A distinctive avian assemblage ( Aves : Passeriformes ) in Western Darién , Panama is uncovered through a disease surveillance program

Basic knowledge about the distribution of flora and fauna is lacking for most tropical areas. Improving our knowledge of the tropical biota will help address contemporary global problems, including emerging tropical diseases. Less appreciated is the role that applied studies can have in improving our understanding of basic biological patterns and processes in the tropics. Here, I describe a novel avifauna assemblage uncovered in Western Darién province in the Republic of Panama that was uncovered during a vector-borne disease surveillance program. I compared the passerine bird species composition at 16 sites using records from recent ornithological expeditions sponsored by the Smithsonian Tropical Research Institute in Central and Eastern Panama. Based on the results of a Mantel test, geographic distance did not correlate with pairwise distinctiveness of sites. Instead, based on an index of distinctiveness modified from the Chao-Jaccard index, most sites were more or less similarly distinctive, with one site, Aruza Abajo, significantly more distinctive than the rest. I found that the distinctiveness of this site was due not only to the presence of several rare and range-restricted taxa, but also to the absence of taxa that are common elsewhere. This finding provides more evidence of high species composition turnover (beta-diversity) in the Panamanian biota, which appears to be driven by a combination of soil and climate differences over narrow distances. Rev. Biol. Trop. 62 (2): 711-717. Epub 2014 June 01.

For most tropical areas, we continue to lack basic knowledge regarding distributional patterns of biodiversity as well the processes that create and maintain this biodiversity.At the same time, this knowledge gap impedes our ability to address problems of conservation, agriculture, and mitigation of emerging tropical diseases using the best available science.Recently, many natural history museums have begun to close this knowledge gap in both basic and applied biology through expeditions in collaboration with applied scientists (Winker, 2004;Pyke & Ehrlich, 2010).While it is routinely acknowledged that basic scientific discovery informs applied scientific inquiry (Yasué et al. 2006), less attention is given to the role that applied scientific endeavors can play in improving our basic knowledge about biodiversity patterns, especially in the tropics.Here, I describe an instance where a collection trip to survey birds and mosquitos for equine encephalitic virus uncovered a unique and overlooked bird assemblage in Panama, a country with one of the best-studied avifauna in the Neotropics.
The Smithsonian Tropical Research Institute Bird Collection (STRIBC) was founded in 2007 in order to better describe patterns of avian biodiversity across the Isthmus of Panama as well as to understand interactions between birds and other environmental factors and biological agents, especially with regard to emerging tropical diseases.In 2010, at the onset of an outbreak of equine encephalitis in Eastern Panama that had claimed at least two human lives and caused considerable loss of livestock, the Panamanian Ministry of Agriculture requested that the STRIBC sample wild birds and mosquitoes in Aruza Abajo (geographical coordinates: 8.36, −77.95,Fig. 1), an affected site in Western Darién Province.Equine encephalitis outbreaks occur periodically in Eastern Panama, but almost nothing is know about the ecology of viral transmission in the area, including which mosquito species are local vectors (Navia-Gine et al. 2013) or whether rodents or birds are natural hosts (Arrigo, Adams, & Weaver, 2010).
At Aruza Abajo, I was impressed by many of the birds we encountered.For example, we collected the second Middle American specimen of the flycatcher Tolmomyias flaviventris, the third, fourth and fifth Middle American specimens of the woodpecker Colaptes punctigula, and several specimens of Phaethornis anthophilus, being the most abundant hummingbird at the site.This species had only been recorded in Panama from five specimens collected more than 45 years ago.The ornithological literature for Panama (Ridgely & Gwynne, 1989;Angehr & Dean, 2010) refers to the Eastern portion of Darién province as being the most distinctive in Panama, but it is clear from these sources and others (González, Eberhard, Lovette, Olson, & Bermingham, 2003;Siegel & Olson, 2008) that Western Darién province has been systematically overlooked by ornithologists.In order to test my observation that the avifaunal assemblage in Western Darién province, and especially Aruza Abajo, was distinctive, I compared our records from Aruza Abajo to those of other locations where my colleagues and I have collected birds in Panama since 2002.

MATERIALS AND METHODS
I generated species incidence lists for 16 locations (Fig. 1) where the STRIBC or STRIsponsored foreign natural history museums have collected birds in central and Eastern Panama since 2002.I participated in the collecting efforts at 15 of the 16 locations.Field work at each site varied from two and 13 days.Most expeditions were disease ecology surveys, while a few were solely biodiversity surveys, but all shared the same goal of sampling broadly across the local bird community, rather than focused efforts to collect particular species of interest.In all cases, the majority of collecting employed standard ornithological mistnets, but occasionally collecting was supplemented by shotgun, where permitted.Data come from the digital collections databases managed by STRI.Although our fieldwork included collecting both passerines and non-passerines, in order to improve consistency among sites, I restricted my analyses to resident passerines only.All scientific collecting in Panama sponsored by STRI was approved by ANAM, Panama's Environmental Authority (permit numbers: DNPN-01-2002, DNPN-01-2003, DNPN-01-2004, DNPN-01-2004, SE/A-60-10, SE/A-137-10, SE/A-96-09, SE/A-44-10, SE/A-66-11, SE/A-2-12); likewise, scientific collecting undertaken by the STRI Bird Collection is done under IACUC approval (permits: 2007-03-03-15-07, 2011-0927-2014-03, 2013-0801-2016).
I used the Chao-Jaccard similarity index modified for replicated incidence data (Chao Chao, Chazdon, Colwell, & Shen, 2005) to measure passerine species assembly distinctiveness among the 16 localities.For each site, I tallied daily incidence (i.e.presence vs. absence).Therefore, for a given site, the total incidence for any given species could vary from 0 (never collected at that site) to n, where n equals the number of days of fieldwork at that site.I chose this replicated incidence approach because effort at each site varied in several ways: the number of days at a site, the number of nets employed and the hours of operation in a day, whether shotgun collecting was allowed, and because the total maximum number of collected specimens per species from a given site varied among species based either on our Panamanian collecting permit or our overall research objectives.Details about number of field personnel, mistnets deployed, and mistnet-hours was not available.Therefore, replicated incidence rather than a raw abundance approach should best deal with these differences in sampling effort, as well as the problem of incomplete sampling at a given site (Chao et al., 2005(Chao et al., , 2006)).It is certain that at none of the 16 sites our collection efforts approached a full sampling of the local passerine community.The modified Chao-Jaccard index (ChaoJaccard inc-est ) was computed pairwise among all 16 sites in the program EstimateS 9 (Colwell, 2013).However, for the rest of this paper, I will be discussing assemblage distinctiveness, which I define as simply 1-ChaoJaccard inc-est .Thus, a distinctiveness value of 0 represents two sites where avifaunal assemblages overlap completely, whereas a value of 1 indicates two sites with completely different bird faunas.I tested for a relationship between pairwise site distinctiveness and the geographic distance using a Mantel test implemented in R (R Core Team 2012).In order to compare the distinctiveness among sites, I generated 16 t-tests, comparing each sites average pairwise distinctiveness measure to that of all other comparisons (that did not include that site).To correct for multiple comparisons, I employed a sequential Bonferroni correction which constrains the familywise error rate (e.g.overall P-value) to 0.05.

RESULTS
Across the 16 sites, 183 species of resident passerines were recorded.Among the 16 sites, raw passerine species richness varied between 16 and 71 (mean S: 42.4±18.0),while species incidence (i.e. total species days) varied from 20 and 219 (mean: 88.3±51.6).No species occurred on all 16 collecting tally lists, and only one species (Mionectes oleagineus) was recorded from 15 locations, with five additional species collected in at least 13 sites.Instead, the distribution of species shows a long tail of relatively rare taxa: 116 species (63.3% of all species recorded in the study) were collected in three or fewer sites and nearly a third of all species in the study (N=60) were collected in only one location.Data for all 16 collecting sites are available at the Figshare depository (DOI: 10.6084/m9.figshare.941075).
Among the 16 sites, distinctiveness averaged 0.60, and ranged from 0.52 to 0.76, with the latter representing Aruza Abajo.Among all pairwise comparisons, the two sites estimated to be most similar were Old Gamboa and El Salto, whereas the most distinctive were found to be Mendoza and Aruza Abajo.A Mantel test failed to rejected the null hypothesis that pairwise distinctive was independent of geographic distance (r=0.092,p=0.17,Fig. 2).The results of the 16 pairwise t-tests found that three locations had significantly higher distinctiveness scores than the remaining pair-wise comparisons before correcting for multiple tests (villa del Carmen: p=0.03;Bayano: p=0.009;Aruza Abajo: p=2.0×10 -7 , Fig. 3), with only Aruza Abajo remaining significant after sequential Bonferroni correction (corrected a=0.003).
Several of the passerines collected at Aruza Abajo represented near-endemic species found only in Eastern Panama and adjacent Northern Colombia, including the flycatcher Tolmomyias flaviventris, the wren Campylorhynchus albobrunneus, and the conebill tanager Conirostrum leucogenys.These taxa were collected only at Aruza Abajo, or in the case of the wren, at one additional site.However, other taxa that were collected exclusively or almost exclusively at Aruza Abajo include more widespread taxa such as the antbird Myrmeciza longipes and the wren Cantorchilus leucotis.Although these species can be found across much of Panama, they were more routinely collected in Aruza Abajo relative to the remaining 15 sites.Finally, it is important to note that while these findings refer to only passerines it is likely that the assemblage of non-passerines at Aruza Abajo is similarly distinctive; the hummingbird and woodpecker examples mentioned earlier are two species of non-passerines that were almost never collected outside Aruza Abajo.
The avifauna of Aruza Abajo is distinctive not just for what is there, but also for what is apparently missing.Only 11 of the 27 species that were collected at half of the study sites were collected at Aruza Abajo, whereas the median representation of these taxa at the other 15 sites was 20 of 27.One possible explanation for missing species is that they are replaced by novel taxa.The most likely example of taxon replacement is the antshrike Thamnophilus atrinucha, which tied for second place among most frequently collected species across the 16 sites (N=14), but appears to be replaced at Aruza Abajo by Thamnophilus nigriceps, an antshrike endemic to Eastern Panama and adjacent Northern Colombia.T. nigriceps was collected at just four sites in this study.
In general, our results agree with earlier studies that have highlighted the high betadiversity found in Panamanian biotic communities.Unlike Chust's finding for Panamanian trees (Chust et al., 2006), there is no evidence that distance drives the turnover in bird species in central and Eastern Panama, agreeing with a more recent study of trees of central Panama (Jones et al., 2013).Often, beta diversity in Panamanian systems has been attributed to the strong rainfall gradient between Caribbean and Pacific Panama, especially in the canal area (e.g.phytophagous beetles: Ødegaard, 2006;and birds: Rompré, Robinson, Desrochers, & Angehr, 2007).Alternatively, Jones et al. (2013) suggest that soil type along with rainfall best predict species turnover of trees and ferns in central Panama.Interestingly, evidence that soils contribute to patterns of tropical bird species distributions comes from recent studies that document a unique avifauna in Amazonian white-sand forests (Alonso & Whitney 2003;Alonso, Metz, & Fine, 2013).Determining the cause of the distinctive avifauna assemblage in Western Darién province, and whether such differences can be found in other flora and faunal assemblages, remains to be determined.It is also important to note that variation in the bird assemblage among Panamanian field sites may be related to the extremely high degree of phylogeographic variation (i.e.within species genetic variation) that has been routinely observed in lowland Panamanian plants (Jones, Cerón Souza, Hardesty, & Dick, 2013) and animals (e.g.birds: González et al., 2003;Miller et al., 2008;Miller, Bermingham, Klicka, Escalante, & Winker, 2010;frogs: Wang, Crawford, & Bermingham, 2008;and bats: Clare, Lim, Fenton, & Hebert, 2011;Hauswaldt, Ludewig, vences, & Pröhl, 2011).
A poor understanding of faunal dynamics in tropical areas not only limits our ability to understand patterns of tropical biodiversity but also impedes our response to applied scientific problems such as vector-borne tropical diseases.In the case of the fatal 2010 outbreak of equine encephalitis in Western Darién, recent phylogenetic evidence indicates that the virus is likely endemic in the area (S.Weaver, pers.comm.), yet as demonstrated by this study, even the most basic understanding of potential vertebrate host communities in the area is lacking.These findings highlight the continuing need to re-examine the depth of basic biodiversity knowledge even in supposedly well-known tropical systems, and demonstrates the value of applied research programs to improving our knowledge of tropical biodiversity patterns.

Fig. 2 .
Fig. 2. visualization of a Mantel test comparing pairwise distinctiveness measures and geographic distance between 16 ornithological collecting sites in Central and Eastern Panama (r 2 =0.0085, p=0.17).Filled circles represent pairwise comparisons that include Aruza Abajo, unfilled circles represent all other pairwise comparisons.