Introduction

In most bird species the sexes are morphologically indistinguishable, ca 50 % of adults (60 % of passerines) and the majority of nestlings (Price and Birch 1996). Blood samples are often used for sex-typing birds; however, taking blood can cause stress, discomfort and may, on occasion, damage wings and/or introduce infection (Joint Working Group on Refinement 2001). It also requires training and appropriate permits. Therefore, less invasive techniques are preferred, especially when studying endangered species—for example, using shed feathers, museum specimens and swabs. Furthermore, studies of fertility and sex ratio require the ability to sex-type degraded tissue (e.g. unhatched eggs, Brekke et al. 2010). However, allelic dropout can occur when attempting to PCR-amplify large products from degraded samples (Toouli et al. 2000) and dropout causes errors in sex-typing (Robertson and Gemmell 2006). We therefore attempted to develop a primer set that amplifies small PCR products (<150 bp) on both the W and Z chromosomes to enable sex-typing of degraded samples.

Methods

Following Dawson et al. (2010), a zebra finch Taeniopygia guttata EST microsatellite sequence DV945670 (Replogle et al. 2008) was identified with strong homology to the chicken Gallus gallus Z chromosome. We created a consensus sequence from these homologous sequences using MEGA3 (Kumar et al. 2004) and designed a primer set using PRIMER3 v0.4.0 (Rozen and Skaletsky 2000). Both the forward and reverse primers were 100 % identical to both the zebra finch Z and chicken Z chromosomes (no homologous W chromosome sequence was available). In order to create a primer set suitable for amplifying degraded samples, we designed the primer set to amplify a small product (<150 bp) whilst avoiding the use of degenerate bases (Table 1). Primer sequences, melting temperatures and the expected and observed allele sizes in zebra finch and chicken are provided (Table 1). The locus (DV945670) was homologous to mRNA sequences of the guanine nucleotide binding protein (G protein), q polypeptide (GNAQ) present in many taxa. Both primer sequences were 100 % identical to 9/10 birds assessed, including passerines, penguins and other seabirds, eagle, duck and chicken (details provided in the footnotes of Table 1).

Table 1 A new sex-typing marker for birds (Z37B), designed from the zebra finch Taeniopygia guttata Z and chicken Gallus gallus Z chromosome
Full size table

Genomic DNA was extracted from non-degraded samples (bird blood, blood slides), samples of low quantity (shed and plucked feathers, buccal swabs), degraded tissue [dead embryos from unhatched eggs, dead nestlings, museum (toe pad) samples] and crocodile skin using an ammonium acetate protocol (Richardson et al. 2001) or commercial kits (for the museum specimens and mouth swabs). Full details of the samples and extraction methods are provided in Supplementary Table 1. The primer set was tested by sex-typing individuals of 42 avian species including one ratite, the brown kiwi Apteryx australis (25 families and 15 orders; Table 2), and the saltwater crocodile Crocodylus porosus. Individuals of known sex (both females and males) were included for 40 of the bird species (for two species known females were available but no known males; Table 2). Sexes were previously identified based on morphology, behaviour and/or sex-typing markers (Table 2). PCR reactions were performed in 2-µl (10-µl for museum samples) volumes containing ca 10 ng of lyophilised genomic DNA, 1 or 5 µl of QIAGEN Multiplex PCR Master Mix and 0.2 µM of each primer (with the forward primer fluorescently labelled with HEX). PCR amplification was performed using a DNA Engine Tetrad thermal cycler. PCR amplification conditions were 95 °C for 15 min; followed by 35 cycles of 94 °C for 30 s, 56 °C for 90 s, 72 °C for 1 min, and finally 60 °C for 30 min. PCR products were loaded on a 48-capillary ABI 3730 DNA Analyzer and genotypes assigned using GeneMapper software (Applied Biosystems).

Table 2 Assessment of the Z37B marker for sex-typing 42 species of birds belonging to 25 families in 15 orders using various tissue types including non-degraded blood, degraded tissue (dead unhatched embryos, dead nestlings and museum specimens) and samples of low quantity DNA (plucked/shed feathers and mouth swabs)
Full size table

Results

All 42 bird species tested amplified, as did the saltwater crocodile (amplicons = 81–110 bp, saltwater crocodile = 110 bp; Table 2). Twenty-one species (50 %) were successfully sexed: including passerines, shorebirds, rails, seabirds, eagles and the brown kiwi (Table 2). In all species sexed, the diagnostic W allele (81–92 bp) was smaller than the Z allele (92–100 bp; Table 2). The difference in size between the W and Z alleles within a species was small (2–19 bp, Table 2) and resolving this difference required an ABI DNA Analyzer.

Individuals were successfully sex-typed from the degraded tissues including unhatched embryos, dead nestlings and museum toe-pads (of the hihi Notiomystis cincta), and from samples of low-quantity DNA, i.e. plucked feathers (hihi, northern fulmar Fulmarus glacialis) and buccal swabs (corncrake Crex crex; Table 2). Individuals whose DNA was extracted from non-degraded blood samples were also successfully sex-typed (Table 2). Z37B successfully sex-typed individuals when included as part of a microsatellite multiplex set (hihi and corncrake; PB unpublished data).

Many species displayed a single allele (i.e. of same size) in both sexes and could therefore not be sexed (43 %; Table 2), probably due to failure of the W locus to amplify (or possibly a lack of difference in size between the Z and W amplicons). Eight species (19 %) displayed polymorphism in the Z locus and for three of these species (7 %) all females were homozygous—suggesting that the W locus failed to amplify (probably due to primer–W chromosome base mismatches; Table 2). In some species, the W locus might require a lower PCR annealing temperature to amplify, such as 50 °C. We recommend the use of Qiagen Multiplex Master Mix for PCR sex-typing because it more often enables amplification even when there are some target–primer base mismatches (DAD unpublished data). Although not causing error here, Z (and/or W) polymorphism can lead to error when interpreting sexes, unless a second sex-typing marker and/or known sexes are included (Dawson et al. 2001, Robertson and Gemmell 2006).

Marker Z37B is of utility for sex-typing degraded samples. It provides an alternative marker to validate sex-typing data. Most of the passerine species tested could be sex-typed with this marker, suggesting it will be of utility for sex-typing many of the ca 5,000 species in this order. In addition, the successful sex-typing of non-passerines including shorebirds, rails, seabirds, eagles and kiwi, suggests Z37B will be of utility in a wide range of species.