Identification and validation of molecular markers associated with pachymetra root rot and brown rust resistance in sugarcane using map- and association-based approaches

Marker-assisted selection for traits that are difficult to screen for, such as resistance to many sugarcane diseases, has the potential to facilitate the development of improved cultivars in sugarcane. Pachymetra root rot (PRR) and brown rust resistance ratings were obtained over two years for 192 I1 progeny ( progeny produced by two heterozygous, non-inbred parental lines) of a sugarcane ( Saccharum spp. hybrid) cross between two elite sugarcane clones, Q117 and 74C42. Approximately 1000 single-dose markers, including microsatellite (SSR), amplified fragment length polymorphism (AFLP) and restriction fragment length polymorphism ( RFLP) markers, were scored across the population and maps containing approximately 400 markers were constructed for each parent. At p <= 0.01, two genomic regions, one from the female Q117 map and a different region from the 74C42 male map, plus an unlinked bi-parental simplex marker (single-dose marker present in both parents) were identified as associated with PRR over both years of data collection. These regions explained between 6 and 16% of the phenotypic variation. An additional region was identified in the female map as associated with PRR at p <= 0.01 in one year and p <= 0.05 in the second year. This region explained between 4 and 8% of the phenotypic variation. For brown rust, two genomic regions, one from the female map and one from the male map, plus an unlinked marker from both maps, were identified as associated with brown rust resistance at p <= 0.01 over two years of phenotypic data. Each region explained between 7 and 18% of the phenotypic variation. Several additional regions were identified in both maps as associated with brown rust at p <= 0.01 in one year and p <= 0.05 in the second year. These regions also explained between 5 and 11% of the phenotypic variation. To validate these markers and determine whether they would be useful in alternative germplasm, markers from each genomic region associated with PRR or brown rust were screened across a set of 154 elite sugarcane clones; PRR and brown rust ratings were available for 131 and 72 of the clones, respectively. For PRR, three of the 6 markers tested remained significantly associated ( p <= 0.01) with resistance ratings in the elite clone set. For brown rust, only one of the seven markers tested remained significantly associated ( p <= 0.01) with resistance in the elite clone set, with one other marker associated at p <= 0.05. These results suggest that these markers could be broadly effective in selecting for PRR and/or brown rust resistance in sugarcane breeding programs.