Building an integrated, high-throughput pipeline to accelerate grapevine improvement in New Zealand

Abstract

The value that perennial crop improvement can offer is constrained by the lag between when growers define a problem and when genetic solutions reach the vineyard. During this period, production systems, markets, and growing environments can shift. Reducing this delay requires accelerating progress on two technical challenges: (i) generating populations with novel trait combinations and (ii) identifying improved individuals at scale.

New Zealand’s Sauvignon Blanc 2.0 programme has used somatic embryo mutagenesis to rapidly address a diversity bottleneck, producing >8,000 novel Sauvignon Blanc clones in three years. Afield-ready database links unique vine identifiers to provenance, treatment history, and batch genotyping metadata, providing traceability from plant production through screening.

We describe progress towards an integrated pipeline that combines early-stage genotyping with scalable phenotyping for traits prioritised by the New Zealand wine industry. DNA samples are collected early using automated extraction from young leaves. Genomic and epigenomic data are generated by multiplex nanopore sequencing and mapped to a phased reference genome for variant calling. In parallel, careful plant management in the nursery and vineyard accelerates maturation to advance phenotyping.

Phenotyping spans phenology (e.g., timing of budburst and flowering), disease sensitivity, and abiotic stress responses including drought tolerance and frost susceptibility. With thousands of vines and no replication, measurements must be both accurate and efficient. For each trait, we begin with controlled, targeted stress challenges on a subset of the population to develop biomarkers and calibrate scalable assays. Once these are established, automated imaging and batch scheduling help increase throughput and comparability.

Common trait targets provide opportunities to adapt and standardise assays among international breeding teams. We discuss how shared data, methods, and functional genomic knowledge can shorten feedback loops between discovery and selection to accelerate progress, particularly through counter-season collaboration.