~ Molecular biology of Sturnella meadowlarks

We are collaborating with Diane Caporale's Genetic Analysis Services laboratory at the University of Wisconsin-Stevens Point to answer some ecological questions with molecular methods. Specifically, we are interested in 1) confirming the presence of breeding Eastern Meadowlarks coexisting at the western extremity of their breeding range with Western Meadowlarks at our study site in the Nebraska Sandhills, 2) determining if any of the meadowlark nestlings were Eastern x Western hybrids, 3) characterizing the distribution of Eastern, Western, and potential hybrid nests, and 4) comparing the distribution of Eastern and Western Meadowlark nests with inter-specific count data (exclusively auditory or song-based detections) to assess the reliability of song as an identification trait in a sympatric zone.

Introduction

Wildlife management and conservation requires accurate demographic information to identify population trends and causal mechanisms. Unfortunately, a variety of environmental factors, species-specific behavioral and morphological characteristics, and sampling or observer error often prevent biologists from detecting 100% of the individuals present (i.e., true abundance), resulting in incomplete detectability (Burnham 1981, Thompson 2002). Depending on the degree and spatiotemporal consistency of bias due to incomplete detectability, researchers may make inaccurate population trend conclusions and consequent management recommendations. Wildlife ecologists have developed numerous sampling and modeling designs to account for detectability (Tyre et al. 2003), including that which arises from sampling error or inter-observer variation (Nichols et al. 1986).

One potential cause of observer-based variation in count data is the inaccurate identification of species that are morphologically and behaviorally similar, and such variation may be compounded in species that are relatively less detectable to begin with. Additionally, some of these species may be physically monomorphic, such as the Eastern Meadowlark (Sturnella magna; hereafter EAME) and the Western Meadowlark (Sturnella neglecta; hereafter WEME). Identifying monomorphic species in sympatric zones can be exceptionally difficult since behavioral traits like songs and calls are mostly learned (Thorpe 1958, Brenowitz et al. 1997, Marler 1997), and sometimes even inter-specifically shared (e.g., Szijj 1966).

Lanyon (1999) confirmed the close phylogenetic relationship of EAME and WEME (Fig. 1). The species are nearly identical morphologically, with the exception of inconspicuous and inconsistent plumage variation around the face (Lanyon 1994, Sibley 2000). Such unreliable subtleties do not facilitate species distinction during standardized bird surveys that typically limit observers to a 3-5 minute sampling period. Additionally, bird surveys generally require observers to count numerous individuals of numerous species simultaneously, and also keep track of moving individuals to avoid repeat detections. Thus, standardized multi-species bird surveys do not allow for lengthy focus on individuals to distinguish subtle differences. Consequently, Lanyon (1994) and most popular bird identification guides, such as Sibley (2000), recommend identifying EAME and WEME by song and call. Cody (1969), however, offered that coexistence of EAME and WEME in sympatric zones could lead to inter-specific song convergence, but other researchers argued more convincingly that songs and calls are not only mostly retained in sympatric zones (Lanyon 1960) and allopatric zones (Ordal 1976), but may also be the primary reproductive isolating mechanisms (Szijj 1966). Inter-specific song differentiation appears to be the norm according to these authors, but most of them also admit to periodically observing interspecific song-sharing or “hybrid songs”.

Apparent hybridization of wild EAME and WEME has been observed by Sutton and Dickson (1965), Szijj (1966), and Rowher (1972), but Szijj (1966) showed that egg fertility and F1 progeny survival were lower from mixed pairs. Lanyon (1979) conducted a long-term captive breeding study that revealed a 90% egg fertility for mixed pairs (not different than the 87% fertility observed from non-mixed pairs), but only 10% of eggs were fertile from pairs with hybrid (F2) adults. Thus, field and laboratory results suggest that EAME and WEME hybridizing is strongly selected against due to low reproductive success.

Lanyon (1994) suggests EAME breeds in Nebraska at the western edge of its range but are mostly restricted to river valleys such as the Platte and Niobrara (Fig. 2). Prior observations at our study site, the UNL Gudmundsen Sandhills Laboratory in the central Nebraska Sandhills (Fig. 3), suggests that breeding EAME may also be occurring at or beyond the western edge of their range in sub-irrigated wet meadows (i.e., interdunal valleys). The wet meadows are low elevation, water table-irrigated, interdunal plant communities that are typically more dense and dominated by cool-season grass, sedge, and forb species, whereas the adjacent upland plant communities on the dunes tend to support less vegetation and primarily warm-season plant species (Fig. 4 and 5). We periodically observed EAME-like songs in the wet meadow sites during the 2006 and 2007 breeding seasons but, given the inter-specific monomorphism, assumed only WEME were present according to published range maps. Further investigation revealed historic records and previous papers indicating that EAME are mostly limited to valleys in Oklahoma (Sutton and Dickson 1965) and river valleys in Nebraska (Rowher 1972), thus indicating the potential presence of breeding EAME.

Although it may be safe to assume that meadowlark individuals we observe producing EAME-like calls are, in fact, EAME, the possibility remains that they could be WEME and/or hybrids. Therefore, unlike any of the aforementioned studies based solely on morphometrics, we are using molecular techniques to confirm if breeding EAME is present at our site. Our more broadly-defined applied research question then is this: does inter-specific song provide a reliable indication of species occurrence and differentiation in a sympatric zone? Our competing research hypotheses are as follows:

Inter-specific song frequency and distribution:

1) Only WEME songs will be observed during the 2008 bird surveys (improbable given the prior observations of EAME-like songs in the wet meadows during 2006 and 2007).

2) WEME and EAME songs will be uniformly and randomly distributed (i.e., no spatial partitioning) across the landscape (also improbable given prior observations).

3) WEME and EAME song distribution will be partitioned along upland and lowland plant communities, but with EAME songs mostly limited to wet meadows while WEME songs occur in wet meadows and uplands (most probable based on prior observations and literature such as Sutton and Dickson (1965)).

Inter-specific sympatry and/or hybridization based on nestling DNA:

1) Only WEME nests will be present across landscape.

2) WEME and EAME nests will be uniformly distributed across landscape without hybridization

3) Hypothesis 2 but with hybridization occurring (but probably at low frequencies based on high offspring sterility rates (Lanyon 1979)).

4) WEME and EAME nest sites will be mostly partitioned along uplands and lowlands, respectively.

5) Hypothesis 4 but with hybridization also occurring.

Methods

We collected three growing feathers (with sterilized forceps) per nestling per meadowlark nest during the 2008 breeding season and froze samples the same day. We will isolate DNA from feathers for about 70 birds using a Qiagen extraction manufacturer’s protocol. We are using the mitochondrial gene COI for comparisons among individuals, which effectively identifies genetic differences within and between bird species (Tavares and Baker 2008). We will amplify the COI gene using PCR and viewing on a gel, and we will sequence PCR products in the UWSP Genetic Analysis Service laboratory.

We will compare DNA sequences within and between plant community types to assess population structure and distribution across the landscape. Specifically, we will spatially correlate species occurrence (WEME, EAME, and/or hybrids), as confirmed genetically, with EAME and WEME song occurrence obtained through standardized bird surveys across the study sites (see Fig. 6 for an example sampling map). For example, given the confirmed occurrence of EAME nests in wet meadows, we would expect a high level of spatial correlation between EAME nests and song occurrence. In addition, and conditional on the confirmation of EAME nests, we suspect that we may find a disproportionately high level of WEME song occurrence in the wet meadows relative to their actual nesting efforts (i.e., WEME nesting in adjacent uplands may be selecting nearby wet meadows for foraging).

Preliminary results and applications

We have successfully isolated DNA from a subsample of nestling feathers, and are currently refining the isolation protocol to maximize DNA acquisition before progressing to DNA replication via polymerase chain reaction (PCR) methodology.

Assuming song is inter-specifically unique, we detected substantial proportions of EAME in the two lowland wet meadow sampling sites but exclusively WEME in the upland dry prairie sites (Fig. 7), and distance sampling-based density estimates (Buckland et al. 2001) illustrate similar trends (Fig. 8).

Traditionally, EAME and WEME were distinguished solely by morphometrics and behavior but this approach may be uncertain in sympatric zones. Thus, our molecular-based results will allow us to confirm the use of song as a reliable species identification trait during standardized bird community sampling surveys. Additionally, assuming sympatry is confirmed, the results will facilitate the investigation of inter-specific genetic, reproductive, and nest site habitat-partitioning relationships.