Ploidy manipulation of the gametophyte, endosperm and sporophyte in nature and for crop improvement: a tribute to Professor Stanley J. Peloquin (1921–2008)

Abstract

Background

Emeritus Campbell-Bascom Professor Stanley J. Peloquin was an internationally renowned plant geneticist and breeder who made exceptional contributions to the quantity, quality and sustainable supply of food for the world from his innovative and extensive scientific contributions. For five decades, Dr Peloquin merged basic research in plant reproduction, cytology, cytogenetics, genetics, potato (Solanum tuberosum) improvement and education at the University of Wisconsin-Madison. Successive advances across these five decades redefined scientific comprehension of reproductive variation, its genetic control, genetic effects, evolutionary impact and utility for breeding. In concert with the International Potato Center (CIP), he and others translated the advances into application, resulting in large benefits on food production worldwide, exemplifying the importance of integrated innovative university research and graduate education to meet domestic and international needs.

Scope

Dr Peloquin is known to plant breeders, geneticists, international agricultural economists and potato researchers for his enthusiastic and incisive contributions to genetic enhancement of potato using haploids, 2n gametes and wild Solanum species; for his pioneering work on potato cultivation through true seed; and as mentor of a new generation of plant breeders worldwide. The genetic enhancement of potato, the fourth most important food crop worldwide, benefited significantly from expanded germplasm utilization and advanced reproductive genetic knowledge, which he and co-workers, including many former students, systematically transformed into applied breeding methods. His research on plant sexual reproduction included subjects such as haploidization and polyploidization, self- and cross-incompatibility, cytoplasmic male sterility and restorer genes, gametophytic/sporophytic heterozygosity and male fertility, as well as endosperm dosages and seed development. By defining methods of half-tetrad analysis and new cytological techniques, he elucidated modes, mechanisms and genetic controls and effects of 2n gametes in Solanum. Ramifications extend to many other crops and plants, in both basic and applied sciences.

Achievements

Based upon a foundation of genetics, cytogenetics and plant reproductive biology, Dr Peloquin and co-workers developed methods to use 2n gametes and haploids for breeding, and used them to move genes for important horticultural traits from wild tuber-bearing Solanum species to cultivated potato for the betterment of agriculture. The resulting potato germplasm included combinations of yield, adaptation, quality and disease resistance traits that were previously unavailable. This elite plant germplasm was utilized and distributed to 85 countries by the CIP, because it not only increased potato yields and quality, it also broadened the adaptation of potato to lowland tropical regions, where humanity has benefited from this addition to their food supply.

INTRODUCTION

Writing about Emeritus Campbell-Bascom Professor Stanley J. Peloquin (Fig. 1) presents a significant challenge as his innovative science, mentoring role and impact on people spans generations and reaches across the world. Dr Peloquin was born in Barron, Wisconsin (USA), on 22 July, 1921 and died on 27 July, 2008 in Madison, Wisconsin. After World War II, he enrolled at Marquette University, Milwaukee, Wisconsin, and obtained an MSc degree in biology in 1948. He then enrolled at the University of Wisconsin, Madison, and studied genetics under the guidance of Professor R. A. Brink and D. C. Cooper and was awarded a PhD degree in genetics in 1952. His thesis was entitled ‘Abnormal embryo and endosperm development in Zea mays following the use of pollen that has been exposed to mustard gas’. From 1951 to 1956, he taught biology at Marquette University, and in 1957 joined the faculty at the University of Wisconsin, Madison, until his retirement in 1994. Dr Peloquin was a faculty member in the Laboratory of Genetics, but his majority appointment was in the Department of Horticulture – a field that he helped to advance although he had no formal training in the subject.

Fig. 1.

Emeritus Campbell-Bascom Professor Stanley J. Peloquin (1921–2008).

In 1984, Dr Peloquin was elected to the US National Academy of Sciences and was awarded an honorary life membership by the Potato Association of America. In 1983, he was named the Campbell-Bascom Professor by the University of Wisconsin and he received Laurea Honoris Causa from the Università degli Studi di Napoli ‘Federico II’ (Italy) in 2002. He was recognized for his innovative research and scientific leadership and they serve as an acknowledgment of his superb skills as a plant geneticist and potato breeder. This article reviews his achievements in science, education, potato improvement and production.

HAPLOIDS

Haploids are individuals having the same number of chromosome sets as a gamete, or half as many as a somatic cell. The scientific contributions of Dr Peloquin in plant genetics began with his research efforts, along with Professor R. W. Hougas, into potato breeding using haploids (Hougas et al., 1958; Peloquin and Hougas, 1958). They initially found a plant with a somatic chromosome number of 24 among progeny of a cross between the tetraploid (2n = 4x = 48) potato cultivar ‘Katahdin’ and the cultivated South American diploid (2n = 2x = 24) species Solanum phureja Juz. & Bukasov (Hougas and Peloquin, 1957). This plant resembled the female parent ‘Katahdin’ and lacked any of the distinctive characters of the S. phureja male parent. They concluded that this 24-chromosome plant was a haploid of ‘Katahdin’, and it arose through parthenogenesis. Following this finding, they developed methods to isolate thousands of potato haploids from diverse tetraploid genotypes (Hougas et al., 1964). These were among the first readily generated haploids in any crop, with numbers comparable with those generated for barley and maize (Kotch and Peloquin, 1987). With the method devised by Peloquin and Hougas, haploids were used to study genetic traits in potato with better resolution than is possible in typical cultivated 4x potato (Peloquin and Hougas, 1960). Haploid production was the first type of ploidy manipulation carried out by Dr Peloquin and co-workers and the first of his many discoveries resulting from research involving interspecific hybridization.

Decapitation of the pistillate parent was found to increase haploid frequency (Peloquin and Hougas, 1959). However, the ‘pollinator’ effect (Peloquin et al., 1996) is the main factor affecting the frequency of potato maternal haploids. Maternal haploids from 4x–2x crosses were associated with hexaploid (6x) endosperm; 2n pollen was not a requisite (Peloquin et al., 1996). This association of decreased embryo ploidy level with increased endosperm ploidy level suggested a common basis, specifically that the absence of egg cell fertilization was coincident with fertilization of the central cell (4x) by both monoploid (1x) sperm cells or, if pollen tube mitosis failed, by a single combined equivalent (2x). Subsequent development of a functional 6x primary endosperm would allow for parthenogenetic development of the associated virgin egg cell, if not induce it. This ‘pollinator’ effect is heritable (Ortiz et al., 1992b). Genetic markers such as pigmentation on seeds allow early identification and elimination of hybrids. Consequently, potential haploids among the progeny of cultivar × pollinator crosses are readily detected. Paternal haploids are also obtained via anther culture, but maternal haploids offer more advantages for potato breeding because paternal haploid production requires gene(s) for androgenic competence, which are not available in all tetraploid potato cultivars (Ortiz, 1998).

Potato haploids provide promising material for genetic studies (Werner and Peloquin, 1991a; Kotch et al., 1992; Ortiz et al., 1993a), cytogenetic analyses (Peloquin et al., 1989a) and germplasm enhancement of this important tuber crop (Hermundstad and Peloquin, 1985a). The crossability of haploids appears to be similar to that of the cultivated diploid species, and most haploid–diploid species hybrids are vigorous plants. Many of these hybrids produce large attractive tubers (Hermundstad and Peloquin, 1986) and are male-fertile (Hermundstad and Peloquin, 1985b). An effective method of incorporating germplasm of wild 2x Solanum species into the cultivated potato is through hybridization with haploids of 4x Solanum tuberosum. This procedure provides for the introduction of desirable traits and genetic diversity from wild species into potato (Jansky et al., 1990). Haploid and wild species genotypes can be evaluated for parental value based on the presence of good tuberization, desirable traits, fertility and 2n gametes in their hybrid progeny. Haploid–wild species hybrids may be intercrossed and evaluated at the 2x level to take advantage of diploid genetic ratios. Selected 2x genotypes may then be brought to the 4x level through sexual polyploidization using 2n gametes.

2n GAMETES

Professor Peloquin's second major contribution to ploidy manipulation was as the founder of potato breeding using 2n gametes (Peloquin et al., 1989b). In the 1950s, large-scale testing of various parents for interploid 4x–2x cross fertility revealed that certain parents consistently departed from the expected failure. Although few or no triploid progeny resulted, many tetraploid progeny were obtained from specific parental combinations, and the numbers were heavily dependent on the 2x genotype (Hanneman and Peloquin, 1967). This discovery led to diverse lines of innovative basic and applied research, including reproductive biology, cytology, cytogenetics, inheritance, quantitative genetics, population genetics, evolution, experimental breeding, applied breeding and agricultural production. They provided new insights into the basic biological mechanisms by which plants generate 2n gametes (Peloquin et al., 1999) and generate polyploids in nature. Science now recognizes 2n gametes as central to the evolution of potatoes and most other polyploidy crop species (Harlan and de Wet, 1975; den Nijs and Peloquin, 1977; Veilleux, 1985; McCoy and Echt, 1992; Ortiz et al., 1992b; Bretagnolle and Thompson, 1995; Katsiotis and Forsberg, 1995; Ortiz, 1997; Ramsey and Schemske, 1998; Hancock, 2005; Heslop-Harrison and Schwarzacher, 2007).

Differences among diploid parents for the ability to produce 4x progeny from 4x–2x crosses was shown to be heritable and resulted from their production of various percentages of diploid pollen, as opposed to haploid (monoploid) pollen. To distinguish categorically the diploid pollen from haploid pollen, the former were dubbed 2n pollen. The definition recognized that the ploidy levels of 2n gametophytes would depend on parental ploidy levels, as do the ploidy levels of haploids. Thus, given a diploid parental plant, haploid pollen would be monoploid and 2n pollen would be diploid; given a tetraploid parent, haploids would be diploid, and 2n pollen would be tetraploid.

Early research on the inheritance of 4x–2x crossability led to the definition of three single-gene mutants, and to the development of an important conceptual distinction between mechanisms and modes of 2n gamete formation. The most important of these, the parallel spindles (ps) mutant, affects relative orientations of meiosis-II spindles in the coenocytic microsporocytes, generates meiotic products genetically equivalent to first division restitution (FDR) gametes, and is under control of the Mendelian recessive ps (Mok and Peloquin, 1975a). The high ps frequency in 4x potato cultivars supports the hypothesis that 2n pollen was involved in the origin of the cultivated 4x gene pool, as a higher ps frequency in modern tetraploid cultivars (0·6–0·82) than in the ancestral 2x population (0·3–0·47) can be expected if they arose from sexual polyploidization but not if they were produced by somatic doubling (Iwanaga and Peloquin, 1982). Mok and Peloquin (1975b) also described two less frequent mechanisms, premature cytokinesis-1 and -2, which are under the genetic control of recessive non-allelic pc-1 and pc-2 mutants and lead to second division restitution (SDR) 2n pollen. Parallel spindles at anaphase II is the most frequent mechanism for producing 2n pollen in potato and in other tuber-bearing Solanum species (Watanabe and Peloquin, 1989, 1992).

The findings of Mok and Peloquin (1975a, b) underscored the fact that 2n gamete populations are defined not only by their parental genes but also by the mechanisms by which they form. Moreover, those mechanisms can be grouped according to the genetic expectations, e.g. FDR versus SDR mechanisms. These differences in genetic ramifications underlined the need for being specific when describing 2n gametes. Professor Peloquin recognized that the term ‘unreduced’ has two connotations, one numerical and one genetic, and so he avoided the term to describe any type of 2n pollen or gamete. Unfortunately, not all researchers realized this fact and its significance, and thus applied it too liberally. Likewise, many scientists, including geneticists, fail to recognize that genetic reduction normally occurs in both the ‘reductional’ and the ‘equational’ divisions of meiosis. Fortunately, these terms have become nearly obsolete.

For 4x–2x crosses, surveys of numerous wild and artificial diploid genotypes for 2x–4x cross fertility revealed variability among diploids for 4x progeny production (Hanneman and Peloquin, 1968). The differences were largely attributable to differing frequencies of functional 2n megagametophytes, and intercrosses initially hinted at simple inheritance, but also to a multiplicity of underlying mechanisms and genes. The biological and genetic dissection of 2x–4x crossability was much slower than that of 4x–2x crossability, due to the much greater difficulty of detailed cytological characterization, which for megagametogenesis conventionally relied on microtome sections. However, Iwanaga and Peloquin (1979) established FDR 2n megagametophyte production by a sex-specific synaptic mutant. Progress quickened following a robust stain-clearing method was developed to facilitate and expedite cytological analysis of megagametogenesis (Stelly et al., 1984). This enabled direct cytological screening of mature megagametophyte populations to determine 2n megaspore frequencies, and cytological analysis of megasporocyte and megagametophyte development to delineate multiple mechanisms of formation (Stelly and Peloquin, 1985, 1986b). Whereas 2n pollen arises predominantly by way of parallel spindles (FDR), 2n megagametophytes seem to arise most commonly by omission of the second division (SDR) (Stelly and Peloquin, 1986a). It is controlled by the recessive meiotic mutant (os) in diploid potato (Werner and Peloquin, 1990). The os gene frequency was found to vary from 0·28 to 0·76 among 2x wild Solanum species (Ortiz and Peloquin, 1991a). Other mechanisms of 2n megagametophyte formation in potato are failure of cytokinesis at telophase II followed by nuclear restitution, and synaptic abnormalities followed by nuclear restitution and a normal second division, which are genetically equivalent to SDR and FDR, respectively (Werner and Peloquin 1987, 1991b).

DIPLOID TUBER-BEARING SOLANUM SPECIES AND POTATO GENETIC ENHANCEMENT

Professor Peloquin viewed the utilization of germplasm resources in breeding programmes as one of the greatest opportunities in modern agriculture, and one of its greatest challenges. Diploid wild species and cultivars of Solanum constitute a vast genetic reservoir of genetic diversity in potato breeding (Jansky and Peloquin, 2006). However, ploidy and other naturally occurring barriers markedly hinder gene transfer from the diploids into agriculturally suitable types. Dr Peloquin was a leading force in circumventing these barriers. A key strategy was systematically to screen parts of the exotic 2x potato gene pool for 4x–2x or 2x–4x crossability and 2n gametes (2n pollen). The facile generation of tetraploid progeny provided quick avenues for hybridization and introgression, but many of the immediate F₁ hybrids were outstandingly impressive in production capacity, but ill-suited in terms of maturity, tuber type and certain other traits related to domesticated production. Back-crossing of certain hybrids to cultivars was shown to mitigate significantly the agricultural shortcomings of original 4x–2x hybrids. The resulting hybrid populations were of such vigour and disease resistance that they served as an important impetus for development of ‘true-potato seed’ (TPS) in tropical areas. Moreover, transmission of most potato viruses is rare via true botanical seed, but common through tubers, so the novel genotypes were also relatively free of pathogenic viruses. These remarkable benefits contributed significantly to the revolutionary practice of producing potatoes from TPS, especially in tropical climates.

Efforts to use diploids in the synthesis of F₁ hybrids suitable for traditional clonal production led to the discovery that agriculturally more suitable types were derived more expeditiously by crossing the exotic 2x plants to haploids of potato cultivars. The haploids tuberize under field conditions in temperate regions, facilitating both evaluation and maintenance. These attributes greatly complemented shortcomings of the diploid species. Most cultivar haploids were found to be pollen-sterile, relegating them to use as female parents, and the exotic materials as pollen parents to produce ‘haploid-species hybrids’. Many of these haploid-species hybrids were discovered to produce 2n gametes, which greatly facilitated subsequent transfer of the diploid germplasm into 4x cultigens. Collectively, these advances established that diploid species germplasm could be used to create female-fertile 2n gamete-producing haploid-species hybrids with acceptable tuber characteristics and other desired attributes, such as host plant resistance to pathogens or pests; these haploid-species hybrids provided a conduit to synthesis of 4x offspring valuable as clones and parents. In short, these efforts establish new breeding schemes based on sexual polyploidization.

PLOIDY MANIPULATIONS: TERMINOLOGY AND THEORY

As a perceptive educator and scientist, Dr Peloquin was astutely aware of the inadequacies of a number of terms widely used to describe meiosis, polyploid formation, constitution, breeding and genetics. The desired accuracy could be achieved only if the terminology were linked to cytogenetic behaviour, and such accuracy was essential to discussions of genetics and the breeding of polyploid crops (Carputo et al., 2006). Cognizant that scientific vocabulary, once established, exhibits considerable inertia, he understood that adoption by the broader scientific community of more accurate and specialized terms would occur only with increased comprehension and common usage. He endeavoured to provide both, and vociferously encouraged others to do likewise. He regularly used enhanced terminology in daily communications with students, research assistants and peers, as well as in the laboratory's presentations and publications. Moreover, he was quick to challenge inappropriate usage of the older terms.

Perhaps his most challenging terminology endeavour was to improve usage of the deeply entrenched terms ‘allopolyploidy’ and ‘autopolyploidy’, and in many cases replace them with terms ‘disomic’ and ‘polysomic’ polyploidy. The underlying aim was to establish terminology that emphasized meiotic behaviour and genetic consequences, rather than ancestral origins and cytogenetic composition. Allopolyploidy and autopolyploidy allude to cytogenetic constitution and ancestry, but there is no measure by which to distinguish between them. If their definitions are based on interspecific versus intraspecific ancestry, then they are subject to the basis upon which taxonomists choose to define separate species. Moreover, genomic constitution does not determine behaviour. For example, the pattern of inheritance followed by a newly formed interspecific polyploid hybrid, even if an allopolyploid, might be disomic, polysomic, intermediate or highly distorted. This lack of a consistent relationship between constitution and genetic behaviour was underlined by the discovery in wheat and then other species that single chromosomes and single genes could shift the meiotic pairing and inheritance from disomic to polysomic in at least some ‘allopolyploids’ (Dvorak et al., 2006). The terms disomic and polysomic polyploidy relate directly to patterns of meiotic chromosome pairing and subsequent genetic ratios, which are readily quantified and highly pertinent to genetic manipulation. Their accountability, relevance to functional ramifications, and instructive focus on disomic versus polysomic patterns of meiotic pairing and inheritance made them preferable for teaching and scientific presentations concerned with cytogenetics, genetics and breeding.

Mendiburu and Peloquin (1976) defined sexual polyploidization and depolyploidization (haploidization) as the processes through which euploid zygotes are formed and whose chromosome numbers are respectively increased above or decreased below the ploidy level that would be expected if each parent contributed a gamete carrying half the parental pre-meiotic (somatic) chromosome number. These definitions helped to communicate the existence of plant reproductive variations and their significance as evolutionary and in vivo breeding mechanisms. Sexual polyploidization leads to massive genetic innovation, which should be regarded as its most significant difference from somatic doubling.

Dr Peloquin and co-workers devised genetic enhancement methods that take advantage of ploidy manipulations, i.e. scaling up and scaling down chromosome sets. Ploidy manipulations with S. tuberosum haploids, 2n gametes and wild species remain as one of the most impressive and exciting crop germplasm enhancement methods ensuing from cytogenetics research (Ortiz et al., 2005). Such ploidy manipulations are easily achieved in potato, where they enable the transfer of genes from wild Solanum species to the primary crop gene pool, particularly alleles for improving horticultural traits. In the breeding scheme, hybridization between cultivar haploids and wild species captures the genetic diversity from 2x species. Many of the 2x progeny tuberize and are thus easily screened and maintained as clones. Their 2n gametes transmit much or all of the diversity to 4x progeny through unilateral or bilateral sexual polyploidization. Unilateral events involve either 2n pollen (4x × 2x crosses), or 2n megagametopytes (2x × 4x crosses), whereas bilateral events involve both (2x × 2x crosses). Most progeny from unilateral sexual polyploidization matings in potato are tetraploids, whereas those from bilateral sexual polyploidization usually include diploids and tetraploids, the frequencies being dependent on the relative frequencies of 2n gametophytes. In potato (and many other species) 3x hybrids from 4x × 2x, 2x × 4x, or 2x × 2x crosses are very rare (Hanneman and Peloquin, 1968), which makes this method of gene transfer both highly effective and quite efficient. Moreover, tetraploid progeny can be visually selected by their larger size and increased vigour relative to diploid sibs.

Relative levels of performance among some tetraploids from sexual polyploidization were stunningly superior to cultivars, while those of others were not. These differences were related to their origin, and particularly to the underlying mode of 2n gamete formation behind sexual polyploidization. The observations corresponded to the principle that a high level of heterozygosity is directly related to heterosis for crop yield in polysomic polyploids. The breeding of polyploid cultivars for polygenic traits where non-additive effects are important should generally aim at maximizing heterozygosity due to allelic diversity. In potato, FDR 2n gametes transmit on average 80 % of the heterozygosity of the 2x parent to the 4x hybrid offspring whereas SDR 2n gametes transfer on average less than 40 % of the 2x heterozygosity to 4x hybrids. In this regard, the better diploid parents for 4x × 2x crosses would be those producing FDR 2n gametes, because hybrid vigour associated with high yields may be maximized by multiple alleles per locus in potato. In his last journal article, Dr Peloquin estimated that FDR gametes were more than twice as effective as SDR gametes in transmission of heterozygosity after estimating the transmission of heterozygosity on an individual chromosome basis through pachytene analysis (Peloquin et al., 2008). Their cytological models rely on relative chromosome length, euchromatin and heterochromatin content, centromere location, and observations of usually one chiasma per bivalent from early research in his laboratory (Yeh et al., 1965).

SELF-INCOMPATIBILITY AND CROSS-INCOMPATIBILITY

Gametophytic self-incompatibility controlled by multi-allelic S genes may hamper inbreeding (either through selfing or sib-mating) in 2x potato species (Cipar et al., 1964a). There may be some self-compatible 2x variants (Cipar et al., 1964b), but they can seldom be selfed over more than a few generations (i.e. beyond S₅) due to inbreeding depression. However, one of Dr Peloquin's former students, Dr Robert E Hanneman Jr, identified and characterized a dominant self-compatibility locus in a 2x wild potato species (Hosaka and Hanneman, 1998a, b). This valuable genetic resource can be useful for the development of inbred lines for breeding and genetics research.

Dr Peloquin undertook further self-incompatibility research with his graduate students in other plant species such as Lilium longiflorum (Thumb.) or Easter lily (Ascher and Peloquin, 1968). They chose this species because it possesses several advantages for investigating the nature of self-incompatibility; for example, the large hollow styles permit analyses of proteins and enzymes from pollen tubes removed from the styles after selfing or crossing along with their corresponding styles, as well as from pollen grains and from styles of non-pollinated flowers (Desborough and Peloquin, 1968a). Research by Dr Peloquin's students included the influence of age and temperature on incompatible and compatible pollen tube growth in Easter lily (Ascher and Peloquin, 1966a, b, 1970). They found inactivation of self-incompatibility following temperature pretreatments of styles (Hopper et al., 1967; Hopper and Peloquin, 1976), and X-ray inactivation of the stylar component of the self-incompatibility reaction (Hopper and Peloquin, 1968).

Tuber-bearing Solanum species are separated by geographical and ecological barriers throughout the American continent. Nonetheless, there are some sympatric species that share the same niches but do not readily cross. For example, 4x S. gourlayi Hawkes and 6x S. oplocense Hawkes from Argentina are difficult to hybridize due to pollen–pistil incompatibility on the stigma, in the first third of the style and in the first two-thirds of the style (Camadro and Peloquin, 1991). Cross-incompatibility genes in pistils and dominant specific complementary genes in pollen grains appear to control such interactions. Likewise, Camadro et al. (2004) indicated that other major forces strengthening external hybridization barriers are nuclear-cytoplasmic male sterility, and the endosperm.

MALE FERTILITY, CYTOPLASMIC MALE STERILITY AND RESTORER GENES

Low male fertility and seed set is common in cultivar haploids and in offspring from inter-haploid matings (Ross et al., 1964). Likewise, male sterility may occur in hybrids from crosses between S. tuberosum haploids and either 2x cultivated relatives or some wild Solanum species (Hanneman and Peloquin, 1981; Hermundstad and Peloquin, 1985b). Often, male sterility in interspecific hybrids results from the interaction of sensitive S. tuberosum cytoplasm with the nuclear dominant gene Ms from some 2x species.

Variation in male fertility among 4x–2x families resided in the 4x cultivar female parents, because the FDR 2n pollen-producing male parents had the same genotype (Ms/ms) for the Ms locus (Iwanaga et al., 1991). The Ms gene was deduced to be located far from the centromere. A male fertility restorer (Rt) gene also occurs among cultivars and its frequency was estimated to be 0·20 in Tuberosum Group clones. Ortiz et al. (1993b) found that tetrad sterility associated with lack of microspore release from sporads could result also from interactions between S. stoloniferum Schltdl. & Bouché cytoplasm and male sterility genes. In contrast, Andigena Group and S. demissum Lindley cytoplasm did not interact with the Ms gene to produce male sterility in potato.

This knowledge about male sterility and restorer of fertility loci provides an opportunity to circumvent genetic–cytoplasmic male sterility by conventional methods. For example, haploids from 4x Tuberosum Group cultivars carrying both Rt and ps alleles could be crossed with 2x cultivars or wild species producing FDR 2n pollen and carrying the Ms gene. Some of their derived haploid-species hybrids will be male-fertile and FDR 2n pollen producers and can be used for the incorporation of 2x germplasm with desirable attributes to the 4x level by unilateral sexual polyploidization via 4x × 2x crosses.

Dr Peloquin expanded his interest in cytoplasm–nuclear interactions to maize (Zea mays L.) through the research of one of his students. Ellsworth and Peloquin investigated the influence of the cytoplasm on ear number expression on the main stalk in the progeny of five maize inbred lines (Ellsworth and Peloquin, 1972). Their research indicates that maize breeders interested in increasing ear number per plant could improve success rates by screening for beneficial cytoplasmic–genotypic interactions.

GAMETOPHYTIC/SPOROPHYTIC HETEROZYGOSITY AND MALE FERTILITY

Tetrasomic plant species such as 4x potato routinely produce gametophytes that have two copies of each gene and are thus heterozygous or homozygous for every locus. As a result, gametophytic heterosis may occur. Levels of heterozygosity in the tetrasomic 4x sporophyte as well as the proportions of heterozygosity that are transmitted may have an effect of pollen vigour and viability. In addition, solanaceous self-incompatibility systems are rendered ineffective by heterozygosity of the SI locus.

Simon and Peloquin (1976) found that FDR 2n pollen was more vigorous than SDR 2n pollen and haploid (n) pollen, both before and after storage. They demonstrated that 2n gametes per se were not necessarily advantageous, as no superiority for vigour was found in SDR 2n pollen compared with n pollen. Further research by Ortiz and Peloquin (1994), using 4x hybrid offspring from 4x × 2x and 4x × 4x crosses free of cytoplasmic–genetic male sterility showed FDR 2n pollen-producing 2x parents produced hybrid offspring with significantly higher pollen stainability than those derived from 4x × 4x crosses. The superiority of such pollen populations seems probably to be due to a higher level of heterozygosity in the sporophyte, a proportion of which would be transmitted to haploid pollen, or to vigour of the parent and sporophytic effects on pollen. As argued by Stephenson et al. (2001), pollen depends upon the sporophyte for the resources necessary to develop, germinate and initiate tube growth, so the level of heterozygosity in the pollen-producing parent could affect pollen performance by affecting the ability of the sporophyte to provision its pollen. As already mentioned, with FDR 2n pollen, 80 % of the heterozygosity and a large fraction of epistasis of the 2x parent can be transmitted to its 4x hybrid offspring in potato. Such findings are important for the production of open-pollinated (OP) TPS because pollen stainability is significantly correlated with fruit set in potato.

ENDOSPERM BALANCE NUMBER

The endosperm is a unique feature of angiosperms and clearly reflects a reproductive strategy with immense evolutionary ramifications. Endosperm arises after double fertilization, where one of the two male gametes delivered by a pollen tube unites with the egg to form the zygote, and the other unites with the central cell to form the endosperm. Normal seed development in vivo is contingent on endosperm. Clearly, this specialized tissue has played a significant role in the evolution of angiosperms because of its physiological and genetic relationships to the embryo and overall seed development. One manifestation of this close developmental relationship is its abnormal development, and consequent seed inviability, that normally follows interploidy crosses. In many angiosperm taxa, interploid 4x × 2x and 2x × 4x crosses fail to produce viable triploid progeny, i.e. a ‘triploid block’ (Marks, 1966).

In contrast to other researchers at the University of Wisconsin-Madison who had attempted to define factors governing angiosperm reproductive controls in relationship to ploidy, Dr Peloquin and his former student Dr Robert E. Hanneman Jr used potato and its relatives to develop a distinctly different approach. They and their students delved into the factors governing viable endosperm and seed production as related to ploidy level when shifts in chromosome number result from 2n gametes or sporophyte ploidy level. These studies led to their co-founding of the endosperm balance number (EBN) concept (Johnston et al., 1980). Conceptually, this EBN hypothesis placed control over endosperm function in many plant species at a level beyond ploidy level per se, and beyond that of maternal versus paternal genome dosage. These principles led to significant practical benefits, by enabling relatively reliable predictions on the limits of and strategies for intercrossing potatoes from diverse ploidy levels (Ehlenfeldt and Ortiz, 1995).

According to the EBN concept, sustained endosperm development occurs only when two EBN factors from the female parent unite with one EBN factor from the male parent. Any deviations from the 2 : 1 ratio of maternal–paternal EBN factors lead to faulty endosperm and inviable seed. For example, in 4x × 2x crosses and 2x × 4x crosses, endosperm development is regularly abnormal as the female–male ratios are 4 : 1 and 1 : 1, respectively. However, if 2n gametes function in the 2x parent of 4x × 2x or 2x × 4x crosses, then the ratio is 2 : 1 and endosperm development is normal.

The genetic basis of EBN has yet to be elucidated. However, it appears that EBN is controlled by a few genes rather than the whole genome (Ehlenfeldt and Hanneman, 1988; Camadro and Masuelli, 1995). Crossing behaviour in Solanum species further suggests that the system is multi-genic and regulates both interspecific and intraspecific crosses. Endosperm dosage systems therefore explain many aspects of species evolution, but the system appears to have originated as an ancient means of promoting the fidelity of ploidy levels for each species, or at least the avoidance of sexual sterility associated with triploidy. Interestingly, 2x Solanum species with one EBN that do not form 2n gametes are sexually isolated from 2x two-EBN species. EBN-based isolation may be artificially overcome for breeding purposes by scaling the ploidy levels up or down or using 2n gametophytes (Carputo et al., 1997). For example, 3x F₁ hybrids [derived from an in vitro doubled clone of a 2x one-EBN (2n = 3x, two-EBN) and 2x two-EBN clones] can be successfully used in crosses to tetraploid cultivated potato, resulting in 5x or near 5x BC₁ progenies. Selected 5x BC₁ clones may be further back-crossed both as male and as female parents with 4x potato cultivars.

ISOZYMES, HALF-TETRAD ANALYSIS AND CYTO-TECHNIQUES

Acid gel electrophoresis of tuber proteins and esterase isozymes from Solanum tubers were used for potato cultivar identification and genetic research (Desborough and Peloquin, 1966, 1967, 1968b, 1969a, b). Further research by Simon and Peloquin (1980) elucidated the inheritance of electrophoretic variants of tuber proteins in S. tuberosum haploids. Desborough and Peloquin (1968a) electrophoretically characterized proteins and enzymes of styles, pollen and pollen tubes of self-incompatible cultivars of Easter lily, but did not associate any proteins with self-incompatibility.

The centromere plays a very important role in delivering the chromosome complement to the daughter cells at cell division. Although it is highly desirable to map the positions of centromeres relative to genes by linkage analysis, there are very few methods to do so. Half-tetrad analysis of 4x × 2x offspring can be used for gene–centromere mapping of loci that are heterozygous in the 2x parent (Mendiburu and Peloquin, 1979). Two of the four product strands from each meiotic bivalent in the 2x parent are transmitted via the 2n gamete to the 4x progeny, thus providing a convenient sample of half of the meiotic ‘tetrad’ for genotyping. The nulliplex (aaaa) 4x individuals that result from a cross between a nulliplex 4x parent and a heterozygous (Aa) FDR 2n gamete-producing 2x parent allows one to estimate the gene–centromere map distance. Following this method, Mok et al. (1976) were able to map the self-incompatibility locus (S) very close to the centromere. Additional research using half-tetrad analysis led to gene–centromere mapping of isozyme loci (Douches and Quiros, 1987), and host plant resistance genes to cyst nematode (H₁) Potato Virus X (Rx) and Potato Virus Y (Ry) (Wagenvoort and Zimnoch-Guzowska, 1992). Ortiz and Peloquin (1993) added non-nulliplex 4x parents for gene–centromore mapping of the flower pigmentation locus in potato. Comparative gene–centromere mapping enabled an estimate of meiotic recombination in normal synaptic and desynaptic 2x potatoes producing 2n pollen, and revealed desynapsis could reduce crossing-over by 73 % (Bastiaanssen et al., 1996). They suggested that half-tetrad analysis may be a promising additional approach because accurate centromere positions in relation to other loci were not available in the saturated genetic maps of potato. Further research by Park et al. (2007) showed that in potato only one crossover occurred per chromosome arm, and proved that strong recombination interference occurs between the centromere and the telomere. Conversely, Stelly and Peloquin (1986a), Werner et al. (1992) and Bastiaanssen et al. (1998) used half-tetrad analysis with previously mapped loci relative to their centromeres to determine the modes of 2n megagametophyte formation in potato. These analyses exemplify the tripartite relationship among genotype, mode of 2n gamete formation and gene–centromere distance.

Dr Peloquin and co-workers also developed several techniques for cytological and cytogenetic research that are also useful for classroom teaching. Among these are methods for germinating potato pollen on artificial media (Mortenson et al., 1964), a modified Giesma staining technique for identifying somatic chromosomes of 2x potatoes (Mok et al., 1974), and a stain-clearing technique for observation within whole ovules of potato and thus populations of megasporocytes and megagametophytes (Stelly and Peloquin, 1985; Stelly et al., 1986b).

THE POTATO: HIS MODEL CROP FOR CHROMOSOME MANIPULATIONS AND GENETIC ENHANCEMENT

Dr Peloquin orchestrated numerous demonstrations that 2n gametes and EBN have contributed intrinsically to the origins and extreme genetic diversity in Solanum at multiple ploidy levels. Their implicit evolutionary roles in nature underlined their potential utility in genetic improvement of potato. Dr Peloquin thus developed an experimental breeding programme that he ran in parallel with a more-or-less conventional one. In the experimental programme, Dr Peloquin, co-workers and students hoped to incorporate wild potato germplasm into agriculturally elite types of potato, using 2n gametes, haploids and EBN to improve the agricultural performance of potatoes. The conventional programme was successful in its own right, and provided a ‘measuring stick’ for the experimental programme. The two breeding programmes were superb teaching tools through which students gained indelible experience.

Dr Peloquin and his students elucidated the mechanisms of ploidy manipulations and developed new breeding methods based on such knowledge. They stated that sexual polyploidization results in greater genetic variability, fitness and heterozygosity than does somatic doubling and that the EBN system acts as a screen for either n or 2n gametes, depending on the EBN of parental clones (Carputo et al., 2003). Indeed, in nature, EBN in combination with 2n gametes ensures ploidy integrity of 2x ancestral species while providing the flexibility for either unilateral or bilateral sexual polyploidization.

The superiority of progeny arising from sexual polyploidization involving FDR 2n gametes was revealed by Kidane-Mariam and Peloquin (1974) and, in particular, Mok and Peloquin (1975c). Relative to SDR 2n gametes from diploids and haploid (n) gametes from tetraploids, FDR 2n gametes from diploids were superior in breeding value and were more homogeneous; SDR 2n gametes could be superior to n gametes of 4x parents, but were as heterogeneous. Replicated yield trials that were used to contrast clonal families created with vs. without the involvement of 2n gametes clearly established the significance of FDR 2n pollen in applied potato breeding and that non-additive gene action contributes significantly to tuber yield (Mendiburu and Peloquin, 1977a). Further research showed that heterozygosity, epistasis and genetic diversity, if present in the parental 2x genotype, are largely maintained in 2n gametes formed by FDR, which accounts for the extremely heterotic responses that can occur upon sexual polyploidization (Mendiburu and Peloquin, 1977b).

The discovery of high general combining ability in 4x × 2x crosses for 2x clones that produce FDR 2n pollen contrasted sharply with the breeding behaviour in conventional 4x × 4x crosses, where epistasis and specific combining ability are predominant. This raised the question of whether general combining ability could be selected among elite 4x clones, and perhaps even improved by breeding. Although such improvements would potentially enhance breeding of conventional potato cultivars, it was clear that they would be especially valuable in the synthesis of parents for high-performing F₁ seed as TPS cultivars, for which FDR 2n pollen producers had demonstrable promise. Dr Peloquin and co-workers devised an innovative approach that minimized genetic variation among 2n pollen grains, such that the performance of the progeny means and dispersion from 4x × 2x crosses reflected the gametic input from the respective 4x parent, only.

To create uniform populations of 2n pollen, the Peloquin laboratory identified diploid clones that combined a mechanism for FDR 2n pollen formation, conferred by the ps mutation, with absence or reduction of recombination, conferred by a mutant affecting homologous chromosome synapsis and recombination (Okwuagwu and Peloquin, 1981). Nearly all stainable pollen from the double mutants is FDR-NCO 2n pollen, i.e. FDR 2n pollen with no crossing-over (FDR-NCO), or at least reduced crossing over. These diploid clones provided ideal testers to assess the breeding value of elite 4x cultivars, and to estimate the relative contributions of random meiotic products from 4x parents to quantitative traits (Buso et al., 1999a, 2000).

In most instances, FDR-NCO 2n gametes are expected to be genetically superior to conventional FDR 2n gametes, i.e. with crossing over (FDR-CO), because they transmit to the progeny, without disruption by recombination, almost 100 % of the parental heterozygosity. FDR-CO gametes transfer approx. 80 % of the parental heterozygosity and a large fraction of the epistatic interactions.

FDR gametes from a given parent are expectedly equivalent for genes residing between centromeres and proximal crossover sites (Buso et al., 1999c). Research on 4x × 2x progeny from FDR 2n pollen showed that most loci controlling yielding ability must be between their respective centromeres and proximal chiasmata. Quantitative trait loci (QTL) in these low-recombination chromosome regions may be relatively intractable to conventional 4x × 4x crosses, which may account for the slow progress in increasing total tuber yield. Conversely, high heterosis for yield has been observed after incorporating genetic diversity through 4x × 2x crosses, and shown to be transmissible via FDR 2n gametes.

Peloquin et al. (2009) coined the term ‘horizontal linkage’ to describe the intact transmission of blocks of all genes between the centromere and the first crossover to offspring via FDR 2n gametes. His research provided support for the concept that horizontal linkage provides a powerful mechanism to exploit non-additive genetic variation, especially for potato yield. Because FDR 2n gametes effectively transmit dominant and epistatic interactions to offspring, producing uniform, high-yielding offspring, they are especially valuable to potato breeders. In addition, horizontal linkage is likely to play an important role in the evolution of polyploid species.

In potato, progeny testing is an imperative step for the selection of parental clones at any ploidy level because the parent–offspring correlation is low for tuber yield and other important traits, i.e. parents per se cannot provide a reliable prediction about the performance of their hybrid offspring (Buso et al., 2003). Both Ortiz et al. (1991b) and Buso et al. (1999b) showed the advantages of unilateral sexual polyploidization for multi-trait selection and progeny testing in potato breeding, e.g. fewer replications and locations are required to evaluate tuber yield in 4x × 2x than in 4x × 4x hybrid offspring. Likewise, research by Darmo and Peloquin (1991) and Buso et al. (2002) demonstrated the feasibility of the 4x × 2x breeding approach for producing outstanding hybrid offspring and to select promising individuals therein, highlighting this method for the genetic enhancement of the potato as a viable alternative to the conventional 4x × 4x crosses in both tropical and temperate areas of the world. However, as pointed out by Ortiz et al. (1991a) selection of parents must be done in each location in which the final product, either a family or a clone, will be used for further breeding or by farmers, respectively.

The concepts and methods developed by Dr Peloquin involving haploid breeding, 2n gametes and EBN have been extended to cassava, sweet potato, banana, alfalfa, other forages, berry fruits and other plants of economic importance (Ortiz, 2003). In potato, Dr Peloquin used them to breed not only clonal potato cultivars, but also true-seed potato cultivars suited to world regions with pests and pathogens that otherwise preclude production. His clonally propagated 4x cultivars and even more so the potato cultivars from true seed have been revolutionary in the adaptation of potato to production of this key crop in temperate and tropical regions.

POTATO GENETICS AND CONVENTIONAL BREEDING FOCUSING ON END-USERS

Dr Peloquin and his students often focused on end-user traits as the topic for their potato genetic research. They elucidated the inheritance of potato crisp colour – an important trait in determining the quality and acceptability of cultivars for processing (Thill and Peloquin, 1994), tuberization under long day-lengths (Jansky et al., 2004) or tuber greening – a problem in retail markets (Parfitt and Peloquin, 1981). Epistasis appears to account for most of the genetic variation of the above tuber traits. His breeding research with his students included new potato uses, e.g. vines for silage (Parfitt et al., 1982), or traits for human nutrition, e.g. ascorbic acid (Davies et al., 2002), which could improve tuber resistance to oxidative damage during chilling or cold storage.

Dr Peloquin together with his colleague Don Kichefski of the University of Wisconsin potato research farm at Rhinelander provided new high-yield 4x potato cultivars propagated by tuber division and adapted for fresh market and chipping, which have been grown on many hectares in North America. ‘Snowden’, for example, has been grown on 100 000 ha and has served as an important source of potato processing quality by potato breeders around the world. Potato-bred germplasm by Dr Peloquin's programme is among the most productive throughout the world, as shown by releases in the tropics of the high-yielding 4x × 2x derived DTO-2, DTO-28 and DTO-33 by Centro Internacional de la Papa (CIP, Lima, Peru) in the 1980s, and of the conventionally bred 4x cultivars ‘Wischip’ (1973), ‘Rhinered’ (1976), ‘Oneida’ (1977), ‘Langlade’ (1984), ‘Snowden’ (1990), ‘Niska’ (1991), ‘AC Glacier Chip’ (2001), ‘Red Companion’ (2002), ‘Red Pearl’ (2004), ‘Millennium Russet’ (2005), ‘White Pearl’ (2006) and Megachip (2007) in North America.

TRUE POTATO SEED

A significant applied aspect of Dr Peloquin's achievements in potato improvement is his development of methods to propagate potatoes reliably from true seed (TPS) rather than tubers. TPS has been suggested as an alternative to the production of the crop from tubers since at least the 1940s but Dr Peloquin's perceptive and very valuable application of his seminal discoveries on potato haploids and 2n gametes improved the uniformity and yield of the potato crop from TPS, resulting in a reliable and sustainable potato crop developing regions of the world that are plagued by pests and lack refrigerated tuber storage facilities. In cooperation with the International Potato Center (CIP), this research made the production of virus-free potatoes possible at an affordable cost for poor farmers in the tropical developing world where the production of potatoes in the past had only been possible with prohibitively expensive imported tubers.

TPS has provided a means for extending the geographical range of potato production throughout the world. The production of potatoes from true seed has increased dramatically in areas of India, Bangladesh and China where they are grown between two crops of rice. TPS has been used in 30 countries and accounts for in excess of an estimated 10 000 ha in India and 15 000 ha in China, for example. An economic analysis by CIP of hybrid TPS use in Viet Nam indicated that about 10 % of the total potato production area (grown by small-scale farmers) utilized TPS and this increased their average yield by 75 %, compared with traditional clonal potato production; it benefited 100 000 rural households; and this single technology accounted for an increased household income of 1·2 %. Similar economic benefits were estimated from the utilization of TPS in India and Indonesia.

TPS research in the Peloquin laboratory ranged from genetic to horticultural in nature. Open-pollinations (OPs) were explored as a means to reduce seed production costs, while trying to retain uniformity and mean performance (Kidane-Mariam et al., 1985). Selfing rates were examined to select an appropriate TPS breeding method (Camadro and Peloquin,1982). Distinct crossing schemes were assessed for their ability to produce uniform TPS offspring (Macaso-Khwaja and Peloquin, 1983). The effects of inbreeding on TPS, OP pollen fertility and tuber yield were examined by Arndt et al. (1990) and Shonnard and Peloquin (1991a). Genetic markers were used to select hybrid TPS seedlings (Arndt and Peloquin, 1990), whereas the effects of seedling selection and planting methods on TPS performance were studied by Shonnard and Peloquin (1991b). Their research led to TPS cultivar options such as OP from 4x potato cultivars and FDR-derived 4x, or TPS hybrids resulting from 4x × 2x crosses, among others.

Based on this accumulated knowledge, Ortiz and Peloquin (1991b) proposed an inexpensive, 4x hybrid TPS production strategy from 2x × 2x crosses: one haploid-species hybrid (female parent) would be selected for male fertility, self-incompatibility, a high frequency of 2n megagametophytes and no 2n pollen, and an unrelated haploid species for male fertility and high 2n pollen production. In addition, the two parents would be selected for profuse flowering, attractiveness to bumble-bees and other desirable characteristics. The male and female parents would be planted on adjacent hills with bumble-bees as pollinators. The elimination of emasculation, pollen collection and hand pollination could reduce 4x hybrid TPS costs by more than 50 %.

CULTIVATING A NEW GENERATION OF SCIENTISTS

The efforts described above reflect visionary insight and leadership of the late Emeritus Campbell-Bascom Professor Stanley J. Peloquin. During these pursuits, he trained large numbers of students and visiting scientists while pursuing these achievements (Fig. 2). Many of the international students and scientists were supported via sustained research interactions with the CIP. The underlying research collaboration model with the CIP combined the international network and mission of the CIP with the cross-section of elite scientific training at a top-tier university and was immensely successful in terms of impacting research, education and real-world needs.

Fig. 2.

Left: Dr Peloquin and one of his former graduate students at potato harvest time in Rhinelander, northern Wisconsin, during the autumn season in the early 1990s. Right: Professor Peloquin teaching about potato science to his students at the University of Wisconsin-Madison in the early 1990s (photographs courtesy of Wolfgang Hoffmann).

A huge part of Dr Peloquin's impact was his classroom teaching and mentoring of undergraduate and graduate students (Fig. 2). Dr Peloquin was closely involved in establishing the BioCore undergraduate course sequence, and a co-founder of the Plant Breeding and Genetics Program at the University of Wisconsin-Madison. The latter program has been highly successful, with 500+ MSc and PhD degrees awarded over the past 40 years. Dr Peloquin also taught key graduate student courses in the Plant Breeding and Plant Genetics Program, where he enthusiastically instructed several thousand students in plant genetics and cytogenetics (Table 1). Dr Peloquin was among the first group of faculty who began teaching the Biology Core Curriculum (Biocore) at the University of Wisconsin-Madison in the 1960s. He taught his first course (Biocore 204 – Organismal Biology) in BioCore in 1967 together with Professors Stanley Beck (Department of Entomology), Oliver Smithies (Laboratory of Genetics) and John Heslop-Harrison (Department of Botany). Dr Peloquin taught almost continuously in the programme until 1993, and following his retirement, he helped recruit one of his former students (Phil Simon), who has continued Dr Peloquin's legacy teaching in this programme.

Table 1.

Teaching by Professor Stanley J. Peloquin at the University of Wisconsin-Madison

Course	Level	Period	Average class size
Introductory Cytogenetics	Advanced undergraduate/graduate course with lecture and hands-on laboratory	1958–1994	30 students per year
Chromosome Manipulations in Plants	Advanced graduate-level course with lecture	1968–1994	18 students
Plant Genetics	Advanced graduate level-course with lecture	1978–1994	18 students
Cellular Biology	Advanced undergraduate/graduate-level course with lecture	1980–1985	160 students
Organismal Biology	Advanced undergraduate/graduate-level course with lecture	1968–1975	160 students
Ecology, Genetics and Evolution	Advanced undergraduate course with lecture and hands-on laboratory	1988–1993	160 students

As Dr Peloquin would attest, the programme's impact was largely due to the collective efforts and innovations of his students and visiting scientists. Dr Peloquin mentored and trained 90+ Plant Breeding and Plant Genetics graduate students from 34 countries, who pursued careers as prominent leaders and educators at world food plant germplasm centres, universities, research institutes and companies (see full list in Supplementary Information to this article, available online). Additionally, Dr Peloquin trained more than 20 established scientists working with food crops from around the world, who came to be mentored by him and then returned to their countries to apply knowledge gained to improve the production of vegetables, fruits, grains and forage crops. He trained more potato researchers than anyone else in the world, and he was among the most prolific trainers of graduate students in the biological sciences at the University of Wisconsin-Madison.

His talent and sincere dedication to mentoring and training students and research scientists working with food crops throughout the world multiplied his personal impact significantly. His contagious enthusiasm and wide range of scientific knowledge and interests inspired his undergraduate workers, graduate students, colleagues and peers at all levels. His superb teaching ability was recognized with teaching awards as he instilled the thrill of science into numerous undergraduate and graduate students who subsequently travelled around the world to pursue careers as researchers, teachers and administrators. Dr Peloquin's scientific and applied contributions yielded 195 referred papers (list available in the dedication to volume 25 of Plant Breeding Reviews; Ortiz et al., 2005), in excess of 250 abstracts, and numerous honours, most notably his election in 1984 to the United States National Academy of Sciences.

SUPPLEMENTARY DATA

A list of graduate students trained by Professor Stanley J. Peloquin and a list of his visiting scientists are available online at www.aob.oxfordjournals.org.