The process of selecting certain desirable traits for plant breeding may compromise other potentially important traits, such as defences against pests; however, specific phenotypic changes occurring over the course of domestication are unknown for most domesticated plants. Cranberry (Vaccinium macrocarpon) offers a unique opportunity to study such changes: its domestication occurred recently, and we have access to the wild ancestors and intermediate varieties used in past crosses. In order to investigate whether breeding for increased yield and fruit quality traits may indirectly affect anti-herbivore defences, the chemical defences have been examined of five related cranberry varieties that span the history of domestication against a common folivore, the gypsy moth (Lymantria dispar). Direct defences were assessed by measuring the performance of gypsy moth caterpillars and levels of phenolic compounds in leaves, and indirect defences by assaying induced leaf volatile emissions. Our results suggest that breeding in cranberry has compromised plant defences: caterpillars performed best on the derived NJS98-23 (the highest-yielding variety) and its parent Ben Lear. Moreover, NJS98-23 showed reduced induction of volatile sesquiterpenes, and had lower concentrations of the defence-related hormone cis-jasmonic acid (JA) than ancestral varieties. However, induced direct defences were not obviously affected by breeding, as exogenous JA applications reduced caterpillar growth and increased the amounts of phenolics independent of variety. Our results suggest that compromised chemical defences in high-yielding cranberry varieties may lead to greater herbivore damage which, in turn, may require more intensive pesticide control measures. This finding should inform the direction of future breeding programmes.

The development of high-throughput genotyping has made genome-wide association (GWAS) and genomic selection (GS) applications possible for both model and non-model species. The exploitation of genome-assisted approaches could greatly benefit breeding efforts in American cranberry (Vaccinium macrocarpon) and other minor crops. Using biparental populations with different degrees of relatedness, we evaluated multiple GS methods for total yield (TY) and mean fruit weight (MFW). Specifically, we compared predictive ability (PA) differences between univariate and multivariate genomic best linear unbiased predictors (GBLUP and MGBLUP, respectively). We found that MGBLUP provided higher predictive ability (PA) than GBLUP, in scenarios with medium genetic correlation (8–17% increase with corg~0.6) and high genetic correlations (25–156% with corg~0.9), but found no increase when genetic correlation was low. In addition, we found that only a few hundred single nucleotide polymorphism (SNP) markers are needed to reach a plateau in PA for both traits in the biparental populations studied (in full linkage disequilibrium). We observed that higher resemblance among individuals in the training (TP) and validation (VP) populations provided greater PA. Although multivariate GS methods are available, genetic correlations and other factors need to be carefully considered when applying these methods for genetic improvement.

No abstract

The first genetic map of cranberry (Vaccinium macrocarpon) has been constructed, comprising 14 linkage groups totaling 879.9 cM with an estimated coverage of 82.2 %. This map, based on four mapping populations segregating for field fruit-rot resistance, contains 136 distinct loci. Mapped markers include blueberry-derived simple sequence repeat (SSR) and cranberry-derived sequence-characterized amplified region markers previously used for fingerprinting cranberry cultivars. In addition, SSR markers were developed near cranberry sequences resembling genes involved in flavonoid biosynthesis or defense against necrotrophic pathogens, or conserved orthologous set (COS) sequences. The cranberry SSRs were developed from next-generation cranberry genomic sequence assemblies; thus, the positions of these SSRs on the genomic map provide information about the genomic location of the sequence scaffold from which they were derived. The use of SSR markers near COS and other functional sequences, plus 33 SSR markers from blueberry, facilitates comparisons of this map with maps of other plant species. Regions of the cranberry map were identified that showed conservation of synteny with Vitis vinifera and Arabidopsis thaliana. Positioned on this map are quantitative trait loci (QTL) for field fruit-rot resistance (FFRR), fruit weight, titratable acidity, and sound fruit yield (SFY). The SFY QTL is adjacent to one of the fruit weight QTL and may reflect pleiotropy. Two of the FFRR QTL are in regions of conserved synteny with grape and span defense gene markers, and the third FFRR QTL spans a flavonoid biosynthetic gene.

Since its domestication 200 years ago, breeding of the American cranberry (Vaccinium macrocarpon) has relied on phenotypic selection because applicable resources for molecular improvement strategies such as marker-assisted selection (MAS) remain limited. To enable MAS in cranberry, the first high-density SSR linkage map with 541 markers representing all 12 cranberry chromosomes was constructed for the CNJ02-1 progeny from a cross of elite cultivars, CNJ97-105-4 and NJ98-23. The population was phenotyped for a 3-year period for total yield (TY), mean fruit weight (MFW), and biennial bearing index (BBI), and data were analyzed using mixed models and best linear unbiased predictors (BLUPs). Significant differences between genotypes were observed for all traits. Quantitative trait loci (QTL) analyses using BLUPs identified four MFW QTL on three linkage groups (LGs), three TY QTL on three LGs, and one BBI QTL which colocalized with a TY QTL. Local BLAST of a cranberry nuclear genome assembly identified homologous sequences for the mapped SSRs which were then anchored to 12 pseudo-chromosomes using the linkage map information. Analyses comparing coding regions (CDS) anchored in the cranberry linkage map with grape, kiwifruit, and tomato genomes were Electronic supplementary material The online version of this article (

No abstract

Tall fescue (Festuca arundinacea Schreb.) is grown as a forage, turfgrass, and conservation grass from the Gulf Coast to Canada in the eastern USA. The presence of endophytes in grasses can confer resistance to insect pests, enhance growth, and improve tall fescue persistence under high temperature and drought stress. Most tall fescue is infected with Acremonium coenophialum Morgan‐Jones & Gams, a fungal endophyte. Our objectives were to determine the effects of endophyte infection on water relations and survival of water stressed (i.e., drought‐stressed) tall rescue under greenhouse conditions. Endophyte‐free and infected plants of three tall fescue selections were vegetatively propagated and grown together in 38‐L weighing lysimeters to allow interplant competition for available soil moisture. Irrigation water and fertilizer were applied for 12 wk after planting before irrigation water was withheld for 8 wk. Water‐relation characteristics including full‐turgor osmotic potential (ψπ100), water potential at zero turgot (ψL0), relative water content at zero turgot (RWC0), apoplastic water fraction (β), bulk modulus tissue elasticity (ϵ), and the turgid weight to dry weight ratio (TW/DW) were determined before and after water stress. Osmotic adjustment (Δψπ) was also determined. Endophytein fection did not affect ψπ100, ψL0, RWC0, and β either before or after water stress and did not affect Δψπ. Endophyte infection did affect ϵ and the TW/DW ratio, but this effect was not sufficient to alter ψL0. Total tillers, tiller survival, plant survival, and recovery weights were similar for endophyte‐free and ‐infected plants. No evidence for endophyte‐mediated drought tolerance was observed in this study. Based on this work, endophyte‐mediated drought resistance may be due to alterations in drought avoidance.

No abstract