Introduction
The Azure-rumped Tanager Tangara cabanisi is a species of restricted range occurring in the Pacific slope mountains of Guatemala and Chiapas, Mexico, where it has been reported from a narrow altitudinal belt between 1,000 and 1,700 m (Parker et al. Reference Parker, Hilty and Robbins1976, Hilty and Simon Reference Hilty and Simon1977, Vannini Reference Vannini1989, Heath and Long Reference Heath and Long1991, Cooper Reference Cooper2003, Tenez Reference Tenez2005, Valdez et al. Reference Valdez, Marroquín, Orellana, Pérez, Sandoval, Villar, Godínez, Acevedo, Morales and Reyes1999, Eisermann and Avendaño Reference Eisermann, Avendaño and Cano2006, Reference Eisermann and Avendaño2007a). South of Lake Atitlán it was first reported in the Pachuj Reserve in 1999, later also in the Los Andes Reserve in 2000 (Eisermann and Avendaño Reference Eisermann, Avendaño and Cano2006), and in the Los Tarrales Reserve in 2004 (Jones Reference Jones2004).
Information on the ecology of the Azure-rumped Tanager originates mainly from observations in the Sierra Madre de Chiapas. Brodkorb (Reference Brodkorb1939) collected the first specimen there in cloud forest, and Heath and Long (Reference Heath and Long1991) provided information on habitat preferences, where by this time it had only been recorded in the interior and on the edge of broadleaf forests. They reported Ficus cookii as the dominant tree species in Azure-rumped Tanager habitat in the El Triunfo Biosphere Reserve. Ficus cookii belongs to the F. aurea species complex (Berg Reference Berg2007). Heath and Long (Reference Heath and Long1991) estimated suitable tanager habitat as 1,125 km2 in Chiapas. The species is considered ‘Endangered’ at a global level because of its small area of distribution and loss of habitat, primarily through the replacement of broadleaf forests by coffee plantations (Collar et al. Reference Collar, Gonzaga, Krabbe, Madroño Nieto, Naranjo, Parker and Wege1992, BirdLife International 2008).
The Guatemalan Pacific slope highlands are one of the country’s main areas for coffee cultivation. Recently, the Azure-rumped Tanager has been observed in both habitat types, humid broadleaf forest and coffee plantations (Dietsch Reference Dietsch2000, BirdLife International 2008). For Guatemala, the extent of suitable habitat has not yet been quantified. Here we quantify the use by the species of forest and coffee plantations, and then analyse the extent of suitable habitat throughout Guatemala. Because we observed feeding tanagers mainly in Ficus aurea trees, we furthermore determine the correlation of the population density of this tree with the abundance of Azure-rumped Tanager. We provide implications for conservation of this species.
Methods
Study area
The Azure-rumped Tanager’s known area of distribution in the western Pacific slope mountains of Guatemala is characterised by a chain of several major volcanoes. The altitudinal belt between 1,000 and 1,900 m is shaped by very pronounced slopes of > 30%. The mean annual rainfall amounts to 2,000–4,600 mm, but can locally reach a maximum of 6,000 mm annually. The wet season extends from April to November with a mean monthly precipitation of 150–800 mm, and the dry season from December to March with mean monthly precipitation of 5–150 mm (MAGA 2002).
Annual mean minimum temperature is 15º C; annual mean maximum is 25º C (MAGA 2002). The natural climax vegetation in this area is humid evergreen broadleaf forest, with broadleaf-conifer forest in higher and less humid areas. Much of the forest < 1,600 m altitude has been replaced by coffee plantations and other agricultural crops in the 19th century.
To determine habitat use by Azure-rumped Tanagers, we conducted counts along trails in the neighbouring private nature reserves Los Tarrales, Los Andes, and San Jerónimo Miramar, department Suchitepéquez, where the species had been recorded previously (Eisermann and Avendaño Reference Eisermann and Avendaño2007a). The rectangle encompassing all trails had a size of 1,950 ha (geographic coordinates of the centre point of the rectangle: 14º 32’ N 91º 10’ W; Figure 1)
Study area at Atitlán volcano. The inset map shows the location of the study area in Guatemala.
Landcover
For land cover analyses in the potential area of distribution of Azure-rumped Tanager in Guatemala we used the most recent national landcover mapping at a scale of 1:50,000, based on Landsat images and ground truthing with local experts (MAGA 2006). Land cover in the study area on Atitlán volcano was mapped based on orthoimages with a ground resolution of 0.4 m, based on aerial photographs from December 2005 (MAGA 2010) and ground-truthed field observation in 2008. We mapped Ficus aurea trees within a total strip width of 120 m along all transects from orthoimages with some ground-truthing. Software ESRI ArcView GIS 3.2 was used for all spatial analyses.
Vegetation structure
We produced foliage-height profiles modified after Blondel and Cuvillier (Reference Blondel and Cuvillier1977) by determining vegetation density in 20 strata. We used a 9 m long telescopic pole, which was marked from 0.5 to 9 m, and a laser rangefinder (Nikon ProStaff Laser 440). Along 15 of 21 transects, four evenly spaced sites were defined using a handheld GPS device. At each site, foliage coverage was measured at six points at the perpendicular distance of 10, 20, and 30 m on both sites of the transect. At each point the telescopic pole was positioned vertically, noting whether plants touched the pole on marks at 0.5, 1, 1.5, 2, 2.5, 3, 4, 5, 6, 7, 8, and 9 m. Above this height, we used the laser rangefinder to determine whether plants touched at the imaginary elongation of the pole at 10, 12, 15, 20, 25, 30, 35, and 40 m. All measurements were taken by the same observer (KE). The percentage of presence of plant parts was used as vegetation density index.
Bird surveys
Counts of Azure-rumped Tanagers were conducted along 21 line transects in forest and in coffee plantations (Table 1, Figure S1, in online supplementary material) applying the distance sampling method (Buckland et al. Reference Buckland, Anderson, Burnham, Laake, Borchers and Thomas2001). Because habitat distribution was patchy, none of the transects in coffee plantations was farther than 150 m away from the nearest forest edge. After familiarisation with the different vocalisations of the tanagers (Eisermann et al. Reference Eisermann, López, Berry, de León Lux and Burge2011), two observers (SA, GL) walked along the transect at an average speed of 1 km h−1 between 06h15 and 10h00 (sometimes until 15h30) under favourable weather conditions (i.e. without rain and no or low wind [wind speed not more than 3 on the Beaufort scale]). All tanagers heard and seen were recorded, together with the following data: number of individuals per flock, perpendicular distance from the transect (measured with a laser rangefinder), estimated height of birds above ground and height of vegetation at the same point. Geographic coordinates of all records were recorded on the trails using a handheld GPS device (Garmin 60CSx). When tanagers were observed feeding, food source was noted. Voucher specimens of food plants were deposited in the herbarium BIGU of the School of Biology of San Carlos University, or in the herbarium AGUAT of the Faculty of Agronomy of the San Carlos University, both located in Guatemala City (Table S1). Coincidental observations of Azure-rumped Tanager were recorded together with the geographic coordinates determined with a GPS device.
Number of Azure-rumped Tanagers recorded at each transect line with an effective strip width of 60 m.
* ci: intensive coffee plantation; cn: non-intensive coffee plantation; fh: forest at 1,400–1,900 m; fl: forest at 900–1,300 m
Available foot trails were used for line transect counts, which were repeated between one and 21 times from 19 June to 8 August 2008. Because data from repeated transect lines are not independent, we pooled data from all repetitions and calculated with the total transect length of all repetitions, as recommended by Buckland et al. (Reference Buckland, Anderson, Burnham, Laake, Borchers and Thomas2001). Density was estimated using software Distance 5.0 Release 2 (Thomas et al. Reference Thomas, Laake, Strindberg, Marques, Buckland, Borchers, Anderson, Burnham, Hedley, Pollard, Bishop and Marques2006).
To compare mean encounter rates between forest and coffee plantations, we truncated data at 30 m. Each transect line was considered a replication unit and differences in mean values where compared with a randomisation test with α = 0.05 using SsS software (Engel Reference Engel1998). Means are given together with standard deviation if not otherwise indicated. Geometric Mean Regression (Ricker Reference Ricker1984, Krebs Reference Krebs1999) was applied to analyse the correlation between the distribution of Ficus aurea trees and encounter rate of Azure-rumped Tanagers using SsS software (Engel Reference Engel1998) with α = 0.05.
Distribution and population size of Azure-rumped Tanager in Guatemala
As a basis for derivation of total population size of Azure-rumped Tanagers in Guatemala, we compiled data on the presence of the species in other areas from published references, from our own occasional observations, from data in eBird Guatemala (http://ebird.org/content/guatemala/), and from Christmas Bird Counts on Atitlán volcano (http://www.audubon.org/bird/cbc/hr/index.html). We calculated the total population size for Guatemala assuming that the population density throughout the remaining prime habitat is similar to the density determined at Atitlán volcano.
Results
Vegetation cover and structure at Atitlán volcano
The Atitlán volcano is one of the major forest patches in the Guatemalan Pacific slope mountains, although much of the forest below 1,600 m has been replaced by coffee plantations. All coffee plantations within the study area were shaded monocultures according to a classification by Perfecto et al. (Reference Perfecto, Vandermeer, Mas and Soto Pinto2005), with a shade cover of 10–30% and 1–5 species of shade trees. Within the study area we discriminated between intensive and non-intensive coffee plantations, classified by differences in shade coverage and density of coffee shrubs. Vegetation cover within the study area was as follows: broadleaf forest covered 1,411 ha (72% of the study area), intensive coffee plantations 337 ha (17%), non-intensive coffee plantations 105 ha (5%), other intensive plantations (ponytail Beaucarnea recurvata, tea Camellia sinensis and others; complete nomenclature of all plant names is indicated in Table S1) 59 ha (3%), scrub 12 ha (1%), and orchard-like settlements 14 ha (1%) (Figure S1).
Foliage-height profiles show that broadleaf forest along transects was up to 40 m high, with a dense understorey of herbs and shrubs (including palms) and a dense canopy (Figure 2a,b). Vines were included in the class of shrubs, which were more abundant in forest at 900–1,300 m than in high-elevation forest at 1,400–1,900 m. Coverage of Ficus aurea trees in the canopy was higher in high-elevation forest.
Foliage height profiles of a) broadleaf forest at 1,400–1,900 m, b) broadleaf forest at 900–1,300 m, c) non-intensive coffee plantation, and d) intensive coffee plantation. n1 indicates the number of transects and n2 the number of points where vegetation cover was measured. Percent cover represents the mean percentage of plant touches on a pole in the indicated height classes.
Coffee plantations had a shrubby understorey with an open canopy of shade trees (Figure 2c,d). Non-intensive coffee plantations had a low density of coffee shrubs in the understorey, where they compete with native shrubs and herbs. Herbs were absent in intensive coffee plantation because of intensive use of herbicides. The canopy of shade trees in non-intensive plantations was dominated by Inga spp. trees (mainly Inga micheliana, but also I. spuria, I. rodrigueziana, and I. laurina) with a dense cover. Inga trees in intensive plantation were heavily pruned and these plantations also had non-native species of shade trees (Acrocarpus fraxinifolius, Eucalyptus sp.).
Habitat use, density, and abundance of Azure-rumped Tanagers at Atitlán volcano
Azure-rumped Tanagers were recorded on the south-eastern slope of Atitlán volcano from 860 to 1,900 m in primary humid broadleaf forest, in coffee plantation, and in edge habitat. A total of 134 observations of Azure-rumped Tanager involving 282 birds were recorded during transect counts between 19 June and 8 August 2008. In the forest interior (> 30 m from habitat edge) 92 observations (202 birds) were recorded; in forest edge habitat (< 30 m to forest edge) 21 observations (38 birds); in the interior of non-intensive coffee plantations three observations (six birds); in non-intensive coffee plantation edge habitat 11 observations (21 birds); and in intensive coffee plantation edge habitat 7 observation (15 birds). No tanagers were recorded in the interior of intensive coffee plantations. Because the sample size of tanagers in the interior of coffee plantations is too small to calculate a detection probability in this habitat, we compare encounter rates between habitats. For that we used only data of tanagers observed in the interior (> 30 m from the edge) of the sampled habitat, and truncated data at 30 m in order to lower the effect of higher detection probabilities in coffee plantations than in forests. Mean number of individuals recorded per 100 m of transect with a strip width of 60 m (Table 1) was significantly higher in forest (both types of forest pooled) than in coffee plantations (both types of coffee plantations pooled) (randomisation test, differences in means 0.303, P < 0.05, based on 1,000 permutations). Mean number of tanagers was also significantly higher in high-elevation forest (tanager’s prime habitat) versus non-intensive coffee plantation (randomisation test, differences in means 0.507, P < 0.05, based on 792 permutations).
Although tanagers were recorded as low as 860 m above sea level, mean encounter rate in forests 1,400–1,900 m (0.53 ± 0.59 birds/100 m) was significantly higher than in forests at 900–1,300 m (0.02 ± 0.03 birds/100 m) (permutation test, differences in means 0.517, P < 0.05, n = 972 permutations), indicating that high-elevation forest is the tanager’s main habitat. Sample size in forests at 1,400–1,900 m was sufficient to model detection probability. Along seven transect lines in forest at 1,400–1,900 m with a total sampling effort of 49.2 km, 89 detections of Azure-rumped Tanager were recorded. Data from distance sampling procedure fitted best a hazard-rate distribution, using Akaike’s Information Criterion (AIC) for model selection and best goodness-of-fit tested with chi-square test for each interval. A good model fit was achieved with data truncation at 65 m and classification in five distance intervals (Figure 3). Mean flock size was 2.17 individuals (SE 0.07; 95% confidence interval: 2.02–2.33). Estimated density of Azure-rumped Tanagers in forest at 1,400–1,900 m was 26 flocks km−2 (95% confidence interval: 15–43 flocks km−2) or 56 birds km−2 (95% confidence interval: 33–93 birds km−2) from June to August 2009. Thus, population size of Azure-rumped Tanager in the approximately 5 km2 of forest between 1,400 and 1,900 m within the study area at Atitlán volcano is estimated to be 165–465 individuals.
Histogram of the distance data for Azure-rumped Tanagers detected in forest between 1,400 and 1,900 m, and detection function obtained by fitting the hazard-rate model (software Distance 5.0 Release 2) to the data which were truncated at 65 m, and grouped into five distance classes.
The largest flock seen during transect counts consisted of five birds. During occasional observations of Azure-rumped Tanager, birds were seen usually in pairs or small groups of 3–4 individuals during the breeding season. After fledging, larger groups were seen. A flock of seven birds was seen on 3 September 2009 (GL pers. obs.), among them were two young birds. At least 14 tanagers were observed on 13 August 2008 (GL pers. obs.) moving in the canopy of a Ficus aurea tree in primary forest, among them were at least five young birds. We never observed Azure-rumped Tanagers in true mixed foraging flocks. On occasions when we recorded them together with other species, for instance with White-winged Tanager Piranga leucoptera on 11 July 2008 (SA and GL pers. obs.) in Los Tarrales, and with Common Bush-tanager Chlorospingus ophthalmicus on 15 August 2009 in the forest reserve of San Rafael Pie de la Cuesta (KE and CA, pers. obs.), they were apparently together in the same fruit tree by coincidence, and did not move on together.
The height of observed Azure-rumped Tanagers in vegetation was recorded during transect counts for 90 detections in forest and edge, and for 16 detections in coffee plantation and edge. The height above ground ranged in forest from 8–36 m (24.3 ± 6.1, n = 90) and in coffee plantation from 8–30 m (18.9 ± 7.2, n = 16). Tanagers stayed mainly in the upper vegetation strata, mean relative height in vegetation was 83% ± 17 (n = 90) in forests and 83% ± 20 (n = 16) in coffee plantations. In coffee plantations tanagers were recorded mainly in solitary large trees (Ficus aurea and F. costaricana) or in the canopy of shade trees (Inga spp.). Occasionally tanagers were also seen very low in the understorey, foraging on fruiting shrubs in the forest or collecting nest material from coffee shrubs as low as 1 m above ground.
During transect counts a total of 41 observations of feeding tanagers were recorded, of which 35 were feeding on fruit of the strangler fig Ficus aurea (Moraceae), one on Ficus costaricana, two on Rhamnus discolor (Rhamnaceae), one on Trema micrantha var. strigillosa (Ulmaceae), one in mistletoe Phoradendron heydeanum (Verbanaceae), and one in Perrottetia longistylis (Celastraceae). During occasional observations throughout the tanager’s range in Guatemala, the species was seen feeding also on fruit of Miconia sp. (Melastomataceae) (KE and CA pers. obs.), on Cecropia obtusifolia (Cecropiaceae) (KE, JLL, EB pers. obs.), on Chomelia brachypoda (Rubiaceae), on Cestrum sp. (Solanaceae) (KE and CA pers. obs.), and on Urera sp. (Urticaceae) (J. L. Yuxón pers. comm.) (see Table S1 for complete nomenclature of all plant names). Tanagers were also seen foraging among inflorescences of Meriana macrophylla (Melastomataceae), probably on invertebrates (KE and CA pers. obs.).
The number of observations of feeding tanagers in Ficus aurea suggests a high importance of that food source. F. aurea is semi-deciduous and the trees are leafless for some weeks in the dry season and identifiable on orthoimages. The density of F. aurea trees within a strip width of 120 m ranged from 0–2 trees ha−1 along all forest transects (Table S2). The slope (0.626 ± 0.149 SE) of the Geometric Mean Regression line comparing density of F. aurea trees from Table S2 versus number of Azure-rumped Tanagers per 100 m of transect from Table 1 is significantly different from 0 (Student’s t-test, t = 4.21, df = 10, P < 0.005), indicating a positive correlation between the abundance of Azure-rumped Tanager and density of F. aurea trees (Figure 4).
Encounter rate of Azure-rumped Tanagers in relation to the density of Ficus aurea trees. The slope of the geometric mean regression line is significantly different from 0 (Student’s t-test, t = 4.21, df = 10, P < 0.005).
Distribution in Guatemala
The Azure-rumped Tanager has been reported in Guatemala from nine topographic units and a total of 16 sites from 1987 to 2009, located in the departments of San Marcos, Quetzaltenango, Sololá, Suchitepéquez, and Chimaltenango (Table S3, Figure 5a). The easternmost record is from Unión Victoria (14º 34’N 91º 03’W), former Finca El Salvador, municipality Pochuta, 5 km northeast of Pochuta. The potential area of distribution of Azure-rumped Tanager in Guatemala can be limited by the upper and lower altitudinal limits of all records, as well as the easternmost record and the Mexican border. This area has a size of 1,170 km2.
Distribution and habitat availability of Azure-rumped Tanager in Guatemala. a) Locations of all records in Guatemala, altitudinal zones, and department limits. b) Landcover in the potential area of distribution, limited by the Mexican border in the West and the easternmost record, and by contour lines of 900 m and 1,900 m, representing approximate altitudes of the lowest and highest records.
From information on habitat use in our study area at Atitlán volcano we derived the coverage of adequate habitat within the potential area of distribution. Based on a nationwide land cover mapping on the scale 1:50,000 (MAGA 2006), broadleaf forest covers approximately 250 km2 or 21% of the entire potential area of distribution of Azure-rumped Tanager in Guatemala (Figure 5b). Most of the potential area of distribution is covered by coffee plantations (approximately 800 km2 or 68%). Largest forest patches are located south of San Marcos (northern part of the municipalities Tumbador, Nuevo Progreso, and La Reforma), on the southern slope of Lacandón volcano, on the south-eastern slope of Chicabal volcano, on the southern slopes of Santa María, Santo Tomás, and Atitlán volcanoes, and on the southern part of the municipality Santiago Atitlán, and in the valley of the Río Madre Vieja southeast of San Lucas Tolimán (Figure 5b). Assuming that the density of Azure-rumped Tanagers in all broadleaf forests is similar to the value determined in the study area at Atitlán volcano, total population of Azure-rumped Tanagers in Guatemala is estimated to be 8,250–23,250 birds.
Discussion
Distribution and conservation status
The type specimen of Azure-rumped Tanager was collected in Costa Cuca (Cabanis Reference Cabanis1866, Sclater Reference Sclater1868; type specimen ZMB 18921 reviewed). In the 19th century, Costa Cuca was not a town or village, but a region of about 1,300 km2, which is today the southern part of the department of Quetzaltenango (Gallini Reference Gallini2009). The type locality has been erroneously attributed to the village Flores Costa Cuca (elevation 500 m) in several publications (Heath and Long Reference Heath and Long1991, Collar et al. Reference Collar, Gonzaga, Krabbe, Madroño Nieto, Naranjo, Parker and Wege1992). All recent records are from the altitudinal range of 860–1,900 m, which is a narrow belt about 10 km wide. Therefore the type specimen had most likely been collected in the upper part of the former Costa Cuca area. After the type specimen of Azure-rumped Tanager was collected in the 19th century, only recently have additional localities been reported in the modern department of Quetzaltenango (Heath and Long Reference Heath and Long1991, Cooper Reference Cooper2003, Eisermann and Avendaño Reference Eisermann, Avendaño and Cano2006, Reference Eisermann and Avendaño2007a). This study documents an expansion of the altitudinal (860–1,900 m) and longitudinal range (east of the valley of the Nicán river). This species may be expected even further east in Guatemala, such as on the southern slope of Fuego and Agua volcanoes in the department of Sacatepéquez, which is supported by an anecdotal record at Pacaya volcano on 12 April 2010 (Dan Poalillo pers. comm.).
The Azure-rumped Tanager has been reported previously from humid broadleaf forest and also from coffee plantations (Dietsch Reference Dietsch2000, BirdLife International 2008). Our comparison of encounter rates in forest and coffee plantation shows that the species is more abundant in humid broadleaved forest. In addition, most of our records from coffee plantations were within a distance of 30 m from forest edge. Thus, the Azure-rumped Tanager is primarily a forest bird, which also uses forest edge and adjacent open habitat as do many other tanagers (Isler and Isler Reference Isler and Isler1987).
Most of the prime habitat of Azure-rumped Tanager in Guatemala has been converted to coffee plantations and only 250 km2 of potential habitat remain (Figure 5b). The tanager is fairly common and occurs with a density of 33–93 birds km-2 on the southern slope of Atitlán volcano. It is also common at other sites such as in the departments of San Marcos (Cooper Reference Cooper2003) and Quetzaltenango (Table S3). However, the species is vulnerable to population declines, because most of the suitable habitat is located in a zone especially favourable for agriculture, except for the steepest canyons and mountain slopes. According to recent information on protected areas (CONAP 2009), within the potential area of distribution, only 370 km2 (32%) occur within protected areas, and 138 km2 (55%) of the remaining broadleaf forest in this area are legally protected (Table S4), but none of the reserves is a strictly protected area in IUCN categories I or II (IUCN 1994). Efficient conservation of habitat cannot be guaranteed in the short term in any of the reserves. Although many of the privately protected areas currently are among the best protected preserves in Guatemala, driven by a strong personal interest of owners, national law of protected areas (Decreto 4–89, Congreso de la República de Guatemala) and its rules (Acuerdo Gubernativo 759–90) do not require a long-term commitment of owners for conservation, nor do they require continued conservation from future land owners. Guatemala’s state reserves lack adequate management and law enforcement, which makes them vulnerable to illegal invasions followed by an advancing agricultural border, illegal extraction of forest resources and hunting (ParksWatch 2002, 2005, Godoy Reference Godoy, Chape, Spalding and Jenkins2008). The remaining suitable habitat of Azure-rumped Tanager in Chiapas was estimated to be 1,120 km2 in the 1980s (Heath and Long Reference Heath and Long1991), but rapid human population growth in Chiapas (1.6% annual growth rate from 2000 to 2005; INEGI 2008) combined with an ongoing human pressure on the Sierra Madre de Chiapas (Richter Reference Richter2000) suggest recent considerable loss of habitat there. The total extent of occurrence in both countries (minimum convex polygon encompassing all sites of occurrence) is less than 2,500 km2. We expect a further decline of the extent of suitable habitat because of continued growth of human population, which in Guatemala is expected to double between 2010 and 2050 to a total of 27.9 million (CEPAL 2007). The maintenance of the IUCN Red List status of ‘Endangered’ is justified according to the IUCN (2001) Red List criteria EN B1a+b(ii,iii,v).
Conservation implications and population monitoring
The survival of the Azure-rumped Tanager depends on conservation of natural habitat and habitat restoration in and outside of protected areas. The necessity of conservation efforts in human-used landscape in Guatemala has been recognised during the identification of Important Bird Areas (Eisermann and Avendaño Reference Eisermann and Avendaño2007b, Reference Eisermann and Avendaño2009a,Reference Eisermann, C., Devenish, Diaz Fernández, Clay, Davidson and Yépez Zabalab), and it has also been recognised on a regional level in Mesoamerica (Harvey et al. Reference Harvey, Komar, Chazdon, Ferguson, Finegan, Griffith, Martínez-Ramos, Morales, Nigh, Soto-Pinto, van Breugel and Wishnie2008, Chazdon et al. Reference Chazdon, Harvey, Komar, Griffith, Ferguson, Martínez-Ramos, Morales, Nigh, Soto-Pinto, van Breugel and Philpott2009, Gardner et al. Reference Gardner, Barlow, Chazdon, Ewers, Harvey, Peres and Sodhi2009). Because most of the tanager’s range in Guatemala is covered with coffee plantations, these areas need to be considered in a conservation strategy. Coffee is being cultivated under a variety of management systems, ranging from intensive unshaded monoculture with an intensive use of agrochemicals, to rustic coffee plantations, which are cultivated under the shade of the natural forest. Between these extremes exist a variety of plantation types differentiated by species richness of shade trees and shade cover (Perfecto et al. Reference Perfecto, Vandermeer, Mas and Soto Pinto2005, López-Gómez et al. Reference López-Gómez, Williams-Linera and Manson2007, Hernández-Martínez et al. Reference Hernández-Martínez, Manson and Contreras Hernández2009). The conservation value of coffee plantations has been discussed since the 1990s (Komar Reference Komar2006), and plantations with a diverse shade have been identified as the most valuable for biodiversity (Perfecto et al. Reference Perfecto, Rice, Greenberg and Van der Voorts1996, Perfecto et al. Reference Perfecto, Mas, Dietsch and Vandermeer2003, Philpott et al. Reference Philpott, Arendt, Armbrecht, Bichier, Dietsch, Gordon, Greenberg, Perfecto, Reynoso-Santos, Soto-Pinto, Tejeda-Cruz, Williams-Linera, Valenzuela and Zolotoff2008).
Most coffee plantations in the potential range of Azure-rumped Tanager in Guatemala are shaded monocultures. The use of forest edge and single trees within coffee plantations by foraging tanagers suggests that the area of suitable habitat could be augmented considerably by some shade alterations among plantations. The dense network of maintenance roads in coffee plantations could be used to establish lines of native trees, especially Ficus aurea, which would provide both food and nesting sites. In many plantations we saw Eucalyptus spp. trees planted along roads, so there would be little loss of cultivated area by replacing Eucalyptus or other non-native trees with Ficus aurea and other fruit trees. The creation of tree rows could even benefit the coffee plantations. Pollination experiments in Costa Rica showed that pollination of coffee blossoms is higher within a distance of 100 m from forest fragments and that forest-based pollinators increased coffee yield by 20% (Ricketts Reference Ricketts2004, Ricketts et al. Reference Ricketts, Daily, Ehrlich and Michener2004). Increased shade and plant species richness in coffee plantations increases bird abundance (Philpott et al. Reference Philpott, Arendt, Armbrecht, Bichier, Dietsch, Gordon, Greenberg, Perfecto, Reynoso-Santos, Soto-Pinto, Tejeda-Cruz, Williams-Linera, Valenzuela and Zolotoff2008). Bird exclosure experiments in Chiapas showed that avian predation on Lepidoptera larvae was significantly higher in plantations with high floristic diversity (Perfecto et al. Reference Perfecto, Vandermeer, López Bautista, Ibarra-Núñez, Greenberg, Bichier and Langridge2004). A similar test in Jamaican coffee plantations concluded that birds reduced significantly the infestation by coffee berry borers Hypothenemus hampei, Coleoptera: Scolytidae (Kellermann et al. Reference Kellermann, Johnson, Stercho and Hackett2008), which is also a common pest in Guatemalan coffee plantations. Johnson et al. (Reference Johnson, Levy, Kellermann and Robinson2009) pointed out that greater shade cover in a Jamaican coffee plantation increased predation on arthropods by birds, but it also increased fungal leaf damage on the coffee shrubs. Local environmental conditions vary widely, and the task is to find for each farm the balance between high coffee yield and satisfying support for biodiversity conservation. Incentives and certification programmes could play an important role in promoting these efforts. Habitat restoration measures should be accompanied by population monitoring of Azure-rumped Tanager. To continue monitor the distribution of the tanager we would appreciate receiving reports of observation sent by e-mail or submitted to eBird Guatemala at <http://www.ebird.org/content/guatemala/>.
Supplementary Material
The online supplementary materials for this article can be found at journals.cambridge.org/bci
Acknowledgements
We appreciate the logistical support by the people of Los Tarrales Reserve. SA and KE are thankful to Andy Burge and Mónica Barrios for providing a temporary home at El Vesubio/Los Tarrales. We thank Olga Hazard and Jaime Freire (Los Andes), as well as Mark Brissani and Arnoldo Villagrán (Reserva San Jerónimo Miramar) for permitting us to work in their private reserves. We appreciate the field assistance and hospitality provided by Jamie Staples, Carlos Mazariegos, Marcelino Orozco and Saúl Sandoval (San Rafael Pie de la Cuesta), Mario Castillo and Fernando García (Las Nubes Reserve), the late Federico J. Fahsen (Pachuj Reserve), Olga Hazard, Jaime Freire, Jesús Lucas Yuxón, Juan Calvo Cortéz (Los Andes Reserve), people of Loma Linda, Mario Aguilar (Patrocinio Reserve). We thank Juan José Castillo Mont, Mario Veliz, and David Mendieta for the identification of plant specimens. We thank Kenset Rosales and Rafael Ávila of Consejo Nacional de Areas Protegidas (CONAP) for providing updated information on Guatemala’s protected areas. We thank Sylke Frahnert and Pascal Eckhoff of the Museum für Naturkunde in Berlin, Germany, for their assistance during reviewing specimens in the ZMB. We appreciate the review of the manuscript by Joseph M. Wunderle and Phil Atkinson, and the revision of English usage by Peter Lowther through the Association of Field Ornithologists’ programme of editorial assistance. Optics for the Tropics donated binoculars. This study was financially supported by Cayaya Birding, Los Tarrales Reserve, and PROEVAL RAXMU Bird Monitoring Program.
