At the 4th International Plant Phenotyping Symposium meeting of the International Plant Phenotyping Network (IPPN) in 2016 at CIMMYT in Mexico, a workshop was convened to consider ways forward with sensors for phenotyping. The increasing number of field applications provides new challenges and requires specialised solutions. There are many traits vital to plant growth and development that demand phenotyping approaches that are still at early stages of development or elude current capabilities. Further, there is growing interest in low-cost sensor solutions, and mobile platforms that can be transported to the experiments, rather than the experiment coming to the platform. Various types of sensors are required to address diverse needs with respect to targets, precision and ease of operation and readout. Converting data into knowledge, and ensuring that those data (and the appropriate metadata) are stored in such a way that they will be sensible and available to others now and for future analysis is also vital. Here we are proposing mechanisms for "next generation phenomics" based on our learning in the past decade, current practice and discussions at the IPPN Symposium, to encourage further thinking and collaboration by plant scientists, physicists and engineering experts.

Australian farmers and scientists have embraced the use of new pasture legume species more than those in any other country, with 36 annual and 11 perennial legumes having cultivars registered for use. Lucerne (Medicago sativa), white clover (Trifolium repens), and red clover (T. pratense) were introduced by the early European settlers and are still important species in Australia, but several other species, notably annual legumes, have been developed specifically for Australian environments, leading to the evolution of unique farming systems. Subterranean clover (T. subterraneum) and annual medics (Medicago spp.) have been the most successful species, while a suite of new annual legumes, including serradellas (Ornithopus compressus and O. sativus), biserrula (Biserrula pelecinus) and other Trifolium and Medicago species, has expanded the range of legume options. Strawberry clover (T. fragiferum) was the first non-traditional, perennial legume commercialised in Australia. Other new perennial legumes have recently been developed to overcome the soil acidity and waterlogging productivity constraints of lucerne and white clover and to reduce groundwater recharge and the spread of dryland salinity. These include birdsfoot trefoil (Lotus corniculatus), Talish clover (T. tumens), and hairy canary clover (Dorycnium hirsutum). Stoloniferous red clover cultivars and sulla (Hedysarum coronarium) cultivars adapted to southern Australia have also been released, along with a new cultivar of Caucasian clover (T. ambiguum) aimed at overcoming seed production issues of cultivars released in the 1970s. New species under development include the annual legume messina (Melilotus siculus) and the perennial legume narrowleaf lotus (L. tenuis) for saline, waterlogged soils, and the drought-tolerant perennial legume tedera (Bituminaria bituminosa var. albomarginata). Traits required in future pasture legumes include greater resilience to declining rainfall and more variable seasons, higher tolerance of soil acidity, higher phosphorous utilisation efficiency, lower potential to produce methane emissions in grazing ruminants, better integration into weed management strategies on mixed farms, and resistance to new pest and disease threats. Future opportunities include supplying new fodder markets and potential pharmaceutical and health uses for humans and livestock. New species could be considered in the future to overcome constraints of existing species, but their commercial success will depend upon perceived need, size of the seed market, ease of establishment, and management and safety of grazing animals and the environment. Molecular biology has a range of potential applications in pasture legume breeding, including marker-assisted and genomics-assisted selection and the identification of quantitative trait loci and candidate genes for important traits. Genetically modified pasture plants are unlikely to be commercialised until public concerns are allayed. Private seed companies are likely to play an increasingly important role in pasture legume development, particularly of mainstream species, but the higher risk and more innovative breakthroughs are likely to come from the public sector, provided the skills base for plant breeding and associated disciplines is maintained.

Subterranean clover (Trifolium subterraneum L.) is the most widely sown annual pasture legume species in southern Australia, valued in the livestock and grains industries as a source of high-quality forage and for its ability to fix atmospheric nitrogen. From its initial accidental introduction into Australia in the 19th Century and subsequent commercialisation in the early 1900s, 45 cultivars have been registered in Australia. These consist of 32 cultivars of ssp. subterraneum, eight of ssp. yanninicum, and five of ssp. brachycalycinum and range in flowering time from 77 to 163 days from sowing, enabling the species to be grown in a diversity of rainfall environments, soil types, and farming systems. Eleven of these cultivars are introductions from the Mediterranean region, 15 are naturalised strains collected in Australia, 18 are the products of crossbreeding, and one is derived from mutagenesis. Cultivars developed in Italy have been commercialised for the local market, whereas other cultivars developed in Spain, Portugal, and France have not had commercial seed production. Important traits exploited include: (i) selection for low levels of the oestrogenic isoflavone formononetin, which causes reduced ewe fertility; (ii) increased levels of dormancy imposed by seed-coat impermeability (hard seeds) for cultivars aimed at crop rotations or unreliable rainfall environments; (iii) strong burr-burial ability to maximise seed production; (iv) resistance to important disease pathogens for cultivars aimed at medium- and high-rainfall environments, particularly to Kabatiella caulivora and root rot pathogens; (v) resistance to pests, particularly redlegged earth mites; and (vi) selection for unique leaf markings and other morphological traits (where possible) to aid cultivar identification. Cultivar development has been aided by a large genetic resource of ~10 000 accessions, assembled from its centre of origin in the Mediterranean Basin, West Asia, and the Atlantic coast of Western Europe, in addition to naturalised strains collected in Australia. The development of a core collection of 97 accessions, representing almost 80% of the genetic diversity of the species, and a genetic map, provides a platform for development of future cultivars with new traits to benefit the livestock and grains industries. New traits being examined include increased phosphorous-use efficiency and reduced methane emissions from grazing ruminant livestock. Economic analyses indicate that future trait development should focus on traits contributing to increased persistence and autumn–winter productivity, while other potential traits include increased nutritive value (particularly of senesced material), increased N2 fixation ability, and tolerance to cheap herbicides. Beneficial compounds for animal and human health may also be present within the species for exploitation.

Camelina ( Camelina sativa (L.) Crantz) is an oilseed known for its potential as a low-input biofuel feedstock and its high levels of beneficial fatty acids. We investigated the role of geographical origin in genetic variation and fatty acid content, expecting to find significant variability among 53 accessions and a link between ecogeography and both origin and key oil traits. Amplified fragment length polymorphism (AFLP) fingerprinting revealed high levels of diversity within the 53 accessions. Even though sampling was relatively biased towards the Russian–Ukrainian area, this region was identified as a genetic diversity hotspot and possible centre of origin for camelina. The accessions were categorized by principal coordinate analysis using molecular marker data, enabling identification of links between geographical distribution and these categories. The influence of geographic location on four canola oil quality measures in camelina was evaluated using a geographic information system. These measures were (1) more than 30% α-linolenic acid, (2) less than 3% erucic acid, (3) less than 10% saturated fatty acids, and (4) a ratio of α-linolenic to linoleic acid greater than 1. The results clearly confirm that camelina oil quality characteristics are strongly influenced by environmental factors. The unprecedented high genetic diversity in this group of accessions offers an excellent opportunity to investigate valuable genes for successful adaptation of camelina to specific ecogeographical conditions such as drought.

Weeds can be major environmental and economic burdens in New Zealand. Traditional methods of weed control including manual and chemical approaches can be time consuming and costly. Some chemical herbicides may have negative environmental and human health impacts. One of the proposed important steps for providing alternatives to these traditional approaches is the automated identification and mapping of weeds. We used hyperspectral imaging data and machine learning to explore the possibility of fast, accurate and automated discrimination of weeds in pastures where ryegrass and clovers are the sown species. Hyperspectral images from two grasses (Setaria pumila [yellow bristle grass] and Stipa arundinacea [wind grass]) and two broad leaf weed species (Ranunculus acris [giant buttercup] and Cirsium arvense [Californian thistle]) were acquired and pre-processed using the standard normal variate method. We trained three classification models, namely partial least squares-discriminant analysis, support vector machine, and Multilayer Perceptron (MLP) using whole plant averaged (Av) spectra and superpixels (Sp) averaged spectra from each weed sample. All three classification models showed repeatable identification of four weeds using both Av and Sp spectra with a range of overall accuracy of 70–100%. However, MLP based on the Sp method produced the most reliable and robust prediction result (89.1% accuracy). Four significant spectral regions were found as highly informative for characterizing the four weed species and could form the basis for a rapid and efficient methodology for identifying weeds in ryegrass/clover pastures.

Oilseed rape ( Brassica napus ) in Australia and China have similar origins, with introductions from Europe, Canada, and Japan in the mid 20th century, and there has been some interchange of germplasm between China and Australia since that time. Allelic diversity of 72 B. napus genotypes representing contemporary germplasm in Australia and China, including samples from India, Europe, and Canada, was characterized by 55 polymorphic simple sequence repeat (SSR) markers spanning the entire B. napus genome. Hierarchical clustering and two-dimensional multidimensional scaling identified a Chinese group (China-1) that was separated from a “mixed” group of Australian, Chinese (China-2), European, and Canadian lines. A small group from India was distinctly separated from all other B. napus genotypes. Chinese genotypes, especially in the China-1 group, have inherited unique alleles from interspecific crossing, primarily with B. rapa, and the China-2 group has many alleles in common with Australian genotypes. The concept of “private alleles” is introduced to describe both the greater genetic diversity and the genetic distinctiveness of Chinese germplasm, compared with Australian germplasm, after 50 years of breeding from similar origins.