=44) and Ammodramus humeralis (n=49), in the PPAs and in the grazed native grassland, respectively. In relation to sample units, the Mantel test indicated that count points were independent as there was no significant spatial correlation (r=0.02, p=0.15). Both species richness and abundance did not present significant differences between sampled areas (F1,14=1.91, p=0.21 and F1,14=6.61, p=0.07, respectively). On the other hand, species composition differed significantly between areas (F1,14=2.85, p=0.001). A clear segregation of species was observed between PPAs and grazed native grassland in the CCA (Fig. 2).

Table 1.

Total species number of grassland bird species sampled in Permanent Preservation Areas (PPAs) and grazed native grasslands (GNG), preferred grass height (sensuAzpiroz et al., 2012) and conservation status (IUCN, 2015; DOE, 2014). Acronyms of Fig. 2 also shown.

Species  PPAs  GNG  Grass heighta  Conservation statusb 
Agelaioides badius (Agba)  Broad  LC 
Ammodramus humeralis (Amhu)  23  49  Broad  LC 
Anthus hellmayri (Anhe)  11  Short  LC 
Anumbius annumbi (Anan)  Broad  LC 
Cariama cristata (Cacr)  Broad  LC 
Cistothorus platensis (Cipl)  Tall  NT (R) 
Colaptes campestris (Coca)  Broad  LC 
Columbina talpacoti (Cota)  Broad  LC 
Culicivora caudacuta (Cuca)  Tall  VU (R, G) 
Donacospiza albifrons (Doal)  Broad  LC 
Emberizoides herbicola (Emhe)  27  19  Tall  LC 
Emberizoides ypiranganus (Emyp)  10  Tall  LC 
Embernagra platensis (Empl)  27  Tall  LC 
Falco sparverius (Fasp)  Broad  LC 
Furnarius rufus (Furu)  Broad  LC 
Gallinago paraguaiae (Gapa)  Short  LC 
Geranoaetus albicaudatus (Geal)  Broad  LC 
Gnorimopsar chopi (Gnch)  Broad  LC 
Mimus saturninus (Misa)  Broad  LC 
Molothrus bonariensis (Mobo)  Broad  LC 
Nothura maculosa (Noma)  Broad  LC 
Progne tapera (Prta)  Broad  LC 
Pseudoleistes guirahuro (Psgu)  Broad  LC 
Rhynchotus rufescens (Rhru)  Broad  LC 
Sicalis luteola (Silu)  33  Broad  LC 
Sporophila caerulescens (Spca)  Broad  LC 
Sporophila cinnamomea (Spci)  Tall  VU (G) 
Sporophila pileata (Sppi)  Tall  VU (R) 
Tapera naevia (Tapa)  Tall  LC 
Tyrannus savana (Tysa)  20  Broad  LC 
Vanellus chilensis (Vach)  11  Short  LC 
Volatinia jacarina (Voja)  Tall  LC 
Xanthopsar flavus (Xafl)  13  Broad  VU (R, G) 
Xolmis irupero (Xoir)  Broad  LC 
Zenaida auriculata (Zeau)  Broad  LC 
Zonotrichia capensis (Zoca)  44  38  Broad  LC 
Total  194  255     

Association with the height vegetation, short or tall; the broad category indicates species that are less dependent of the grasslands height.


VU, vulnerable; NT, near threatened; LC, least concern; R, regional threat; G, global threat.

Fig. 2.

Ordination diagram generated by the Canonical Correspondence Analysis (CCA) of bird species composition in relation to vegetation structure. Dark circles represent areas of grazed native grasslands; empty circles represent Permanent Preservation Areas (PPAs). Arrows represent mean plant height and mean plant density, respectively. Species names acronyms in Table 1.

The CCA showed an association between birds species composition and vegetation structure. For example, typical tall grass species are associated with PPAs, while grazing selects species are less dependent on tall grasslands.

However, the CCA model was not significant when we tested it with an ANOVA including the two vegetation variables. Both variables were significant when we analyzed them separately: height (F1,14=1.84, p=0.002) and density (F1,14=1.7, p=0.02). Thus, mean plant height explained 12% of the variation in species composition and mean plant density explained 11%. Also, we detected significant differences in plant height and density when we tested PPAs versus grazed native grassland sites with an ANOVA (F1,14=68, p=0.001 and F1,14=40, p=0.001), respectively (Fig. 3).

Fig. 3.

Boxplots of mean plant height (A) and mean plant density in each one of the sampling sites (B), showing differences among sampled sites, with their maximum and minimum values and standard error. Different letters indicate significant differences between treatments.

Four of the species recorded are threatened in regional or global categories (Sporophila pileata, Sporophila cinnamomea, Culicivora caudacuta, Xanthopsar flavus) and one as regionally near threatened species (Cistothorus platensis). The three first species were exclusively observed in the PPAs, whereas X. flavus was only found in the grazed native grasslands. C. platensis was recorded in both areas.


Our results demonstrate a turnover of species between areas, bird species recorded in the PPAs that depend on tall grass and preserved vegetation (e.g. Sporophila sp.) were replaced by species associated with short grass and (besides) by species that are less dependent on the height of the vegetation in the grazed native grasslands (e.g. Anthus sp.). These results demonstrate that presence of PPAs and grazed native grasslands with moderate levels of grazing are essential for the conservation of grassland birds, because two groups of species are benefited, short and tall grass dependent. Grassland birds differ in their sensitivity to disturbance of grasslands by grazing (Azpiroz et al., 2012). Areas with lower or no grazing are necessary to ensure viable populations of specialized tall grass species (Isacch and Cardoni, 2011). On the other hand, in grazed grasslands, a short grass assemblage of birds is benefited (Azpiroz et al., 2012).

Specific variations in habitat characteristics, such as the difference in recorded plant height and density, demonstrated to determine differences in bird species composition in the study areas. In the studies developed by Fisher and Davis (2010) and Dias et al. (2014), the authors indicated that plant height was the main factor associated with diversity and composition of bird species in grasslands. This difference in vegetation demonstrates the importance of grassland mosaics with plants of different heights for bird diversity. Many grassland birds require a mosaic with different habitat patches to fulfill their reproductive requirements (Dias et al., 2014). Currently, these mosaics are generated by patches with different grazing intensity, thereby contributing to the conservation of a great part of bird diversity (Isacch and Cardoni, 2011). On the other hand, Vickery et al. (1999) recommend that some areas should not be managed, for obtaining the maximum diversity of grassland birds. They also recognize that specific management practices can satisfy the specific requirements of a subset of species, but cannot benefit the others. Thus, this habitat heterogeneity is necessary to ensure species coexistence and maintaining the diversity of grassland bird communities (Codesido et al., 2013; Dias et al., 2014; Hovick et al., 2015).

The register of threatened X. flavus in grazed native grasslands demonstrated the importance of adequate management in private farms for maintain grassland-associated birds (Develey et al., 2008; Fontana et al., 2016). Furthermore, three threatened bird species found in our study were exclusive to the PPAs. These species are typical of tall and well preserved grasslands, in addition to being sensitive to any sort of disturbance (Bencke, 2009). The presence of threatened species in PPAs is satisfactory from the conservation point of view. However, we do not know whether these populations will remain viable in the long run, since indirect effects of afforestation in adjacent areas are still scarcely known. Indirect effects of neighboring tree plantations can affect the physical and biotic characteristics of adjacent areas, such as facilitation of shrub encroachment and increase of nest predation rates (Azpiroz et al., 2012; Reino et al., 2010).

Our results demonstrate the importance of maintaining protected and non-protected grasslands for conservation of grassland bird species. However, it is fundamental to investigate whether populations in the PPAs remain stable or these areas are only a temporary refuge in the face of imminent extinction. Furthermore, the maintenance of grassland PPAs and managed grasslands in a sustainable way might favor a high diversity of bird species in the study region, and help to avoid the processes that lead to extinction of strictly grassland species. Our results also highlighted the necessity of maintaining PPAs, especially in agricultural and grazing areas. Although PPAs in grasslands are protected by law, they are often neglected. For example, in sampled grazed native grasslands, grassy vegetation along the water courses is not protected, and thus cattle have free access to these areas. Therefore, the maintenance of grasslands PPAs by the landowners might be a useful tool to increase protection of the grasslands vegetation in private land, since landowners can act as important agents for the conservation of birds in the grasslands of south Brazil.


This work was supported by CMPC Celulose Riograndense.

Conflicts of interest

The authors declare no conflicts of interest.


We thank E.F. de Araújo who provided information on the study area. We also thank the landowners who authorized bird sampling on their properties. We are grateful to the field assistance of A.M. Pezda, J.K.F. Mähler Jr. and B.A. Maslak. L.I.J. received a scholarship from CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior). S.M.H. holds a research grant from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), process 306816/2010-5. Thanks to two anonymous reviewers for all suggestions that improved our manuscript.

Azpiroz et al., 2012
A.B. Azpiroz,J.P. Isacch,R.A. Dias
Ecology and conservation of grassland birds in southeastern South America: a review
J. Field Ornithol., 83 (2012), pp. 217-246
Bencke, 2009
G.A. Bencke
Diversidade e conservação da fauna dos Campos do Sul do Brasil
Campos Sulinos – conservação e uso sustentável da biodiversidade, pp. 101-121
Bibby et al., 1992
C. Bibby,N. Burgess,D. Hill
Bird Census Techniques
Academic Press, (1992)
CFB, 2012
Código Florestal Brasileiro. Lei N° 12.651, de 25 de maio de 2012
(2012) (accessed 18.03.15)
Codesido et al., 2008
M. Codesido,C.M. González-Fischer,D.N. Bilenca
Asociaciones entre diferentes patrones de uso de la tierra y ensambles de aves en agroecosistemas de la región pampeana, Argentina
Ornitol. Neotrop., 19 (2008), pp. 575-585
Codesido et al., 2013
M. Codesido,C.M. González-Fischer,D.N. Bilenca
Land bird assemblages in different agricultural landscapes: a case study in the Pampas of Central Argentina
BioOne, 115 (2013), pp. 8-16
Develey et al., 2008
P.F. Develey,R.B. Setubal,R.A. Dias
Conservação das aves e da biodiversidade no bioma Pampa aliada a sistemas de produção animal
Rev. Bras. Ornitol., 16 (2008), pp. 308-315
Dias and Scarano, 2007
A.T.C. Dias,F.R. Scarano
Clusia as nurse plant
Clusia – A Woody Neotropical Genus with Remarkable Plasticity and Diversity, pp. 55-72
Dias et al., 2014
R.A. Dias,V.A.G. Bastazini,A.T. Gianuca
Bird-habitat associations in coastal rangelands of southern Brazil
Iheringia Sér. Zool., 104 (2014), pp. 200-208
DOE, 2014
Diário Oficial do Estado do Rio Grande do Sul. Espécies da Fauna Silvestre Ameaçadas de Extinção no Estado do Rio Grande do Sul. Decreto n° 51.797, de 08 de Setembro de 2014
Porto Alegre, RS, Brasil, (2014)
Filloy et al., 2010
J. Filloy,G.A. Zurita,J.M. Corbelli
On the similarity among bird communities: testing the influence of distance and land use
Acta Oecol., 36 (2010), pp. 333-338
Fisher and Davis, 2010
R.J. Fisher,S.K. Davis
From Wiens to Robel: a review of grassland-bird habitat selection
J. Wildl. Manag., 74 (2010), pp. 265-273
Fontana et al., 2016
C.S. Fontana,G. Dotta,C.K. Marques
Conservation of grasslands birds in south Brazil: a land management perspective
J. Nat. Conserv., 14 (2016), pp. 83-87
Google Earth, 2014
Google Earth
Google Earth, Version 7.1.5
(2014) (accessed 12.11.14)
Hoekstra et al., 2005
J.M. Hoekstra,T.M. Boucher,T.H. Ricketts
Confronting a biome crisis: global disparities of habitat loss and protection
Ecol. Lett., 8 (2005), pp. 23-29
Hovick et al., 2015
T.J. Hovick,L.D. Elmore,S.D. Fuhlendorf
Spatial heterogeneity increases diversity and stability in grassland bird communities
Ecol. Appl., 25 (2015), pp. 662-672
Isacch and Cardoni, 2011
J.P. Isacch,D.A. Cardoni
Different grazing strategies are necessary to conserve endangered grassland birds in short and tall salty grasslands of the flooding pampas
Condor, 113 (2011), pp. 724-734
IUCN, 2015
IUCN Red List of Threatened Species, Version 2015.4
(2015) (accessed 19.03.15)
Jacoboski et al., 2016
L.I. Jacoboski,A. Mendonça-Lima,S.M. Hartz
Structure of birds communities in eucalyptus plantations: nestedness as a pattern on species distribution
Braz. J. Biol., 76 (2016), pp. 583-591
Oksanen et al., 2015
J. Oksanen,F.G. Blanchet,M. Friendly
Vegan: Community Ecology Package. Rpackage Version 2.3-1
(2015) (accessed 21.09.15)
Overbeck et al., 2007
G.E. Overbeck,S.C. Müller,A. Fidelis
Brazil's neglected biome: the south Brazilian campos
Perspect. Plant Ecol. Evol. Syst., 9 (2007), pp. 101-116
Overbeck et al., 2015
G.E. Overbeck,E. Vélez-Martin,F.R. Scarano
Conservation in Brazil needs to include non-forest ecosystems
Divers. Distrib., 21 (2015), pp. 1455-1460
R Development Core Team, 2015
R Development Core Team
R: A Language and Environment for Statistical Computing
Vienna, Austria, (2015) (accessed 21.09.15)
Reino et al., 2010
L. Reino,M. Porto,R. Morgado
Does afforestation increase bird nest predation risk in surrounding farmland?
Forest Ecol. Manag., 260 (2010), pp. 1359-1366
Vickery et al., 1999
P.D. Vickery,P.L. Tubaro,J.M.C. Da Silva
Conservation of grassland birds in the western hemisphere
Copyright © 2017. Associação Brasileira de Ciência Ecológica e Conservação