THE JOURNAL OF TROPICAL BIOLOGY AND CONSERVATION

BIOTROPICA 41(1): 66–73 2009

10.1111/j.1744-7429.2008.00458.x

Specialization on Guadua Bamboo Seeds by Three Bird Species in the Atlantic Forest of Argentina Juan I. Areta1,2,3,5 , Alejandro Bodrati2,3 , and Kristina Cockle3,4 1 CICyTTP-CONICET, 2 Grupo

Materi y Espana, ˜ 3105 Diamante, Entre R´ıos, Argentina

FALCO, Calle 117 Nro 1725 e/67 y 68, 1900 La Plata, Buenos Aires, Argentina

3 Proyecto

Selva de Pino Parana, ´ Fundacion ´ de Historia Natural Felix ´ de Azara, Departamento de Ciencias Naturales y Antropolog´ıa, CEBBAD–Universidad Maimonides. Valent´ın Virasoro 732, C1405BDB Buenos Aires, Argentina ´

4 Centre

for Applied Conservation Research, Faculty of Forestry, University of British Columbia, 2424 Main Mall, Vancouver, BC, V6T 1Z4, Canada

ABSTRACT Most bamboos are semelparous. Their synchronous masting events occur on a cycle of 3–120 yr and represent an extremely pulsed resource for granivorous birds. Although many bird species feed occasionally on bamboo seeds, there are constraints to specializing on such a fluctuating resource and few bird species are known to specialize on bamboo seeds. Three of these bird species are endemic to the Atlantic forest of South America: the purple-winged ground-dove Claravis godefrida, buff-fronted seedeater Sporophila frontalis, and Temminck’s seedeater Sporophila falcirostris. All three species are irregularly recorded in the province of Misiones, Argentina. We compared the temporal and spatial patterns of records of these birds in Misiones to masting events of the five common bamboos: yatev´o Guadua trinii, takuaruz´u Guadua chacoensis, takuap´ı Merostachys claussenii, pitinga Chusquea tenella, and takuaremb´o Chusquea ramosissima. All bird records coincided with times and places where Guadua bamboos (G. trinii and G. chacoensis) were known or estimated to have seeds. None of the bird species occurred during masting events of Merostachys or Chusquea, unless Guadua was also masting. We discuss relevant ecological and morphological features of the birds that might bear on their association with Guadua bamboos and that might be key to their conservation. Abstract in Spanish is available at http://www.blackwell-synergy.com/loi/btp. Key words: Claravis; granivore; mast event; specialist; South America; Sporophila.

WORLDWIDE,

MANY MAMMALS, BIRDS, AND INSECTS EXPLOIT THE

and food resources provided by the more than 1500 described species of bamboos (Janzen 1976). Because most species of bamboo are semelparous, with mass flowering and die-off events every 3–120 yr (usually 15–60 yr) depending on the species (McClure 1966, Janzen 1976), bamboo specialist animals experience extreme variability in their habitat and food sources. Avian bamboo specialists show two very different strategies. Most are insectivores that inhabit woody bamboos. For example, the bamboo woodpecker Gecinulus viridis of SE Asia gleans and probes for ants in live and dead bamboo, and nests in cavities in bamboo (Short 1973). These species may be relatively sedentary, but they must experience population declines, switch to alternative habitat, or travel widely during bamboo die-off events. They benefit from long-term habitat stability but must face short-term reductions in habitat that do not affect all generations of birds. Granivorous birds may also be bamboo specialists. Although many bird species have been reported foraging on bamboo seeds (Janzen 1976, Lebbin 2006) and flowers (Franklin 2005), few are considered to specialize on this resource. Bamboo seed specialists must rely on an ephemeral source of food available only after large intervals of time at most locations. During their lifetime, most individuals face a lack of their FLUCTUATING HABITAT

Received 20 March 2008; Revision accepted 19 July 2008. author; e-mail: [email protected]

5 Corresponding

66

preferred food supply, at which time they must wander to survive (Neudorf & Blanchfield 1994). For example, the pied mannikin Lonchura fringilloides of East Africa appears to rely, at least in part of its range, on seeds of the bindura bamboo Oxytenanthera abyssinica, a medium-sized, upright bamboo that produces seeds approximately every 30 yr (Jackson 1972, Restall 1995). At a given location, pied mannikins may feed on bamboo during masting events that last up to 7 yr, moving to different locations when few bamboo seeds remain. Thus, the availability of seeds of regularly masting species is predictable, but only over a period longer than the life span of most granivorous birds (i.e., seed availability is ecologically predictable but individually unpredictable). Their nomadic movements make bamboo seed specialists particularly difficult to study, and it is not surprising that most bamboo-bird studies focus on insectivores, whereas the ecology and habitat preferences of bamboo granivores remain poorly known. Bamboo specialization is a prominent feature of the Neotropical avifauna. Avian bamboo specialists occur from the southern temperate forests of Patagonia to the northern tropical rain forests (Lentino & Restall 2003, Reid et al. 2004). More than 90 species of Neotropical birds in seven families were classified by Stotz et al. (1996) as bamboo-associated. An overlooked feature of bamboo specialization is that bamboo specialist birds may depend on a single genus or species of bamboo, at least in some parts of their range. For example, at least 22 diurnal insectivores in the Peruvian

 C 2008 The Author(s) C 2008 by The Association for Tropical Biology and Conservation Journal compilation 

Bird Specialization on Guadua Bamboo Seeds

Amazon (Kratter 1997, Lebbin 2007) and the insectivorous whitebearded antshrike Biatas nigropectus in Argentina (Bodrati & Cockle 2006) specialize on Guadua bamboo, avoiding sympatric bamboo genera. Since the ecology and distribution of bamboo specialist birds may be closely linked to specific bamboos, it is important to understand the relationships between bird species and bamboo species (Bodrati & Areta 2006). Specialization on bamboo seeds seems most likely to arise in bird species that inhabit regions with high species richness of bamboos, where asynchronous flowering by different bamboo species would provide a constant supply of seeds. One such region is the Atlantic forest of South America (Judziewicz et al. 1999). Richness of woody bamboos varies considerably across the Atlantic forest, from more than 30 species in parts of Brazil (Judziewicz et al. 1999) to only ca eight species in Misiones, Argentina (Nicora & R´ugolo de Agrasar 1987, Sulekic et al. 1999). In Misiones, there are five common species of bamboo and six species of granivorous birds considered by Stotz et al. (1996) to be associated with bamboo. Among the bamboos, takuaremb´o Chusquea ramosissima Lindman and pitinga C. tenella Nees have small seeds and form dense tangles in the understory, while takuap´ı Merostachys claussenii Munro, yatev´o Guadua trinii (Nees) Ruprecht, and takuaruz´u G. chacoensis (Rojas Acosta) Londo˜no & P. Peterson have larger seeds and are tall and upright. Among the six granivorous bird species mentioned by Stotz et al. (1996), chestnut-headed tanager Pyrrhocoma ruficeps (Strickland) 1844, uniform finch Haplospiza unicolor (Cabanis) 1851, and blackish-blue seedeater Amaurospiza moesta (Hartlaub) 1853 are recorded regularly and appear to forage on a wide variety of seeds and insects (J. I. Areta & A. Bodrati, pers. obs.). In contrast, purple-winged ground-dove Claravis godefrida (Temminck) 1811, buff-fronted seedeater Sporophila frontalis (Verreaux) 1869, and Temminck’s Seedeater Sporophila falcirostris (Temminck) 1820 are poorly known, are probably more specialized on bamboo seeds, and exhibit marked temporal discontinuities that could be a result of the masting cycles of bamboos (Collar et al. 1992). All three are listed as threatened by BirdLife International (2007): C. godefrida is critically endangered, S. frontalis and S. falcirostris are vulnerable. Here, we use historic records and our own field data to test the strength of the relationships between records of C. godefrida, S. frontalis, and S. falcirostris and masting events of each bamboo genus. We also discuss features of the natural history of these birds that bear on their association with bamboo and might have conservation implications.

METHODS BAMBOO DATA.—To determine the distribution and timing of flowering events of the five widespread woody bamboo species present in Misiones, we sought reports in the literature, examined seedand flower-bearing specimens in the herbaria of the Museo de La Plata (LP) and the Instituto Darwinion (SI), interviewed long-time residents of Misiones, and contributed our own data from 5 yr in Misiones (2003–2008).

67

BIRD DATA.—To determine the distribution and timing of records of C. godefrida, S. frontalis, and S. falcirostris in Misiones, we sought specimens at the Museo Argentino de Ciencias Naturales (MACN, Buenos Aires), published data, and personal communications from third parties. We searched for these species throughout the province of Misiones during 450 d between 2003 and 2008, including flowering stands of G. chacoensis, M. claussenii, and C. ramosissima. For each record, we obtained (whenever possible): locality, habitat, date, number of birds, type of evidence, author, and source. DATA ANALYSIS.—To compare the timing and location of records of the three bird species to the timing and location of masting events of bamboos, we made a table and map that included all bird records and all known and estimated masting events over the 52-yr period from 1957 (first bird record) to 2008. To determine whether years with bird records coincided with years of bamboo masting more often than expected by chance, we conducted randomization tests with 2000 iterations. For each bird species, we generated 2000 data sets by sampling randomly from our original data (with replacement). For each random dataset we determined the number of years in which the presence of the bird species coincided with masting of any bamboo. We calculated the probability of finding a coincidence as extreme as ours if the presence of the bird species was random with respect to masting of bamboo. We repeated the randomization test for each combination of bird species and bamboo genus, to assess the strength of the relationship between the presence of each bird species and masting events of each bamboo genus. Since we had only rough estimates of bamboo timing for historic masting events, and some plants would have flowered early or late, for statistical calculation we considered each bamboo genus to be in flower from 1 yr before until 1 yr after main masting events. All statistical analyses were conducted using R 2.5.1 (R Development Core Team 2007).

RESULTS DISTRIBUTION AND MASTING EVENTS OF BAMBOOS IN MISIONES.— Guadua trinii is widespread throughout Misiones province, whereas G. chacoensis is restricted to the coast of the Iguaz´u and Paran´a rivers (Figs. 1 and 2). Guadua trinii favors well drained and rocky soils with moderate slopes, and G. chacoensis is found on sandy soils along watercourses (M. Srur pers. comm.). Merostachys claussenii and C. ramosissima, are found throughout the province, but not always at the same sites. Chusquea tenella is found mostly in the steeper terrain of eastern Misiones (Bertoni 1918; Parodi 1936, 1941). Guadua trinii masts synchronously within a given locality, and its intermasting period has been firmly established at 30 yr in the southern part of its distribution (Dutra 1938, Parodi 1955, Klimaitis 2004). However, like other bamboos (Franklin 2004), it seems to show a wave of flowering, with masting recorded in 1920–27 in Buenos Aires (Parodi 1955) but not until 1934 in Rio Grande do Sul (Dutra 1938), and most recently in 1982 in Buenos Aires but not until July 1988 in the Iguaz´u area (northern Misiones;

68

Areta, Bodrati, and Cockle

FIGURE 1.

Masting events of five bamboo species and their relationship to the presence of three bamboo seed specialist birds from 1957 to 2008, with documented

bamboo masting events (dark gray), inferred bamboo masting events (light gray), and records of birds (black) at localities in the Atlantic forest of Misiones (Argentina): 1—Iguaz´u area; 2—Arroyo Urugua-´ı area; 3—Parque Provincial Esmeralda; 4—Corpus Christi.

Appendix S1). Based on these data, a G. trinii masting event can be inferred to have occurred beginning roughly in 1958 in northern Misiones, and probably earlier in southern Misiones. The flowering cycle of G. chacoensis is not well known (Londo˜no & Peterson 1992, Young & Judd 1992). The intermast period might be as short as 27 yr (Bertoni 1917) or as long as 35 yr (Young & Judd 1992). Concordantly, there is evidence of seed production in northern Argentina in 1943–1944, 1973–1978, and 2005–2008 (Appendix S1), suggesting an intermast period of ca 30 yr. The most recent flowering in Misiones began in southern Misiones around 2004, with all culms dead by February 2007 (J. I. Areta, pers. obs.; N. Fari˜na pers. comm.), while in northern Misiones (Iguaz´u area), flowering began in 2005, and by January 2008 seed production had still not reached its peak (Appendix S1). Not all culms in a clump flower simultaneously, and individual plants might flower continuously for around 3 yr (Londo˜no & Peterson 1992). Merostachys claussenii is currently considered a synonym of M. burchellii (Dutra 1938, Sendulsky 1995), but taxonomy and biology of this genus remain little studied (Z. R´ugolo, pers. comm.). Merostachys claussenii flowered in Misiones in 2004–2006, and evidence suggests that masting occurs every 30 yr, lasting ca 3 yr at a given location (Appendix S1). Chusquea bamboos often have an intense masting period of ca 2 yr (Seifriz 1950, Pearson et al. 1994, Widmer 1998). Specimens and records of flowering events of C. tenella suggest synchronous flowering over a wide geographic area, with an intermast period of 15–16 yr (Appendix S1). The most recent masting event in Misiones was reported in 1993, and we therefore expect masting to occur again in 2008–2010. Unlike the other four bamboo species we studied, C. ramosissima does not show a regular pattern of masting events over a wide area (Appendix S1), however, Dutra (1938) found that two consecutive flowering events occurred in 23 yr in Rio Grande do Sul, Brazil. In at least two cases, specimens in vegetative state and specimens with flowers or seeds were collected in the same years at adjacent localities (e.g., in 1910 at Pto. Iguaz´u and 1945 at Cataratas del Iguaz´u). The recent records of flowering C. ramosissima in Misiones, Argentina, in 1995, 1997, and 2005 are most likely explained by asynchrony in the timing of flowering events of cohorts within a small geographic area. If this is the case, then C. ramosissima might not be a strict masting species. Although flowering waves have been reported for some Chusquea species (Widmer 1998, Stern et al.

1999), the pattern of flowering of C. ramosissima does not fit this kind of flowering asynchrony, and we suspect that C. ramosissima might flower sporadically at irregular intervals (see Campbell 1985). A study of C. ramosissima demography in northern Misiones agrees with our observations (L. Montti, pers. comm.). Without more precise data and based on the data presented in Appendix S1, we estimate 5 yr per Guadua masting period, 3 yr per Merostachys masting period, and 2 yr per Chusquea masting period (Fig. 1). Although C. ramosissima must have flowered more times from 1957 to 2008, we include the years when flowering is known to have occurred (Figs. 1 and 2). GEOGRAPHIC AND TEMPORAL DISTRIBUTION OF BIRD RECORDS IN MISIONES.—We obtained a total of 27 bird records in four localities: six records of C. godefrida, five of S. frontalis, and 16 of S. falcirostris (Figs. 1 and 2; Appendix S2; Bertoni 1913; Navas & B´o 1986, 1987; Savigny in press, Wege & Long 1995). The bird records are clumped in four distinct time intervals (Figs. 1 and 2; Appendix S2). RELATIONSHIP BETWEEN BAMBOO MASTING AND PRESENCE OF BIRDS.—The bird records are obviously clumped, occurring at times and places where either G. trinii or G. chacoensis had seeds or were estimated to have seeds (Figs. 1 and 2). Nine records mentioned masting of Guadua bamboos in the vicinity; the other 18 records made little or no mention of habitat (Appendix S2). None of the records mentioned Chusquea or Merostachys, either with or without seeds. For all three bird species, the coincidence we observed between timing of bird records and timing of Guadua flowering events would have been unlikely if bird records were distributed randomly through time (Table 1). Although we searched extensively for the three bird species in masting stands of G. chacoensis, M. claussenii, and C. ramosissima, and nonmasting stands of G. trinii, we only found them in masting stands of G. chacoensis. For C. godefrida and S. frontalis the sample size is small and the results are only marginally significant, but for all three species of birds there is clearly a coincidence with timing of Guadua masting, and not with the masting of other bamboo genera.

DISCUSSION HABITAT USE.—The temporal and spatial patterns of the records of three avian bamboo specialists in Misiones (C. godefrida, S. frontalis

Bird Specialization on Guadua Bamboo Seeds

69

TABLE 1. Results of randomization tests for relationship between years with bird records and years within 1 yr of bamboo flowering. P-values indicate the probability of achieving a coincidence in bird records and bamboo masting as extreme as ours if the bird records were randomly distributed in time with respect to bamboo masting events. Data are from Appendices S1 and S2. Also see Figures 1 and 2. Bamboo genus

S. falcirostris

Any bamboo Guadua

0.134 0.011

Merostachys Chusquea

0.698 0.443

Any bamboo Guadua Merostachys

0.298 0.104 0.247

Chusquea Any bamboo Guadua

0.264 0.243 0.080

Merostachys Chusquea

0.342 0.910

S. frontalis

C. godefrida

FIGURE 2. Approximate distributional range of five bamboo species in Misiones: Guadua trinii, Merostachys claussenii and Chusquea ramosissima (gray); Guadua angustifolia (diagonal hatching); Chusquea tenella (horizontal hatching), and localities corresponding with bird records as numbered in Figure 1. Real distributions are patchy for all species at the chosen scale. Boundaries were drawn based on our knowledge of the occurrence of large bamboo stands of each species. Bamboo data are from Appendix S1, bird data are from Appendix S2.

and S. falcirostris) closely fit the patterns of masting events of two widespread woody Guadua bamboo species. The birds were present, from 1957 to 2008, only where and when Guadua bamboos had seeds. Sporophila frontalis was not found at all masting events, possibly because it was absent (which is expected if seeds of preferred bamboos were available elsewhere), or possibly because it was overlooked. The other two species were present at all four Guadua masting events since 1957. Although masting events of C. tenella and M. claussenii often partly overlapped the masting events of one or both species of Guadua, geographic locations of the bird records and observations of feeding and nesting in Guadua bamboos strongly suggest that it was the masting of Guaduas, rather than Merostachys or Chusquea, that allowed these birds to occur in Misiones. These birds may also rely on Guadua bamboos in parts of Brazil and Paraguay. In Minas Gerais, Brazil, Vasconcelos et al. (2005) noted that they did not record these three bamboo seed specialists during masting events of Parodiolyra micrantha and Chusquea attenuata, and they proposed that these species may have been driven

P

Bird species

to local extinction by habitat destruction, or, alternatively, may not feed on the bamboo species studied. Sporophila frontalis was present at a recent masting event of a species of Guadua in Teres´opolis (Rio de Janeiro, Brazil) but not during a masting event of one Merostachys and two Chusquea species in Minas Gerais (Brazil, Sep–Nov, 2007), where only one S. falcirostris and no C. godefrida were detected (M. F. Vasconcelos, pers. comm.). Although S. frontalis was present at a masting event of Chusquea aff. meyeriana at Fazenda Intervales (S˜ao Paulo, Brazil; Olmos 1996), it did not feed on the seeds of this bamboo and Olmos (1996) concluded that S. frontalis must specialize on other bamboos, probably Merostachys or Guadua. In Paraguay, too, S. falcirostris went unrecorded during extensive fieldwork from 2000 to 2002 when no Guadua was known to be masting (A. Bodrati, pers. obs.), but M. Vel´azquez (pers. comm.) has compiled several records from 2004 to 2008, when G. chacoensis was masting. Although seeds of Merostachys and Chusquea do not appear to be important sources of food for these three bird species in Argentina, they may be important resources for these and other species in other parts of South America. Merostachys bamboo was flowering when Sick (1997) recorded great concentrations of S. frontalis in 1952 in Itatiaia (Rio de Janeiro, Brazil). Both S. frontalis and S. falcirostris were recently photographed at a masting event of a Chusquea subg. Rettbergia in 2007 at Ubatuba (S˜ao Paulo, Brazil; C. Rizzo, G. Trivelato and X. Londo˜no, pers. comm.), where both species were observed consuming the Chusquea seeds and many S. falcirostris males were singing from the tree tops (G. Trivelato, pers. comm.). Several nonexclusive hypotheses may explain why the Merostachys and Chusquea bamboos in Misiones are not a good source of food for the three bird species we studied. First, the

70

Areta, Bodrati, and Cockle

Chusqueas might not produce seeds in large enough quantities as to attract these birds. Chusquea ramosissima does not appear to have large masting events (Fig. 1; Appendix S1; L. Montti, pers. comm.). Furthermore, Olmos (1996) reports that to find husks of C. aff. meyeriana containing edible seeds, individuals of H. unicolor had to squeeze, in their bills, numerous empty husks, suggesting a low rate of return for the energy expended. Second, if the same or other species of bamboo produced seeds at the same time, outside Argentina, birds might not have arrived at masting events in Argentina. Finally, the seeds and spikelets of the Merostachys and Chusqueas we studied differ in size and morphology from those of the Guaduas, perhaps making them inappropriate food sources for these birds (see Food and feeding). In the long term, survival of bird species that specialize on bamboo seeds might depend on the existence of several species of bamboos that produce similar seeds at different times and over very wide geographic areas. At a local level, periods without bamboo seeds are reduced if flowering events of sympatric bamboo species occur in an alternate fashion, as do those of G. chacoensis and G. trinii in Misiones, thus providing birds with more feeding opportunities through time. In coastal Brazil, where these three bird species also exploit bamboo seeds (see Collar et al. 1992 and references therein), there is a much greater local diversity of bamboo (Judziewicz et al. 1999), and intervals between flowering events of different species are much reduced in comparison to Argentina. Thus, these bamboo seed specialists may be less nomadic in coastal Brazil compared to Argentina, and might rely on different bamboo species in different parts of their range. Some records of bamboo-specialist birds are to be expected away from bamboo, and such records should not be taken as supporting lack of specialization on bamboo-seeds (e.g., S´anchez et al. 2006). The birds recorded may be in transit to or searching for bamboo patches. For example, Lentino and Restall (2003) suggested that most sightings of Amaurospiza blue seedeaters away from bamboo could be of birds in movement from one patch of seed-bearing bamboo to another. ABUNDANCE.—When considered globally, S. frontalis is probably overall more abundant than S. falcirostris (Ridgely & Tudor 1989, Collar et al. 1992), but the few records in Misiones (none properly documented) suggest the opposite for Argentina. Claravis godefrida has always been considered scarce throughout its distribution, although it might have been more common in some parts of Brazil (Bertoni 1901, Collar et al. 1992). Both C. godefrida and S. frontalis have been reported to concentrate locally in flocks of 50–100 birds (Collar et al. 1992, Sick 1997), and this seems to occur to a lesser extent in S. falcirostris (Collar et al.1992; J. I. Areta & A. Bodrati, pers. obs.). No large concentrations of these species are known to occur in Argentina. Seed production usually increases over the first part of a masting event (Janzen 1976, Kelly 1994). If a bird species is a bambooseed specialist, its abundance should increase as a masting event progresses, with the arrival of more individuals and the production of chicks through successful breeding. In accordance with this prediction, between August 2007 and January 2008 we observed

an increasing abundance of G. chacoensis seeds and an increasing abundance of S. falcirostris. FOOD AND FEEDING.—The specialization on Guadua that we report may be related to differences in the size and morphology of seeds and spikelets between the five bamboo species we studied. The seeds of C. ramosissima and C. tenella are considerably smaller (7–9 mm) than the seeds of G. trinii, G. chacoensis, and M. claussenii (11– 14 mm; K. Cockle, J. I. Areta & A. Bodrati, pers. obs.). Chusquea bamboo seeds may be important food items of the peg-billed finch Acanthidops bairdii, slaty finch Haplospiza rustica, and H. unicolor, and were present where these birds were recorded breeding (S´anchez & Hern´andez 1990, S´anchez 1993, Olmos 1996, S´anchez 2005). These birds, which habitually feed on seeds of Chusquea bamboos (Stiles & Hespenheide 1972, Solano Ugalde & Arcos Torres 2006), have slender and conical bills. In contrast, S. frontalis, S. falcirostris and another bamboo specialist, the slate-colored seedeater S. schistacea, have large high bills, with the lower mandible larger than the upper mandible, and shaped as in some baleen whale jaws (e.g., bowhead whale). This is illustrated by a comparison between S. falcirostris, whose lower mandible measures 57 percent of total bill height, and typical seedeaters, whose lower mandibles measure 42–46 percent of total bill height (K. Cockle, pers. obs.). These peculiarly shaped bills might be especially suited to feeding on the seeds of some bamboos (e.g., Guadua spp.), but might be less effective for, or even preclude, feeding on other seeds (e.g., Chusquea spp.). The exact relationship between bamboo seed specialization and bill shape remains poorly understood; however, it might be the single most important feature that could explain habitat specificity of both S. frontalis and S. falcirostris. Conceivably, bamboo-seed specialists should have ways of coping with the lack of bamboo seeds to which they might be repeatedly exposed through their lives. Feeding on alternative seed sources has been reported for S. falcirostris (Hypolitrum sp. seeds in forest clearings, Sick 1997; unidentified tree seeds; Collar et al. 1992) and C. godefrida (papaya and other large fruits, Bertoni 1901, Collar et al. 1992 and references therein). Feeding on insects may also be common. The bristles around their bills suggest that both S. falcirostris and S. frontalis rely heavily on unusually large insects for food at least at some times (V. Remsen, pers. comm.). We have observed S. falcirostris capturing insect prey by sallying and by pursuing them with sustained flights, as mentioned by Sick (1997) for S. schistacea, probably the closest relative of S. falcirostris (Parker 1982, Lijtmaer et al. 2004). Studies on foraging tactics and their relationship to seed, inflorescence and branch complement morphology should provide further insight into the feeding opportunities and constraints posed by bamboo seeds. For example, the nucoid caryopsis and apsidate branch pattern of Merostachys bamboos differs drastically from the traditional caryopsis and simple branch pattern of Guadua bamboos (McClure 1966, Judziewicz et al. 1999). Thus, very different foraging tactics might be needed to feed on these genera or on different species within these genera (e.g., each M. claussenii seed is supported by a long, thin stem, which may be too weak to support the weight of a S. falcirostris or S. frontalis). The apparent

Bird Specialization on Guadua Bamboo Seeds

geographical variation in use of Chusquea might be explained by its enormous morphological diversity (Judziewicz & Clark 1991, Judziewicz et al. 1999). BREEDING.—Although bamboo seed specialists might be able to survive for fairly long periods without bamboo seeds, masting events would provide chances for massive reproduction events that could keep the population numbers viable during protracted periods with few or no breeding opportunities. Breeding has been recorded for C. godefrida during simultaneous masting of G. trinii and C. ramosissima in Rio de Janeiro (Chebez 1986), and for its closest extant relative, C. mondetoura, during masting of Chusquea sp. bamboo (Greeney et al. 2007). Sporophila falcirostris is considered to breed during bamboo masting events in Brazil (Sick 1997), and the only two breeding records from Argentina occurred at masting events of G. trinii (Castelino 1990) and G. chacoensis (Appendix S2). Sporophila frontalis was reported nesting in Brazil (U. St¨auble, pers. comm.; Appendix S2) when G. chacoensis was known to be masting nearby. Sporophila schistacea also usually breeds during bamboo masting events (Stiles & Skutch 1989, Neudorf & Blanchfield 1994, Stutchbury et al. 1996) but has been reported occasionally breeding in second growth vegetation without bamboo (Restall et al. 2006). CONCLUSION.—Available evidence on the distribution, movements, habitat use, abundance, feeding ecology, and breeding, of C. godefrida, S. falcirostris, and S. frontalis supports the idea that they specialize on Guadua bamboo seeds in Argentina. The high vagility of bamboo seed specialists, typical of nomadic birds that follow ephemeral food resources (Keast 1961, Kratter 1997, Sick 1997), may render these species relatively robust to habitat fragmentation. However, because many resource patches are required to support a single individual over time, and because alternative food sources may play a key role when bamboo seeds are lacking, these species might be highly vulnerable to habitat loss of both their preferred bamboo stands and the intervening areas with alternative food sources. In some parts of Minas Gerais, S. frontalis appears to be locally extinct, perhaps because of the loss of connectivity between the Serra da Mantiqueira and the Serra do Mar rather than the disappearance of the local bamboos (Vasconcelos 2002). An alternative explanation is to consider that bamboos that might provide food for the species have not flowered for the last 25–30 yr (M. Vasconcelos, pers. comm.). Loss of large bamboo stands in North America might have caused the extinction of the Bachman’s warbler Vermivora bachmanii, the only insectivorous bird in North America thought to specialize on bamboo (Remsen 1986). It may also have contributed to the demise of the granivorous passenger pigeon Ectopistes migratorius, now extinct (Janzen 1976). Much remains to be studied on bamboo seed specialist birds, and such studies are key for informing conservation efforts. In Misiones, the next masting event of G. trinii is expected to occur in 2018, and will likely provide excellent opportunities for further studies on bamboo seed specialists. We also encourage Paraguayan and Brazilian ornithologists to record numbers of bamboo seed specialists, observe their behavior, identify the dominant seed-bearing bamboos where these birds are observed, and collect detailed data

71

on the timing of masting of Guadua and other bamboos. This information is necessary to determine the importance of the different bamboo species for these three birds outside of Argentina. Special attention should be given to those bamboo genera closely related to Guadua, such as Eremocaulon and Apoclada (Soderstrom & Londo˜no 1987, Londo˜no & Clark 2002). Conservation of bamboo seed specialists remains a difficult task, given the lack of information about their ecology. Life histories of these species have not been adequately studied and might differ drastically from other closely related species that do not specialize on bamboo (Kratter 1997). These fragile, unique, and spectacular birds and habitats deserve our best effort before it is too late.

ACKNOWLEDGMENTS We wish to thank M. Srur, X. Londo˜no, L. Clark, L. Montti, Z. R´ugolo de Agrasar, L. Iharlegui, and G. Delucci for helping with data and references on bamboo taxonomy and ecology, M. Holzmann and T. Agostini for housing and bamboo-stand provisioning in Puerto Iguaz´u, and J. Van Remsen, J. Worrall, an anonymous reviewer for helpful suggestions that improved the manuscript. We thank N. Fari˜na, C. Ram´on, K. Roesler, E. Jordan, M. Juhant, L. Pagano, J. Segovia, and M. Holzmann for sharing in field observations; R. Restall, D. Franklin, and C. S´anchez for bibliography and discussions about bamboo specialists; J. Herrera and J. Calo for helpful comments on the fieldwork; C. Savigny, C. Rizzo, G. Trivelato, and U.-P. St¨auble for sharing their observations of S. frontalis and S. falcirostris; and M. Vasconcelos and M. Vel´azquez for sharing their knowledge of bamboo birds in Brazil and Paraguay. E. Krauczuk first showed S. falcirostris to JIA in Corpus Christi. We are grateful to M. Drever for writing the code for the randomization test and to Z. Z. Joroscho for greatly improving the map. We also thank all the curators who allowed us access to collections or shared specimen information with us. The Delegaci´on Noreste Argentina de la Administraci´on de Parques Nacionales and the Ministerio de Ecolog´ıa, R.N.R y Turismo de la Provincia de Misiones granted permission to work at PN Iguaz´u and other field sites. For funding and equipment, we are grateful to a Pamela and Alexander Skutch Award, the Rufford Foundation, IdeaWild, Killam Trusts, and Optics for the Tropics. We wish to dedicate this article to L. Parodi, for his long lasting contributions to bamboo systematics and ecology.

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: APPENDIX S1. Records of flowering/seeding of the five woody bamboo species widespread in Misiones, Argentina. APPENDIX S2. Records of the avian study species in the Atlantic forest of Misiones, Argentina. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors.

72

Areta, Bodrati, and Cockle

Any queries (other than missing material) should be directed to the corresponding author for the article.

LITERATURE CITED BERTONI, A. DE W. 1901. Aves nuevas del Paraguay: cat´alogo de las aves del Paraguay. An. Cient. Paraguay 1, ser. 1: 1–216. BERTONI, A. DE W. 1913. Contribuci´on para un Cat´alogo de Aves Argentinas. An. Soc. Cient. Arg. 75: 64–102. BERTONI, A. DE W. 1917. El Tamb´u y la muerte de las tacuaras en Sud-Am´erica. Proc. Second Pan Am. Sci. Cong. 4: 812–814. BERTONI, M. S. 1918. Gramin´aceas de las regiones forestales del Alto Paran´a (Paraguaya, Brasile˜na y Argentina). Avec des conclusiones phytog´eographiques. An. Cient. Paraguay 2, ser. 2: 143–166. BIRDLIFE INTERNATIONAL. 2007. Species fact sheets: Claravis godefrida, Sporophila falcirostris, Sporophila frontalis. (accessed 17 December 2007). BODRATI, A., AND J. I. ARETA. 2006. La Mosqueta Pecho Pardo (Hemitriccus obsoletus) en la Argentina y comentarios sobre su h´abitat y distribuci´on. Ornitol. Neotrop. 17: 597–600. BODRATI, A., AND K. COCKLE. 2006. Habitat, distribution, and conservation of Atlantic Forest birds in Argentina: Notes on nine rare or threatened species. Ornitol. Neotrop. 17: 243–258. CAMPBELL, J. J. N. 1985. Bamboo flowering patterns: A global view with special reference to East Asia. In Gupta, B. K. (Ed.) Symposium on recent advances in plant sciences, pp. 95–118, D. A. V. College, Dehradun, India. CASTELINO, M. 1990. Un ave nueva para la rep´ublica Argentina y segunda menci´on para otra. N´otulas Faun´ısticas 21: 1–2. CHEBEZ, J. C. 1986. Nuestras aves amenazadas, 16: Palomita Morada (Claravis godefrida). Nuestras Aves 4: 21–22. ˜ NIETO, L. G. COLLAR, N. J., L. P. GONZAGA, N. KRABBE, A. MADRONO NARANJO, T. A. PARKER, III, AND D. C. WEGE. 1992. Threatened birds of the Americas, The ICBP/IUCN Red data Book 2, 3rd edition. International Council for Bird Preservation, Cambridge, UK. DUTRA, J. 1938. Les bambus´ees de Rio Grande du Sud. Rev. Sudamer. Bot. 5: 145–152. FRANKLIN, D. C. 2004. Synchrony and asynchrony: Observations and hypotheses for the flowering wave in a long-lived semelparous bamboo. J. Biogeogr. 31: 773–786. FRANKLIN, D. C. 2005. Avian granivores consume flowers, not just seed, of the Top End Bamboo Bambusa arnhemica. Northern Territory Naturalist 18: 45–50. GREENEY, H. F., N. SHERMAN, R. LYNCH, AND I. HARMS. 2007. The nest and eggs of Maroon-chested Ground-dove Claravis mondetoura in south-east Ecuador. Cotinga 28: 71–73. JACKSON, H. D. 1972. The status of the Pied Mannikin, Lonchura fringilloides (Lafresnaye) in Rhodesia and its association with the bamboo Oxytenanthera abyssinica. Rhodesia Science News 6: 342–348. JANZEN, D. H. 1976. Why bamboos wait so long to flower. Ann. Rev. Ecol. Syst. 7: 347–391. JUDZIEWICZ, E. J., AND L. G. CLARK. 1991. An overview of the diversity of New World bamboos. J. Am. Bamboo Soc. 8: 117–122. ˜ , AND M. J. STERN. 1999. American JUDZIEWICZ, E. J., L. G. CLARK, X. LONDONO bamboos. Smithsonian Institution Press, Washington, DC. KEAST, A. 1961. Bird speciation on the Australian continent. Bull. Mus. Comp. Zool. 123: 305–495. KELLY D. 1994. The evolutionary ecology of mast seeding. Trends Ecol. Evol. 9: 465–470. KLIMAITIS, J. F. 2004. Memorias del atardecer. Relatos de un naturalista. Ediciones del Hogar, Berisso, Argentina. KRATTER, A. W. 1997. Bamboo specialization by Amazonian birds. Biotropica 29: 100–110.

LEBBIN, D. J. 2006. Notes on birds consuming Guadua bamboo seeds. Ornitol. Neotrop. 17: 609–612. LEBBIN, D. J. 2007. Habitat specialization among Amazonian birds: Why are there so many Guadua bamboo specialists? PhD Dissertation, Cornell University, New York. LENTINO, M., AND R. RESTALL. 2003. A new species of Amaurospiza Blue Seedeater from Venezuela. Auk 120: 600–606. LIJTMAER, D., N. M. M. SHARPE, P. L. TUBARO, AND S. C. LOUGHEED. 2004. Molecular phylogenetics and diversification of the genus Sporophila (Aves: Passeriformes). Mol. Phyl. Evol. 33: 562–579. ˜ , X., AND P. M. PETERSON. 1992. Guadua chacoensis (Poaceae: BamLONDONO buseae), its taxonomic identity, morphology and affinities. Novon 2: 41–47. ˜ , X., AND L. G. CLARK. 2002. A revision of the Brazilian bamboo LONDONO genus Eremocaulon (Poaceae: Bambuseae: Guaduinae). Syst. Bot. 27: 703–721. MCCLURE, F. A. 1966. The Bamboos. A fresh perspective. Harvard University Press, Cambridge, Massachusetts. NAVAS, J. R., AND N. A., BO´ .1986. Aves nuevas o poco conocidas de Misiones, Argentina. I. Neotr´opica 32: 43–44. NAVAS, J. R., AND N. A., BO´ . 1987. Sporophila falcirostris, nueva especie para la Argentina. Neotr´opica 33: 96. NEUDORF, D. L., AND P. J. BLANCHFIELD. 1994. The Slate-colored Seedeater (Sporophila schistacea): A bamboo specialist? Ornitol. Neotrop. 5: 129– 132. NICORA, E. G. AND Z. E. R´UGOLO DE AGRASAR. 1987. Los g´eneros de gram´ıneas de Am´erica Austral. Hemisferio Sur, Buenos Aires, Argentina. OLMOS, F. 1996. Satiation or deception?: Mast-seeding Chusquea bamboos, birds and rats in the Atlantic Forest. Rev. Brasil. Biol. 56: 391–401. PARKER, T. A., III. 1982. Observations of some unusual Rainforest and marsh birds in southeastern Peru. Wilson Bull. 94: 477–493. PARODI, L. R. 1936. Las bamb´useas ind´ıgenas en la Mesopotamia argentina. Rev. Arg. Agron. 3: 229–244. PARODI, L. R. 1941. Estudio preliminar sobre el g´enero “Chusquea” en la Argentina. Rev. Arg. Agron. 8: 331–345. PARODI, L. R. 1955. La floraci´on de la tacuara brava (“Guadua trinii”). Rev. Arg. Agron. 22: 134–136. PEARSON A. K., O. P. PEARSON, AND I. A. GOMEZ. 1994. Biology of the bamboo Chusquea culeou (Poaceae: Bambusoideae) in southern Argentina. Vegetatio 111: 93–126. R DEVELOPMENT CORE TEAM. 2007. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3–900051-07-0. . REID, S., I. A. D´IAZ, J. J. ARMESTO, AND M. F. WILLSON. 2004. Importance of native bamboo for understory birds in Chilean Temperate Forests. Auk 121: 515–525. REMSEN, J. V. 1986. Was Bachman’s Warbler a bamboo specialist? Auk 103: 216–219. RESTALL R. 1995. Munnias and Mannikins. Christopher Helm, London, UK. RESTALL R., C. RODNER, AND M. LENTINO. 2006. Birds of Northern South America. Christopher Helm, London, UK. RIDGELY, R. S., AND G. TUDOR. 1989. The birds of South America, vol 1. University of Texas Press, Austin, Texas. S´ANCHEZ, C. 2005. First description of the nest and eggs of the Slaty Finch (Haplospiza rustica) and observations on song and breeding behavior. Ornitol. Neotrop. 16: 493–501. S´ANCHEZ, J. E., AND D. A. HERNA´ NDEZ. 1990. El nido y huevos de Acanthidops bairdii (Emberizidae). Brenesia 34: 155–157. S´ANCHEZ, J. E. 1993. Abundancia, estructura de poblaci´on y dieta de Acanthidops. Congreso de Ornitolog´ıa de Costa Rica. I. Res´umenes, San Jos´e, CR, 20-22 de mayo, 1993, 1993-05-20: 11. Editorial San Jos´e, CIPA MNCR - PRMVS - UNA, CR. S´ANCHEZ, J. E., L. CHAVES, AND D. MART´INEZ. 2006. Observaciones de la Tortolita Serranera (Claravis mondetoura) en un h´abitat sin bamb´ues, con notas sobre su vocalizaci´on. Brenesia 65: 79–80.

Bird Specialization on Guadua Bamboo Seeds

SAVIGNY, C. in press. Aportes al conocimiento de la avifauna del norte de Misiones y el Parque Nacional Iguaz´u, Argentina. Nuestras Aves. SEIFRIZ, W. 1950. Gregarious flowering of Chusquea. Nature 165: 635–636. SENDULSKY, T. 1995. Merostachys multiramea (Poaceae: Bambusoideae: Bambuseae) and similar species from Brazil. Novon 5: 76–96. SHORT, L. 1973. Habits of some Asian woodpeckers (Aves, Picidae). Bull. AMNH 152: 253–364. SICK, H. 1997. Ornitolog´ıa Brasileira. Ed. Nova Fronteira, R´ıo de Janeiro, Brasil. ˜ . 1987. Two new genera of Brazilian SODERSTROM, T. R., AND X. LONDONO bamboos related to Guadua (Poaceae: Bambusoideae: Bambuseae). Am. J. Bot. 74: 27–39. SOLANO UGALDE, A., AND A. ARCOS TORRES. 2006. Primer registro reproductivo de Haplospiza rustica en Ecuador. Bolet´ın SAO 16: 20–25. STERN, M. J, K. GOODELL, AND D. K. KENNARD. 1999. Local distribution of Chusquea tomentosa (Poaceae: Bambusoideae) before and after a flowering event. Biotropica 31: 365–368. STILES, F. G., AND H. A. HESPENHEIDE. 1972. Observations on two rare Costa Rican finches. Condor 74: 99–101. STILES, F. G., AND A. SKUTCH. 1989. A guide to the birds of Costa Rica. Ithaca, New York. STOTZ, D. F., J. W. FITZPATRICK, T. A. PARKER, III, AND D. K. MOSKOVITS. 1996. Neotropical Birds. Ecology and Conservation. The University of Chicago Press, Chicago, Illinois.

73

STUTCHBURY, B. J. M., P. M. MARTIN, AND E. S. MORTON. 1996. Nesting behavior of the Slate-colored Seedeater (Sporophila schistacea) in Panam´a. Ornitol. Neotrop. 7: 63–65. SULEKIC, A. A., Z. E. R´UGOLO DEAGRASAR, AND L. G. CLARK. 1999. El g´enero Rhipidocladum (Poaceae, Bambuseae) en la Argentina. Darwiniana 37: 315–322. VASCONCELOS, M. F. 2002. O pixox´o (Sporophila frontalis) nos munic´ıpios do Serro e Santa B´arbara, Minas Gerais: Poss´ıveis casos de extinc¸oes locais. Atualidades Ornitologicas 103: 2. VASCONCELOS, M. F., A. P. VASCONCELOS, P. L. VIANA, L. PALU´ , AND J. F. SILVA. 2005. Observac¸o˜ es sobre aves gran´ıvoras (Columbidae e Emberizidae) associadas a` frutificac¸a˜o de taquaras (Poaceae, Bambusoideae) na porc¸a˜o meridional da Cadeia do Espinhac¸o, Minas Gerais, Brasil. Lundiana 6: 75–77. WEGE D. C., AND A. J. LONG. 1995. Key areas for threatened birds in the Neotropics. BirdLife International. BirdLife Conservation Series No. 5. WIDMER, Y. 1998. Flowering phenology of Chusquea bamboos with special reference to Chusquea talamancensis in Costa Rica. J. Am. Bamboo. Soc 12: 1–20. YOUNG, S. M., AND W. S. JUDD. 1992. Systematics of the Guadua angustifolia complex (Poaceae: Bambusoideae). Ann. Missouri Bot. Gard. 79: 737– 769.

APPENDIX S1. Records of flowering/seeding of the five woody bamboo species widespread in Misiones, Argentina, based on records from the literature, specimens from the Instituto Darwinion (SI), Museo de La Plata (LP), interviews with local inhabitants of Misiones and personal observations.

Guadua trinii 1- Rio Grande do Sul, Brazil, flowering/seeding reported in 1870 (Dutra 1938) 2- Rio Grande do Sul, Brazil, flowering/seeding reported in 1902 (Dutra 1938) 3- Río de la Plata, Buenos Aires, Argentina, flowering reported in 1920 by J. Molfino in litt. to D. Cristobal M. Hicken 1920 (Instituto Darwinion Archives). 4- Lower Paraná Delta, Buenos Aires, Argentina, natural flowering reported in austral spring 1922 and seeding in austral summer 1922-1923 (Parodi 1955). 5- Paraná Delta, Buenos Aires, Argentina, specimen bearing seeds collected on 28 March 1927 (Instituto Darwinion1236). 6- Rio Grande do Sul, Brazil, flowering/seeding reported in 1934 (Dutra 1938) 7- Lower Paraná Delta, Buenos Aires, Argentina, seeding reported in late 1953 (Parodi 1955). 8- Punta Lara, Buenos Aires, Argentina, flowering reported in October 1982 (Klimaitis 2004). 9- Parque Nacional Iguazú, Misiones, Argentina, flowering and seeding began in 1988 (Castelino 1990, Saibene et al. 1996), flowering and seeding continued during 1990 (Castelino 1990), and seeding ended by 1991 (Saibene et al. 1996). 10- Iguazú, Misiones, Argentina, specimen collected in December 1988 (Instituto Darwinion 2179, specimen not seen).

Guadua chacoensis 1- Colonia Nueva Germania, Paraguay, specimen bearing seeds collected in December 1916 (SI2311-10492). 2- Eldorado, Misiones, Argentina, specimen bearing seeds collected on 7 December 1943 (SI14526). 3- Puerto Montecarlo, Misiones, specimen bearing seeds collected on January 1943 (LP900486) 4- San Martin, across from Ituzaingó, Corrientes, Argentina, specimen bearing seeds collected on 8 February 1944 (SI6284).

5- Casa Lata, Chaco, Argentina, specimen bearing seeds collected on 28 July 1944 (SI11719). 6- Isla Apipé Grande, Ituzaingó, Corrientes, Argentina, specimen bearing seeds collected on 8 July 1973 (SI23851). 7- Rincón de San Pedro, Concepción, Corrientes, Argentina, specimen bearing seeds collected on 17 July 1974 (SI30687). 8- Arroyo Riachuelo, Capital, Corrientes, Argentina, specimen bearing seeds collected on 16 January 1974 (SI1859). 9- Rincón Ombú Chico, Ituzaingó, Corrientes, Argentina, specimen bearing seeds collected 3-5 July 1974 (SI25470). 10- Arroyo Riachuelo, Capital, Corrientes, Argentina, specimen bearing seeds collected on 15 August 1974 (SI2402). 11- Rincón de San Pedro, Corrientes, Argentina, specimen bearing seeds collected on 17 December 1974 (SI2870). 12- Arroyo Riachuelo, Capital, Corrientes, Argentina, specimen bearing seeds collected on 21 December 1974 (SI635). 13- Arroyo Riachuelo, Capital, Corrientes, Argentina, specimen bearing seeds collected on 27 September 1975 (SI3125). 14- Peñon Reina Victoria, San Ignacio, Misiones, Argentina, specimen bearing seeds collected on 14 January 1976 (SI28759). 15- Iguazú, Misiones, Argentina, mast seeding reported in 1978 (M. Srur pers. com. 2007). 16- Corpus Christi, Misiones, Argentina, mass seeding observed in 2005 and flowering estimated to have begun in 2004 (JIA pers. obs.). 17- Berón de Astrada, Corrientes, Argentina, flowering specimens collected in September 2005 (SI5, specimen not seen). 18- Iguazú, Misiones, Argentina, mass flowering and seeding observed from October 2006 through January 2008 (pers. obs.).

Merostachys claussenii 1- Rio Grande do Sul, Brazil, flowering reported in 1877 (Dutra 1938). 2- Rio Grande do Sul, Brazil, flowering reported in 1907 (Dutra 1938). 2- Alto Paraná Region, Paraguay, specimen bearing seeds collected in 1909/1910 (SI5418). 3- Valbinhos, Paraná, Brazil, specimen bearing seeds collected on 11 November 1910 (SI10761).

4- Iguazú, Misiones, Argentina, specimen bearing seeds collected on 11 August 1910 (SI408). 5- Ypiranga, Parana, Brazil, specimen bearing seeds collected on 15 January 1915 (SI14367). 6- Loreto, Misiones, Argentina, specimen bearing seeds collected in January 1940 (SI13645). 7- Colonia Montecarlo, Misiones, Argentina, specimen bearing seeds collected on 2 May 1943 (SI33). Flowering reported in 1941, and by May 1943 few culms had seeds (note on label of specimen). 8- San Pedro, Misiones, Argentina, masting event reported from 1972 to 1974 (T. Debarba pers. com. 2008). 9- Reserva Forestal San Jorge, Misiones, last week of September 2004, mass mortality and seeding photographed by JIA and AB 10- San Pedro, Misiones, Argentina, flowers photographed in 2004 by G. Capuzzi, and seeds observed in 2005-2006 (pers. obs.). By 2007, few solitary culms could be found with seeds despite search over wide geographic areas.

Chusquea tenella 1- Tacuarembó, Uruguay, seeding reported in 1896 (Parodi 1941). 2- São Leopoldo?, Rio Grande do Sul, Brazil, seeding reported in 1901 (Dutra 1938). 3- Río Uruguay, Santa Catarina, Brazil, specimen bearing seeds collected on 25 February 1916 (SI17788). 4- São Leopoldo, Rio Grande do Sul, Brazil, seeding reported and specimen bearing seeds collected in March 1916 (Dutra 1938, SI446). 5- São Leopoldo?, Rio Grande do Sul, Brazil, seeding reported in 1932 (Dutra 1938). 6- São Paulo, Brazil, specimen bearing seeds collected on 23 May 1948 (SI). 7- San Pedro, Misiones, Argentina, seeding reported close to the year 1993 (T. Debarba pers. com. 2008).

Chusquea ramosissima 1- São Leopoldo?, Rio Grande do Sul, Brazil, flowering reported in 1893 (Dutra 1938). 2- Puerto Esperanza, Misiones, Argentina, specimen bearing flowers collected in March 1907 (LP008036). 3- Alto Paraná region, Paraguay, specimens bearing flowers and seeds collected in 1909/1910 (SI5944). 4- São Leopoldo, Rio Grande do Sul, Brazil, flowering reported and specimen bearing seeds collected in September 1916 (Dutra 1938, SI418). 5- Leandro N. Alem, Misiones, Argentina, specimen bearing flowers collected in July 1940 (LP050165).

6- Santo Pipó, Misiones, Argentina, specimen bearing flowers collected in July 1945 (LP905010). 7- Arroyo Garuhapé, Puerto Rico, Misiones, Argentina, specimen bearing seeds collected on 12 July 1945 (SI787). 8- Arroyo Carajá del Olimar, Sta. Clara, Dpto. 33, Uruguay, specimen bearing flowers collected on 9 October 1945 (Rosengurt N°B-4836 at LP). 9- Caí, Rio Grande do Sul, Brazil, specimen bearing flowers collected on 18 July 1949 (SI42566). 10- Between Porto Alegre and São Francisco de Paula, Rio Grande do Sul, Brazil, specimen bearing flowers collected on 29 January 1964 (SI25062). 11- Iguazú, Misiones, Argentina, specimen bearing seeds collected on 5 March 1995 (SI5269). 12- PN Iguazú, Misiones, Argentina, specimen bearing seeds collected on 3 April 1997 (SI3941). 13- Reserva Privada Yaguaroundí, Departamento Guaraní, Misiones, Argentina, mast seeding observed in May 2004 (AB and KC, pers. obs.). 14- Parque Provincial Urugua-í, mass mortality observed in January 2005 (pers. obs.). 15- Specimens in vegetative state are known from 1910, 1922, 1926, 1932, and 1945 (LP008305, LP050169, LP008034, LP038793 and LP900471).

APPENDIX S2. Records of Purple-winged Ground Dove (Claravis godefrida), Buff-fronted Seedeater (Sporophila frontalis) and Temminck’s Seedeater (Sporophila falcirostris) in the Atlantic forest of Misiones (Argentina) based on our own observations, third parties data, literature and MACN specimens. No specimens relevant to our study were housed at Museo de La Plata (MLP, La Plata), Museo Antonio Serrano (MAS, Paraná), Fundación Miguel Lillo (FML, Tucumán), Museu de Zoologia Universidade de São Paulo (MZUSP, São Paulo), American Museum of Natural History (AMNH, New York), Field Museum of Natural History (FMNH, Chicago), University of Kansas Museum of Natural History (KU, Kansas), Louisiana State University (LSU, Louisiana), and Yale Peabody Museum (YPM, New Haven).

Purple-winged Ground Dove (Claravis godefrida) 1- Arroyo Urugua-í km 30 (Dpto Iguazú), 18 October 1957, immature male collected by W. H. Partridge (MACN, Navas & Bó 1986, pers. obs.). 2- Wanda surroundings, (Dpto. Iguazú), December 1974, a pair observed by C. Olrog (Olrog 1979). 3- PN Iguazú (Dpto. Iguazú), August 1977, a single individual banded by A. Tarak and M. Christie (Olrog 1979, Saibene et al. 1996). 4- PN Iguazú (Dpto. Iguazú), 19 July 1990, two individuals observed by H. Povedano on the Macuco trail (Saibene et al. 1996, H. Povedano pers. com.). 5- Pto. Iguazú (Dpto. Iguazú), one individual observed in May 1991 by F. Lambert (Wege & Long 1995). 6- PN Iguazú (Dpto. Iguazú), 29 and 30 November 2007, two females were observed 10 km apart by AB and J. Segovia. Both were seen on a dirt road, less than 100 m from seeding stands of G. chacoensis. 7- Misiones, no details (Bertoni 1913).

Buff-fronted Seedeater (Sporophila frontalis) 1- PN Iguazú (Dpto. Iguazú), September 1978, one individual banded and released by A. Tarak (Olrog 1979). 2- PP Esmeralda (Dpto. San Pedro), 25 August 1993, unspecified number of birds seen by E. Krauczuk (Chebez 1994, Mazar Barnett & Pearman 2001). 3- Puerto Iguazú (Dpto. Iguazú), 19 May 2004, male and female observed by Christian Savigny in bamboo and riparian vegetation by the lower Iguazú river (Savigny in press).

4- PN do Iguaçu (Parana State, Brazil), 31 October 2007, a pair observed by Urs-Peter Stäuble building a nest close to a small swamp near the Trilha das Bananeiras (U. Stäuble in litt.). 5- PN Iguazú (Dpto. Iguazú), 16 January 2008, an individual with no white in the front, short creamy-whitish postocular eyebrow and two creamy wing bands (“taquara” according to Sick 1997) was observed by JIA and I. Roesler feeding with S. falcirostris on a clump of G. chacoensis . 6- Misiones, no details (Bertoni 1913).

Temminck’s Seedeater (Sporophila falcirostris) 1- Arroyo Urugua-í km 30 (Dpto. Iguazú), 26 November 1957, a female collected by W. H. Partridge (MACN, Navas & Bó 1987). 2- Arroyo Urugua-í km 10 (Dpto. Iguazú), 4-5 August 1958, two males and two females collected by W.H. Partridge (MACN, Navas & Bó 1987). 3- PN Iguazú (Dpto. Iguazú), August 1977, unspecified number of birds seen by R. Ridgely (Collar et al. 1992). 4- PN Iguazú (Dpto. Iguazú), 15-17 August 1988, a pair observed building a nest near a waterfall (Castelino 1990). 5- PN Iguazú (Dpto. Iguazú), 15 April 1989, three birds tape recorded in stands of flowering G. trinii bamboo (Castelino 1990). 6- PN Iguazú (Dpto. Iguazú), October 1989, unspecified number of birds seen by B. Whitney (Collar et al. 1992). 7- PN Iguazú (Dpto. Iguazú), October 1990, unspecified number of birds seen by B. Whitney (Collar et al. 1992). 8- PN Iguazú (Dpto. Iguazú), June and July 1992, birds seen by C. Saibene (Saibene et al. 1996). 9- PN Iguazú (Dpto. Iguazú), September 1993, unspecified number of birds seen by H. Povedano and S. Di Martino (Mazar Barnett & Pearman 2001, H. Povedano pers. com.). 10- PN Iguazú (Dpto. Iguazú), 19 July 1994, a pair seen by J. Mazar Barnett (Mazar Barnett & Pearman 2001). 11- Corpus Christi (Dpto. San Ignacio), 10 October 2005, at least three males (two of which were tape recorded) and two females, in flowering and seeding stands of G. chacoensis (JIA & E. Krauczuk, pers. obs.). 12- Puerto Iguazú (Dpto. Iguazú), 3 August 2007, one bird heard twice, but not seen, in a seed-bearing clump of G. chacoensis in the town (JIA, pers. obs.).

13- PN Iguazú (Dpto. Iguazú), 4-6 September 2007, at least four males (all tape recorded) and one female seen feeding on G. chacoensis seeds in several stands along upper Río Iguazú near Seccional Apepú (JIA, AB & KC, pers. obs.). 14- Puerto Iguazú (Dpto. Iguazú), 22 November 2007, a female video-taped feeding on seeds of G. chacoensis in the town (JIA, pers. obs.). 15- PN Iguazú (Dpto. Iguazú), 28-30 November 2007, 25-30 individuals (males and females; several tape-recorded and photographed) observed feeding on G. chacoensis seeds and heard singing in seed-bearing stands of G. chacoensis along upper Río Iguazú on the road to Seccional Apepú (AB, KC, N. Fariña, J. Segovia, C. Ramón, pers. obs.). 16- PN Iguazú (Dpto. Iguazú), 16-18 January 2008, over 55 individuals (males, females and fledglings; several taperecorded) observed feeding on G. chacoensis seeds and singing in seed-bearing stands of G. chacoensis along upper Río Iguazú from Garganta del Diablo to the Seccional Apepú (JIA, AB, I. Roesler, L. Pagano, E. Jordan, I. Holzmann, pers. obs.).