SCAR Primers Research Articles

Development and application of a SCAR marker to monitor and quantify populations of the postharvest biocontrol agent Pantoea agglomerans CPA-2

Pantoea agglomerans CPA-2 is an effective biocontrol agent of postharvest diseases of citrus and pome fruit. A monitoring technique was developed for its identification and to quantify its populations. The methodology used consisted of (i) searching for a semi-selective medium, (ii) identification of molecular markers and (iii) monitoring population dynamics in a commercial trial. As a semi-selective medium, Malonate Broth Agar supplemented with tetracycline hydroxychloride and incubation at high temperature (max. of 40 °C) facilitated the selective recovery of P. agglomerans CPA-2 colonies. The RAPD technique was applied to a collection of 13 strains of P. agglomerans, including CPA-2. Among the 12 primers tested, OPL-11 amplified a fragment (about 720 bp) specific to strain CPA-2. On the basis of this fragment, two SCAR markers were amplified using a primer pair derived from OPL-11 elongation. A first SCAR marker of 720 bp was specifically amplified for the strain CPA-2 and a second one of 270 bp was obtained for all P. agglomerans strains tested, including CPA-2. Commercial trials demonstrated a significant reduction of decay with the treatment of formulated cells of P. agglomerans CPA-2. Population dynamics of CPA-2 in commercial trials were determined on fruit surfaces and in the environment using both the classical plating technique and PCR with SCAR primers. In general, no significant differences were observed between results obtained from the two methods. On fruit surfaces, 1 day after CPA-2 applied its population by classical methods was 4.37 × 10 6 cfu wound −1 and at the end of the experiment the population increased to 5.8 × 10 5 cfu wound −1. The percentages of colonies identified as P. agglomerans CPA-2 at these sampling times using SCAR primers were 90 and 95%, respectively. Population dynamics in the environment to evaluate the environmental fate of P. agglomerans CPA-2 showed that it has a limited persistence and limited capacity for dispersion.

Inter and intra-population variability of Jatropha curcas (L.) characterized by RAPD and ISSR markers and development of population-specific SCAR markers

Jatropha curcas (Euphorbiaceae) is an oil-bearing species with multiple uses and considerable potential as a bioenergy crop. The present investigation has been undertaken to assess the extent of genetic diversity in a representative set of 42 accessions of J. curcas encompassing different crop growing regions in India along with a non-toxic genotype from Mexico as a prelude for utilization of promising and genetically divergent materials in the breeding programmes. Molecular polymorphism was 42.0% with 400 RAPD primers and 33.5% with 100 ISSR primers between accessions indicating modest levels of genetic variation in the Indian germplasm. The within-population variation based on RAPD polymorphism was 64.0% and was on par with the inter-population variation. Polymorphic ISSR markers have been identified that could differentiate the Indian accessions from the Mexican genotype and two of them were converted to SCAR markers. The SCAR primer pair ISPJ1 amplified a 543 bp fragment in all the Indian populations, while ISPJ2 with a specific amplicon of 1,096 bp was specific to the Mexican genotype. Population-specific bands have been identified for the accession from Kerala (2 RAPD markers), Neemuch-1 from Rajasthan (1 each of RAPD and ISSR markers) and the non-toxic genotype from Mexico (17 RAPD and 4 ISSR markers), which serve as diagnostic markers in genotyping. The study indicates an immediate need for widening the genetic base of J. curcas germplasm through introduction of accessions with broader geographical background.

Identification of AFLP markers linked to the male fertility restorer gene of CMS (Moricandia arvensis) Brassica juncea and conversion to SCAR marker

We have developed a cytoplasmic male sterile (CMS) line of Brassica juncea through somatic hybridization with Moricandia arvensis and introgressed the fertility restorer gene into B. juncea. This fertility restorer locus is unique in that it is capable of restoring male fertility to two other alloplasmic CMS systems of B. juncea. As a first step toward cloning of this restorer gene we attempted molecular tagging of the Rf locus using the amplified fragment length polymorphism (AFLP) technique. A BC(1)F(1) population segregating for male sterility/fertility was used for tagging using the bulk segregant analysis method. Out of 64 primer combinations tested in the bulks, 5 combinations gave polymorphic amplification patterns. Further testing of these primers in individual plants showed four amplicons associated with the male fertility trait. Polymorphic amplicons were cloned and used for designing SCAR primers. One of the SCAR primers generated amplicons mostly in the fertile plants. Linkage analysis using MAPMAKER showed two AFLP and one SCAR markers linked to the male fertility gene with a map distance ranging from 0.6 to 2.9 cM. All the markers are located on one side of the Rf locus.

Molecular identification and detection of Eutypa lata in grapevine

Eutypa lata, the causal agent of Eutypa dieback of grapevines, is difficult to identify on the basis of colony morphology and is often out-competed by other fungi when isolated from wood. To facilitate diagnosis of the pathogen, we designed SCAR primers capable of amplifying DNA of E. lata and constructed a genomic DNA library from which DNA sequences specific to E. lata were identified and sequenced. SCAR primers were used to identify E. lata directly from culture without the requirement for DNA extraction or prolonged incubation periods and could also detect the pathogen in DNA isolated from grapevine wood. RFLP probes were used in slot-blot assays to detect the pathogen in DNA isolated from 1 yr old cane as well as from mature grapevine trunks. The markers developed in this study have the potential to be used as a research tool to gather information on the epidemiology of the disease and to assess the efficacy of potential control agents against E. lata.

Development of a strain-specific quantitative method for monitoring Pseudomonas fluorescens EPS62e, a novel biocontrol agent of fire blight

Pseudomonas fluorescens EPS62e has been selected in a screening procedure for its high efficacy controlling Erwinia amylovora infections in flowers, immature fruits and young pear plants. We developed two monitoring methods which allowed specific detection and quantification of EPS62e by combining classical microbiological techniques with molecular tools. RAPD and unspecific-PCR fingerprints were used to differentiate EPS62e from other P. fluorescens strains. Differential amplified fragments from EPS62e were sequence characterized as SCAR markers and two primer pairs were designed and selected for their specificity against EPS62e. A SCAR primer pair was evaluated and validated for the assessment of population dynamics of EPS62e on pear plants under greenhouse conditions using plating and most probable number assays coupled to PCR. Both techniques were useful in monitoring the biological control agent. The population level of EPS62e after treatment was 7 log CFU (g f.w.) −1, which in turn decreased progressively to 4–5 log CFU (g f.w.) −1 after 17 days and then remained stable until the end of the assay 11 days later. The limit of detection of both monitoring methods developed was around 3 log CFU (g f.w.) −1, thus, providing a reliable tool for the analysis of EPS62e in greenhouse or field trials, and the assessment of threshold population levels for efficient biocontrol of fire blight.

Genetic relationships among strains of Xanthomonas campestrispv. campestrisrevealed by novel rep-PCR primers

Novel primers for rep-PCR were developed with the original software and based on `ancient diverged periodical sequences'. Rep-PCR with these primers was applied to study genetic relationships among 51 Xanthomonas campestris strains. The strains were collected from different countries including Russia, Japan, UK, Germany and Hungary. Reference strains of three X. campestrispathovars and five other Xanthomonas species were included. Based on qualitative differences in amplification profiles, the strains were divided into four major groups. Two subgroups recognised within X. campestrispopulation were similar to RFLP haplotypes. The third subgroup included strains of two other pathovariants and Japanese isolates of X. campestris pv. campestriswhile the fourth group comprised the other species of Xanthomonas. The analysis of the diversity within X. campestris resulted in the conclusion that isolates belong to distinct clonal populations (subgroups). The differences between the subgroups of X. campestris were only slightly smaller than between species of Xanthomonas. A PCR fragment about 600 bp amplified by primer KRPN2 was found in nearly all tested strains of X. campestris.SCAR primers designed for this marker produced a single specific band for strains of X. campestris, but not for other Xanthomonas, Pseudomonas and Erwiniastrains tested. Application of the new primer set for rep-PCR offers a rapid, simple and reproducible method for identification of bacterial strains. The X. campestris-specific SCAR primers may be used in diagnostics of this important plant pathogen.

Characterization of a RAPD fragment unique to species with hairy fruit skin in the genusActinidia

To develop a SCAR primer related to the hairy-fruit trait in the genusActinidia, we took a PCR-RAPD approach using arbitrary 10-mer primers. PCR with the UBC 376 primer generated specific fragments from three species with hairy fruit skin. Those fragments were then cloned to determine their nucleotide sequences. Two SCAR primers were designed from the UBC 376 primer and nucleotide sequences were obtained from the PCR fragments. A SCAR primer, OKC385, specifically amplified a 385-bp fragment from one clone ofActinidia eriantha, four ofActinidia chinensis, and four ofActinidia deliciosa. Deduced amino acid sequences of this fragment showed high sequence homology with plant cellulose synthases, which are involved in the biosynthesis of cellulose, a major cell wall component. The 385-bp fragment was specifically detected only in the seriesPerfectae C.F. Liang of sectionStellatae Li. This type has many hairs on the leaves, fruits, and stems, suggesting that the gene containing the PCR fragment is involved in hair formation in this phylogenetic group. Taken together, our results suggest that the SCAR primer, OKC385, can be used as a specific primer for early selection of the non-hair trait in breeding of the genusActinidia.

Tagging and mapping of a rice gall midge resistance gene, Gm8, and development of SCARs for use in marker-aided selection and gene pyramiding.

Using amplified fragment length polymorphisms (AFLPs) and random amplified polymorphic DNAs (RAPDs), we have tagged and mapped Gm8, a gene conferring resistance to the rice gall midge (Orseolia oryzae), a major insect pest of rice, onto rice chromosome 8. Using AFLPs, two fragments, AR257 and AS168, were identified that were linked to the resistant and susceptible phenotypes, respectively. Another resistant phenotype-specific marker, AP19(587), was also identified using RAPDs. SCAR primers based on the sequence of the fragments AR257 and AS168 failed to reveal polymorphism between the resistant and the susceptible parents. However, PCR using primers based on the regions flanking AR257 revealed polymorphism that was phenotype-specific. In contrast, PCR carried out using primers flanking the susceptible phenotype-associated fragment AS168 produced a monomorphic fragment. Restriction digestion of these monomorphic fragments revealed polymorphism between the susceptible and resistant parents. Nucleotide BLAST searches revealed that the three fragments show strong homology to rice PAC and BAC clones that formed a contig representing the short arm of chromosome 8. PCR amplification using the above-mentioned primers on a larger population, derived from a cross between two indica rice varieties, Jhitpiti (resistant parent) and TN1 (susceptible parent), showed that there is a tight linkage between the markers and the Gm8 locus. These markers, therefore, have potential for use in marker-aided selection and pyramiding of Gm8 along with other previously tagged gall midge resistance genes [ Gm2, Gm4(t), and Gm7].

Monitoring system for the biocontrol agent Pichia anomala strain K using quantitative competitive PCR‐ELOSA

A quantitative‐competitive PCR (QC‐PCR) method with enzyme‐linked oligosorbent assay (ELOSA) was developed to monitor Pichia anomala strain K, a biocontrol agent against postharvest diseases on apples. The procedure involved: (i) extraction of strain K DNA; (ii) coamplification of a constant amount of the extracted DNA (containing a strain K DNA marker of 262 base pairs (bp) specifically amplified with SCAR primers K1 and K2) with an internal standard (IS) titration; and (iii) differential hybridization with two specific biotinylated probes either for the target or for the IS sequence on microplate. The IS sequence differed from the target by only a short internal region of 35 bp providing the differential detection between both sequences. Both target and IS sequences proved to be competitive in PCR as well as in sandwich hybridization. Two copies of the target sequence were detected in the strain K genome by means of enzymatic restriction and Southern blot analysis. Varying amounts of strain K cell suspension were quantified in the phosphate buffer used for recovering cells from the apple surface. An accurate estimate of the strain K population was obtained from 103 to 106 yeast cells per apple. The threshold of the method was found to be at 1000 colony‐forming units per apple, which was around 100 times more sensitive than the plating method for monitoring strain K.

Development of SCAR markers for a Thinopyrum distichum chromosome that appears to be involved in salt tolerance

Sixty-seven F-1 plants from the backcross: Thinopyrum distichum/tetraploid ‘Henoch’ rye//diploid ‘Henoch’ rye were cloned, tested for salt tolerance (2–4 replications per genotype) and ranked. Sixty RAPD primers were tested on the parental lines. Nine of these produced Thinopyrum specific fragments and were applied to the whole population. Correlations between salt tolerance and presence of the Thinopyrum specific bands were mostly significant but small, as would be expected with a trait involving genes on several chromosomes. The nine markers were then also tested on a group of 15 B1 F1 plants from the same cross in which C-banding has identified the Thinopyrum chromosomes. It appeared that OPK6 (± 1060 bp) and OPK17 (± 652 bp) polymorphisms co-segregated with a chromosome putatively identified as belonging to homoeologous group 2 and implied in salt tolerance in a different study. The diagnostic RAPD bands were cloned and sequenced and specific SCAR primers were designed. These amplified DNA sequences specific for chromosome IIJ1 d that will be very useful in future introgression attempts.

RAPD markers linked to eastern filbert blight resistance in Corylus avellana

A total of 1,110 decamer primers were screened for RAPD markers linked to a dominant allele in hazelnut ( Corylus avellana) that confers resistance to eastern filbert blight caused by Anisogramma anomala. Twenty RAPD markers linked in coupling, and five markers linked in repulsion, were found. A seedling population was used to construct a linkage map of the region flanking the resistance locus. The map spans 46.6 cM, with 14 markers on one side of the resistance locus and eight on the other side. Eleven markers showed less than 3% recombination with resistance, including three that showed no recombination. Seven of these 11 markers are sufficiently robust to allow their use in marker-assisted selection. These include AA12(850) which shows no recombination, and six markers on one side of the resistance locus: 173(500), 152(800), 122(825), 275(1130), H19(650) and O16(1250). Marker 268(580), which flanks the resistance locus on the other side, is also suitable for use in marker-assisted selection, but shows 5.8% recombination with resistance. Other markers are less suitable for marker-assisted selection because of sensitivity to changes in primer or MgCl(2) concentration, or the long time required for electrophoresis to separate bands of similar size. The 16 markers closest to the resistance locus were cloned and sequenced. The W07(365) marker, which showed no recombination with the resistance locus but is difficult to score, includes a CT microsatellite repeat. The sequence information will allow the design of SCAR primers and eventual map-based cloning of the resistance allele.

Sex-specific SCAR markers in the dioecious plant Rumex nivalis (Polygonaceae) and implications for the evolution of sex chromosomes.

We developed SCAR primers based on isolated and sequenced male-specific fragments as identified in an AFLP analysis of the dioecious plant Rumex nivalis. PCR amplification using these primers on females and males resulted in fragments exclusively present in males. Co-amplification of the nuclear rDNA internal transcribed spacer 2 together with the male-specific fragment was applied as an internal control for successful PCR reactions to avoid false-negative sex scoring. With a length of about 164 bp, the AFLP fragment was of a similar size as the tandemly arranged, repetitive sequences of 180 bp located on the Y chromosomes of Rumex acetosa. The genetic distances between the Y-chromosomal sequences of R. nivalis and R. acetosa, both members of the section Acetosa, were substantial. We found intra-individual divergence among cloned sequences of the male-specific fragment in R. nivalis. The patterns of interspecific and intra-individual sequence variation found are in accordance with proposed modes of the evolution of sex chromosomes. Y chromosomes possibly arose only once in the genus Rumex and consist mainly of heterochromatic DNA. Due to the almost complete absence of selection on them, Y chromosomes are likely to accumulate large numbers of mutations.

Molecular markers for sex determination in papaya ( Carica papaya L.).

We have developed molecular markers tightly linked to Sex1, the gene that determines plant sex in papaya ( Carica papaya L.). Three RAPD products have been cloned and a portion of their DNA sequenced. Based on these sequences SCAR primers were synthesized. SCAR T12 and SCAR W11 produce products in hermaphrodite and male plants and only rarely in females. SCAR T1 produces a product in all papayas regardless of plant sex. SCAR T12 and SCAR W11 showed no recombination in a population of 182 F2 plants from a 'SunUp' by 'Kapoho' cross. Based on these results a PCR-based technique for rapidly and accurately determining the sex of papaya plants was developed using either W11 or T12 to detect the hermaphrodite or male allele and T1, which amplifies a product regardless of sex type, as a positive control. The sexing technique, using SCAR T12 and SCAR T1 as a positive control, was used to correctly predict hermaphrodite papaya plants in a population of seedlings with an overall accuracy of 99.2%.

Mode of reproduction is detected by Parth1 and Sex1 SCAR markers in a wide range of facultative apomictic Kentucky bluegrass varieties

Gametophytic apomixis in Kentucky bluegrass (Poa pratensis L.) involves the parthenogenetic development of unreduced eggs from aposporic embryo sacs. Marker-assisted selection for the mode of reproduction in P. pratensis would avoid costly and time-consuming phenotypic progeny tests. We developed and tested two SCAR primer pairs that are associated with the mode of reproduction in P. pratensis. The SCAR primers identified the apomictic and sexual genotypes among progenies of sexual x apomictic crosses with very low bias. Furthermore, when tested on a wide range of Italian and exotic P. pratensis germplasm, they were able to unequivocally distinguish sexual from apomictic genotypes. This system should, therefore, allow new selection models to be set up in this species. Abbreviations: AFLP – amplified fragment length polymorphism, BSA – bulked segregant analysis, MAS – marker-assisted selection, PCR – polymerase chain reaction, SCAR – sequence characterized amplified region

Design of grapevine (Vitis vinifera L.) cultivar-specific SCAR primers for PCR fingerprinting

Among 34 grapevine cultivars (Vitis vinifera L.), eight putative genotype-specific RAPD markers, from ’Albarino’, ’Caino blanco’, ’Chardonnay’, ’Folle blanche’, ’Grenache blanc’, ’Malvasia Sitges’, ’Torrontes’ and ’Treixadura’ respectively, were selected to transform into SCAR markers. Of these, seven markers were cloned and then five which showed a positive specific hybridization signal were sequenced. For these five markers, 30 sequence-specific primers ranging from 14 to 29 bases were designed to amplify genomic DNA from 64 grapevine cultivars under more-stringent PCR conditions. Only, two primer pairs, OpA111175p17R/ p17F and OpD10800p14R/p14F, still produced a specific SCAR marker, the ’Folle blanche’ ScA111175 and the ’Malvasia Sitges’ ScD10800 respectively. Moreover, the ScA111175 marker was amplified only in ’Folle blanche’ among the 64 cultivars tested with a large annealing temperature range using either two different Taq DNA polymerases or two separate thermocyclers. In addition, we discuss the initial polymorphism originated by the RAPD technique and suggest a new design of SCAR primers to obtain reliable cultivar-specific SCAR markers from single PCR-based bands for identification purposes.

Reliable identification of the quarantine root‐knot nematodes Meloidogyne chitwoodi and M. fallax by PCR‐based techniques*

The root‐knot nematodes Meloidogyne chitwoodi and M. fallax have been declared quarantine pests in Europe and there is a growing need for practical, reliable techniques to identify these nematodes and distinguish them from M. hapla. Such methods are needed for plant resistance management and research, for design of crop rotation systems and for inspection of harvested crops or soil samples. At the moment, several PCR‐based techniques have been developed, based on rDNA and SCAR primers. Some of these methods can be applied for extremely sensitive detection of a specific species, enabling the detection of 0.001 juvenile or less. These can also detect specific species present in mixtures in proportions of less than 1%. Others enable a precise estimation of the species composition in mixtures and can detect species present in mixtures in proportions as low as 5%. Depending on the question to be asked and the source of the sample to be studied, optimal diagnostic assays can be created by combining relevant methods.

558 Evaluation of Sequence-based PCR Markers in an F1 Pseudo-testcross and a BC1 Hybrid Citrus Population

Specific primers were designed for 61 cloned RAPD fragments and from 10 Citrus EST sequences for the production of SCAR, CAPS, and STS markers for a Citrus grandis `DPI 6-4' × Poncirus trifoliata `Rubideaux' F1 pseudo-testcross population. Fifteen SCAR, three CAPS, and one EST/STS markers were developed. An additional 17 SCAR and CAPS primer pairs developed at the Citrus Research and Education Center for a Citrus grandis `Thong Dee' × (Citrus grandis `Thong Dee' × Poncirus trifoliata `Pomeroy') BC1 population were screened in the pseudo-testcross population. A total of 27 markers were identified and scored in the pseudo-testcross population in which 24 were mapped; 13 in the Citrus parental linkage map on seven linkage groups and 11 in the Poncirus parental map on five linkage groups. In the BC1 population, 20 of 27 markers tested were found to be polymorphic and 13 mapped to seven of nine linkage groups. Of these, 11 were mapped in both populations and could be used for aligning presumed homologous regions on the three linkage maps.

Marker-assisted dissection of the oligogenic anthracnose resistance in the common bean cultivar, ‘G2333’

Two independently assorting dominant genes conditioning resistance to bean anthracnose were identified in an F2 population derived from the highly resistant bean differential cultivar, ‘G 2333’. One gene was allelic to the Co-4 gene in the differential cultivar ‘TO’ and was named Co-4 2 , whereas the second gene was assigned the temporary name Co-7 until a complete characterization with other known resistance genes can be conducted. Two RAPD markers linked to the Co-4 2 allele were identified. One RAPD, OAS13950, co-segregated with no recombinants in two segregating populations of 143 F2 individuals, whereas the second RAPD, OAL9740, mapped at 3.9 cM from the Co-4 2 allele. Two 24-mer SCAR primers (SAS13), developed from the OAS13950 RAPD marker, were dominant and polymorphic, similar to the original RAPD, and supported the tight linkage between the marker(s) and the Co-4 2 allele. The markers were present in germplasm with known resistance alleles at the Co-4 locus. The presence of the markers in two other differential cultivars not previously characterized and in four navy bean cultivars suggests the existence of a gene family for anthracnose resistance at or near the Co-4 locus. Since the Co-7 gene was present only in germplasm which also possessed the Co-4 2 and Co-5 genes, the SAS13 markers were used in combination with standard inoculation techniques to identify F3 lines in which the Co-7 gene was homozygous and the Co-4 2 allele was absent. A similar strategy of marker-assisted dissection is proposed to identify resistant lines in which the Co-5 gene is absent and the Co-7 gene is present by selecting against the OAB3450 marker, which has been shown previously to be linked to the Co-5 gene. These genes cannot be distinguished using traditional screening methods since all current races of the pathogen virulent to the Co-5 gene are avirulent to the Co-4 2 and Co-7 genes. We describe the use of molecular markers tightly linked to resistance genes to facilitate the identification of an uncharacterized resistance gene for which no discriminating race of the pathogen is known.

Development of a SCAR marker linked to the I gene in common bean.

Two 24-mer SCAR primers (SW13) were developed from a previously identified 10-mer RAPD primer (OW13(690)) linked to the I gene, which conditions resistance to bean common mosaic virus (BCMV) in common bean. Linkage between SW13 and the I gene was tested in three F2 populations segregating for both SW13 and the I gene: N84004/Michelite (1.0 +/- 0.7 cM), Seafarer/UI-114 (1.3 +/- 0.8 cM), and G91201/Alpine (5.0 +/- 2.2 cM). SW13 proved to be more specific and reproducible than the OW13(690) RAPD marker. Using different heat-stable DNA polymerases, SW13 amplified a single 690-bp fragment linked to the I gene that more consistently permitted the identification of resistant plants. In addition, the presence of the I gene was detected using SW13 in genotypes originating from different gene pools of Phaseolus vulgaris L., indicating a broad utility of this marker for bean breeding programs.

PCR-Based Method for Sexing Roseate Terns (Sterna dougallii)

A simple method for identifying the sex of sexually monomorphic birds would be widely applicable to studies of population-genetic structure, demography, and behavior. In this paper we describe a method for identifying W-chromosome-specific genetic markers and show that these markers can be used to identify sex from very small samples of tissue, such as single mature feathers and, thus, from individuals of any age. We have applied this technique to Roseate Terns (Sterna dougallii) and further shown that the marker correctly identified sex in a closely related species, the Common Tern (Sterna hirundo). This method can be used to develop markers for other groups of birds, or any organism with chromosomal sex determination. The Roseate Tern is a cosmopolitan species of seabird that occurs in widely separated breeding populations in temperate and tropical regions of the Atlantic, Indian, and West Pacific oceans (Cramp 1985). The population in the northwestern Atlantic, which is the subject of our study, is listed as endangered (U.S. Fish and Wildlife Service 1987). Preliminary observations indicate a female-biased sex ratio in this population (Hatch and Nisbet pers. obs.). Information about the sex of many more individuals at various life cycle stages will be important for characterizing sex-specific behavior and investigating this femalebiased population sex ratio. Like many terns and other seabirds, the Roseate Tern is sexually monomorphic in plumage, and the size differences between males and females are small and not well characterized. Sex-specific behaviors (copulation, courtship feeding) in this species are usually performed away from the nest site and only occasionally can be used to identify the sex of individual birds. Alternative methods for sexing include laparoscopy and handling the adults during the period of egg laying when morphological differences are distinct. However, both of these methods pose unacceptable risks of injury or nest abandonment for members of this endangered population. Another method for sexing birds based on analysis of tissue samples is karyotyping from fibroblasts grown from blood-feather pulp (Parker et al. 1991). This method is laborious and has not been used on terns. Molecular techniques for sexing used on several bird species have been based on hybridizing labeled probes to genomic DNA. The probes are either W-chromosome-specific or they identify multigene families with W-specific members (Griffiths and Holland 1990, Quinn et al. 1990, Dvorak et al. 1992, Griffiths 1992, Millar et al. 1992, Graves et al. 1993, Longmire et al. 1993). These molecular methods are slow and laborious, they require relatively large quantities of DNA, and most of them involve use of radioisotopes. Such methods are neither economical nor practical for field studies involving large sample sizes or nestlings. With the advent of random-amplified polymorphic DNA (RAPD; Williams et al. 1990), a new source of easily accessible genetic markers is available. These markers have been applied to genetic studies in plants (Tingey et al. 1992) and, recently, have been used for sexing birds (Griffiths and Tiwari 1993). We describe a method for sexing Roseate Terns based on the polymerase chain reaction (PCR) that offers considerable advantages in technical simplicity and time. The method was developed with blood samples from adult Roseate Terns. However, because only nanogram quantities of DNA are required, the method is much more versatile and works with single mature feathers. Genetic markers specific to the W chromosome are identified by separately pooling samples of male and female DNA, and then screening the two pools for RAPDs (Michelmore et al. 1991). A typical RAPD reaction yields 6 to 10 bands (loci) that are produced by amplification with one arbitrary 10-base primer. Once a W-linked RAPD marker has been identified, it can be converted to a more specific and reliable PCR-based marker (i.e. a sequence-characterized amplified region, SCAR; Paran and Michelmore 1992) by increasing the length of the original primer at each end of the targeted region. The high specificity of a SCAR marker overcomes several of the drawbacks of RAPDs (Kesseli et al. 1992, 1994, Paran and Michelmore 1992). Methods.-Roseate Terns initially were sexed using an RFLP probe (Zoogen Inc., Davis, California). The method was verified using field-sexed Common Terns. The Roseate Terns sexed by this method were used to identify RAPD markers and develop SCAR primers. Finally, these primers were tested on a larger sample of field-sexed Common Terns. Terns were trapped at nests on Bird Island, Marion, Massachusetts (41?40'N, 70?43'W). Roseate Terns were trapped shortly before their eggs hatched. Common Terns were trapped within 24 h after the first egg was