Perceptions of plant breeding methods–from ‘phenotypic selection’ to ‘genetic modification’ and ‘new breeding technologies’

ABSTRACT

Plant breeding has been closely aligned with the development of civilisations and continues to be important for the supply of nutritious food and a key factor in reducing poverty and hunger. Plant breeding uses a range of techniques for both expanding and exploiting the genetic potential of plants. However, some techniques are deemed higher risk than others despite the end products of both processes at times being indistinguishable. While it is considered that the domestication of some plant species began over 10,000 years ago, it is only in the last 100 years or so that modern plant breeding has been used to develop thousands of cultivars in a range of plant species for food, feed, and recreation. In the last 25 years, genetic modification and, more recently, New Breeding Technologies have been used to introduce new variations into important plant species. This has resulted in mistrust and suspicion, and a range of regulatory systems. Product-based and process-based regulatory systems differ in the information required for decision-making. Methods used for the development and manipulation of plant traits are reviewed in an attempt to understand the reasons why some are deemed more acceptable than others.

KEYWORDS:

• CRISPR
• gene editing
• genomic selection
• heterosis
• hybridisation
• mutagenesis
• polyploidy

Introduction

Plants through their ability to capture sunlight energy and store it have been and will continue to be the foundation of food–whether consumed directly or indirectly. Plant breeding which has contributed towards higher yielding and more nutritious crops and forages is one of the factors underpinning the development of civilisations (Hallauer 2011; Breseghello and Coelho 2013) and has alleviated poverty through initially increasing food supply locally but more recently globally (FAO 2022). Plant breeding brings together both the skill of the breeder to recognise value across a range of traits and understanding through science to engineer plants for the benefit of humanity (Hallauer 2011; Poehlman 2013). It is the application of techniques for both expanding and exploiting the genetic potential of plants (Stoskopf et al. 2019; Bowerman et al. 2023) that has been the key to continued improvements in crops and forages. The aim here is to review methods used for both the development and manipulation of plant traits, investigate their effectiveness, and attempt to understand the reasons why some methods are deemed more acceptable than others.

The importance of plant breeding

The history of plant breeding

Plant breeding is considered to have begun over 10,000 years ago (Hallauer 2011) through the gradual domestication of wild plant species. For example, modern domesticated maize (Zea mays ssp, mays) is derived from teosinte (Z. mays ssp. parviglumis), a wild species native to Mexico and Central America (Matsuoka et al. 2002; van Heerwaarden et al. 2011). Teosinte seed, when ripe fall from the cob, which is only made up of two rows of ‘kernels’, and the seed coat is full of silica and lignin making it indigestible and therefore easily distributed in faecal matter (Hake and Ross-Ibarra 2015). In contrast, modern maize kernels are large, soft coated, easily digestible, and held tightly into cobs up to 20 or more rows.

Similarly, domesticated wheat (Triticum aestivum) originated in the Middle East (Iraq, Syria, Lebanon, Israel, and Palestine) as a result of two successive cycles of natural interspecific hybridisation (Salamini et al. 2002). It was spread by humanity both west into Europe and east into Asia where adaptation to local environments occurred resulting in the development of landraces (Balfourier et al. 2019). Landraces are defined as ‘a cultivated, genetically heterogeneous variety that has evolved in a certain ecogeographical area and is therefore adapted to the edaphic and climatic conditions and to its traditional management and uses’ (Casañas et al. 2017). Landraces differ from ecotypes in that landraces result from being cultivated by farmers and becoming domesticated (Zeven 1998; Ortiz 2020). The more recent genetic modification to wheat occurred 70 years ago with the introduction of dwarfing genes and genes for photoperiod insensitivity during the Green Revolution which improved the harvest index and reduced lodging (Trethowan et al. 2007).

The vegetable brassicas cauliflower, brussels sprouts, broccoli, kale, kohlrabi, and cabbage have all been domesticated over many generations from the same precursor germplasm now known as Brassica oleracea (European Commission 2021). Determining the origin and evolutionary history of B. oleracea has been controversial. However, a recent study using newly generated RNA-seq data (which accurately maps short sequence read to a reference genome allowing quantification of gene(s) expression levels) for a diversity panel of 224 accessions, which represents 14 different B. oleracea crop types, and nine potential wild progenitor species have concluded that the Aegean endemic B. cretica is the closest living relative of cultivated B. oleracea, supporting an origin of cultivation in the Eastern Mediterranean region (Mabry et al. 2021). This study also considered that cultivated plants of this species can revert to a wild-like state with relative ease. The different sub-groups of B. oleracea developed through mutation and introgression from wild species during evolution or human selection (Ramchiary et al. 2011). Genetic analysis of these sub-groups has shown they are closely related despite the large morphological differences (Sadowski and Kole 2016).

Modern plant breeding has however a short history of about 100 years, resulting largely from the connection of Mendel’s principles of inheritance (Laird and Lange 2011; Bateson and Mendel 2013; Stenseth et al. 2022), Darwin’s theory of natural selection (Ruse 1975; Ospovat 1995), Vilmorin’s demonstration that parental selection is more effective if based on progeny testing (Gayon and Zallen 1998), and the use of appropriate experimental design and statistical analysis (Fisher 1925; Provine 1971; Hallauer 2011). This has led to the development of thousands of varieties or cultivars in major food crops and forage plants and continues to do so. Even with the advent of modern biotechnologies for manipulating the genome and linking genes to traits of value (Moose and Mumm 2008; Bowerman et al. 2023), the role of the plant breeders remains central to the delivery of new and innovative cultivars.

Plant traits

Phenotype is the expression of both the genetics of the plant and the environment in which the plant is growing. Despite an increase in knowledge of genomics and methods of manipulating the genome, it is the resulting phenotype that is important in delivering the value of the plant breeding or genetic improvement method used (Lee 2006). Phenotypic plasticity is defined as the degree to which a trait can change with changes in environment and management is itself under genetic control (West-Eberhard 1989). Trait expression can be controlled by either single or multiple genes and these themselves may interact with environmental variables. Single or major genes will follow the principles of Mendel in the way they segregate and express in progeny, while complex traits (such as yield or persistence) which are under multigene control are better understood by determining the breeding value of parental genotypes based on the performance of progeny (Falconer 1960).

Plant traits identified by breeders as important can be many and varied, but success in selecting them depends on their heritability (Nyquist and Baker 1991) which is a measure of the extent to which they can be manipulated from a genetic perspective. Two broad categories of heritability are often distinguished. Broad sense heritability is the ratio of total genetic variance (the sum of additive genetic variance, dominance genetic variance, and epistatic genetic variance) to phenotypic variance; while narrow sense heritability is the ratio of additive genetic variance (which is determined by gene frequency and by the average effect of substituting one allele for another) to phenotypic variance (Dudley and Moll 1969).

Methods for developing and manipulating plant traits

The ultimate success of plant breeding is reliant on both the ability of the breeder and the extent of variation available for the traits being sought or selected. Increasing trait variability and then capturing the desired trait expression is achieved through a number of methods, some of which can be regulated.

Phenotypic selection

The ability to accurately identify, define and measure important plant traits is the crucial factor underlying a successful plant breeding programme. This requires access to a wide range of germplasm within the species of interest, which might include wild relatives and landraces, the relevant environment in which to grow the plant material and the perceptive eye of the breeder.

Selection of the correct phenotype is the basic activity of plant breeding closely followed by using the appropriately controlled mating of plants that have been selected with desirable trait expression. The power of phenotypic selection has led to the development of modern crops and forages. Plant breeding methods have been extensively reviewed previously (Allard 1960; Mayo 1987; Poehlman 2013; Acquaah 2015; Stoskopf et al. 2019).

Interspecific hybridisation

While numerous books and articles have been written to examine and defend a variety of definitions of species (George and Mayden 2005; Zachos 2016), here species, as the primary natural taxonomic unit, is defined by being genetically isolated, able to interbreed within the species but unlikely to interbreed with another species (Wilkins 2007). However, using embryo rescue interspecific hybrids can be produced. Embryo rescue is an in vitro technique used when crossing between plants results in embryo abortion or degeneration (Pramanik et al. 2021). In plants, this along with polyploidy has occurred most frequently in ornamentals such as Rosa, Chrysanthemum, Gladiolus, Alstroemeria, Lilium, and orchids (Van Tuyl and Lim 2003). Plant biologists have long considered hybridisation as an important force in adaptation to environmental effects and to further speciation (Grant 1972; Bowley and Taylor 1987; Rieseberg et al. 2004). In crop and forage plants successful artificially-made interspecific hybrids include Triticale–a hybrid between wheat and rye (Secale cereale) (Randhawa et al. 2015; Losert et al. 2017); raphanobrassica–a hybrid between kale (Brassica oleracea) and radish (Raphanus raphanistrum subsp. sativus) (Dumbleton et al. 2022) (Table 1). In horticulture, interspecific hybrids have been widely used in tomato (Solanum lycopersicum) (Vanlay et al. 2022), sugarcane (Saccharum) (Irvine 1999), cucumber (Cucumis) (Tak et al. 2016), Prunus species (Layne and Sherman 1986; Szymajda et al. 2015), apple (Malus) (Korban 1986; Gross et al. 2012), chestnut (Castanea) (Pereira-Lorenzo et al. 2016), papaya (Carica species) (Drew et al. 1998), and banana (Musa species) (De Langhe et al. 2010), (Table 1). For some horticultural species that are propagated through grafting, such as apple (Malus x domestica), the use of rootstocks from wild relatives has been used to improve drought tolerance (Liu et al. 2012). Some rootstocks are themselves interspecific hybrids (Niu et al. 2019).

Table 1. Artificially-made interspecific hybrids used in agriculture and horticulture.

Interspecific hybrid	Trait enhancement	Reference
Chestnut (Castanea)	Blight resistant Chinese (C. mollissima) and Japanese (C. crenata) species used in a backcrossing programme with American chestnuts (C. dentata) to produce blight resistant timber chestnuts; Other crosses provide Phytopthora spp. (ink disease) in Europe	Pereira-Lorenzo et al. 2016
Papaya	Enhanced papaya ringspot virus-type P (PRSV-P) resistance	Drew et al. 1998
Tomato (Kumato –hybrid SX 387)	Colour ranging from a green to reddish brown or purple, varying in flavour from almost no flavour to sweeter than typical tomatoes due to a higher fructose content	United States Patent 2009
Triticale	Combine the superior agronomic performance and end-use qualities of wheat with the stress tolerance (both biotic and abiotic) and adaptability of rye	Randhawa et al. 2015; Mergoum et al. 2019
Raphanobrassica	Combines the superior biomass yield, quality, and palatability of kale, with the increased drought, insect, and disease tolerance of radish	Chakwizira et al. 2021; Dumbleton et al. 2022

Polyploidy

Genome doubling or polyploidy can occur in two ways –allopolyploidy, resulting from hybridisation between two different species, or autopolyploidy, resulting from the duplication of a single genome. Greater new genetic diversity and hybrid vigour is most likely to result from allopolyploidy (Bell et al. 2013). Many crop species are natural polyploids (Mason and Batley 2015)–including potato (Solanum), sweet potato (Ipomoea), cassava (Manihot), taro (Colocassia), yams (Dioscoria), wheat (Triticum), sugarcane (Saccharum), cotton (Gossypium), canola (Brassica), strawberry (Fragaria), and kiwifruit (Actinidia) (Sattler et al. 2016).

Polyploidy has also been generated in some plant species to improve yield and nutritive quality (Sattler et al. 2016). Polyploids can be induced by sexual polyploidisation using a range of chemicals including colchicine, nitrous oxide, oryzalin, caffeine, trifluraline, puromycin, and benzobenil (Younis et al. 2014), resulting in the union of unreduced male and female gametes that have not undergone normal meiosis and still have a 2n constitution (Ramanna and Jacobsen 2003). The anti-mitotic agent colchicine can also be used to disrupt mitosis resulting in chromosome doubling in meristematic cells giving rise to a polyploid shoot (Planchais et al. 2000; Nair 2004; Ranney 2006). While not all polyploids deliver advantages, the opportunities they might provide are overcoming barriers to hybridisation, development of sterile cultivars, restoration of fertility in wide hybrids, improved pest resistance and stress tolerance, and enhanced vigour (Ranney 2006).

Doubling the chromosome number to create tetraploids has been used in forages such as perennial ryegrass (Lolium perenne) primarily to increase nutritive value and digestibility (Balocchi and López 2009). Further, it has been demonstrated that tetraploids of ryegrass are more likely to have higher yields under severe drought conditions and cold than diploids (Kemesyte et al. 2017; Lee et al. 2019a), but they are potentially more susceptible to damage from root feeding invertebrates (Tozer et al. 2017).

Heterotic breeding systems

Heterosis or hybrid vigour can be created by crossing genetically diverse and often inbred populations (Frankel 1983; Shen et al. 2006; Stupar et al. 2008). Mendel observed heterosis in his pea populations, as did early horticulturists in the eighteenth century (Ashby 1937), but the first designed experiments to understand this phenomenon were carried out by Darwin and documented in ‘The effects of cross- and self-fertilization in the vegetable kingdom’ published in 1876 (Mather 1955), Hybrid vigour has been effectively used in crop and vegetable species resulting in superior performance relative to the parental lines (Fujimoto et al. 2018). Production of hybrid seed relies on the prevention of self-fertilisation which can be achieved using a male-sterile female parent. Male sterility can be induced by cytoplasmic male sterility (CMS), nuclear-controlled environment-sensitive genic male sterility (EGMS), or genic male sterility (Chen and Liu 2014; Kim and Zhang 2018). CMS is caused by mitochondrial genes together with nuclear genes, while genic male sterility is caused by nuclear genes alone, and EGMS depends on environmental condition such as photoperiod or temperature. CMS-induced male-sterile female lines can only be maintained by crossing with a maintainer male fertile line (due to mitochondrial DNA only being maternally inherited) while EGMS-induced male-sterile female lines can be self-propagated without a maintainer line. CMS resulting from either spontaneous or artificial mutations has been used in maize, rice (Oryza sativa), sunflower (Helianthus annuus), a range of brassica species (Yamagishi and Bhat 2014), sorghum (Sorghum bicolor), wheat, common bean (Proteus vulgaris), pepper (Capsicum annuum), carrot (Daucus carota), and sugar beet (Beta vulgaris) (Chen and Liu 2014). In species such as maize with a natural separation of male and female reproductive organs, removing the male structure through detasseling is another method of creating hybrids. However, it is a laborious method and the use of cytoplasmic male sterility which is maternally inherited is preferred (Laughnan and Gabay-Laughnan 1983; Levings 1993). Heterosis can result in significant yield increases over open-pollinated cultivars of up to 120% in oilseed brassica crops (Brandle and McVetty 1989), 113% in maize (Zanoni and Dudley 1989), 125% for 10-ear weight in sweet corn (Dickert and Tracy 2002), 134% in rice (Virmani et al. 1982), and 130% in wheat (Duvick 1999). Prior to the development of hybrid maize in the 1930s, open-pollinated cultivars were yielding about 2 tons per ha of grain in the USA while hybrid cultivars in 2018 were yielding close to 12 tons per ha of grain (Hochholdinger and Baldauf 2018).

Chemical and radiation-induced mutagenesis

Mutations are heritable changes to the genetic material of an organism (Mba et al. 2010; Mba 2013). Spontaneous mutations are the primary source of genetic variation in organisms (Kharkwal 2012). Mutagenesis can be induced to increase trait variability by exposing the seed to certain chemicals, x-rays, or radiation (Kodym and Afza 2003; Mba 2013; Oladosu et al. 2016; Ahmar et al. 2020). Mutations either spontaneous or induced are random events, and it is the breeders’ ability to recognise useful and beneficial variations that remains the key to a successful outcome. The International Atomic Energy Agency (IAEA) Mutant Variety Database (IAEA 2022) is a repository of voluntarily contributed information on officially released mutant crop varieties of both seed and vegetative crops. This database indicates the largest number of cultivars released that involved induced mutation breeding was rice (with 873 cultivars), followed by barley (Hordeum vulgare) (307), chrysanthemum (285), wheat (265), soybean (Glycine max) (182), maize (89), and horticultural fruit species (81). The highest number of mutation-induced cultivars are used in Asia (with 2087 cultivars), followed by Europe (960), North America (211), Africa (82), Latin America (53), and Australia/Pacific (9).

Reverse genetics and breeding

Reverse genetic approaches identify gene function through three methods (Kumar et al. 2022): (a) phenotypic changes due to knocking out/altering/silencing the target gene(s) using molecular techniques such as RNAi (interferes with the translation of target mRNA transcript eventually suppressing the gene expression (Saurabh et al. 2014)), virus induced silencing of gene (VIGs) (Senthil-Kumar and Mysore 2011), and homologous recombination which creates covalent linkages between DNA in regions of highly similar or identical sequence (Schuermann et al. 2005), (b) through expression analysis when a target candidate gene is inserted into the species of interest, and (c) screening the target gene(s) in the mutant populations developed by random disruption of genes through mutagens and then determining the phenotypic result (Jankowicz-Cieslak and Till 2015). The latter uses a technique called Targeting Induced Local Lesions IN Genomes (TILLING) (McCallum et al. 2000). TILLING does not involve transgenic modifications but uses gene-specific primers to amplify DNA and then specific nucleases to identify mismatches between wild-type and mutant DNA (Henikoff et al. 2004). This process has the potential for removing the need to screen large putative mutant populations for the desired phenotype and has been used in a large number of crops, including rice, maize, wheat, sugar beet, barley, soybean, pea (Pisum sativum), bean (Phaseolus vulgaris), tomato (Solanum lycopersicum), and the vegetatively propagated banana (Musa spp.) (McCallum et al. 2000).

Single seed descent

Used in self-fertilising crops single seed descent is used to identify phenotypic traits or characteristics that exceed those of the parents for economically important traits such as seed yield (Chahota et al. 2007) and fix these traits for expression in succeeding generations (Lenaerts et al. 2019) through advancing lines to homozygosity. It has the advantages over the normal pedigree breeding method of requiring less time and labour, and transgressive segregation (i.e. producing individuals in subsequent generations that are superior to both the parents) can be fixed and genetic advance obtained (Snape and Riggs 1975; Chahota et al. 2007; Jumbo et al. 2011). Single seed descent genotypes have been effectively used to study the genetic variation within a species, as demonstrated in a large collection of Durum wheat (Triticum durum) (Pignone et al. 2015).

Doubled haploidy

Haploids are sporophytes that have the gametic chromosome number (n). Double haploids can be spontaneous at low frequency but can also be induced to generate homozygous lines without the need for using several generations of selfing (Germanà 2011) and then use these for obtaining hybrid cultivars and/or provide access to recessive genes and for biotechnological manipulations (Lübberstedt and Frei 2012; Prigge and Melchinger 2012). Anther culture is often the preferred method for double haploid production in many crops (Sopory and Munshi 1996). It is only achievable in species able to be tissue cultured, but this does include rice for which this breeding method has led to the release of many cultivars (Grewal et al. 2011; Mishra and Rao 2016).

Apomixis

Apomictic plants produce seeds with embryos derived directly from maternal tissue, not through sexual fertilisation. It has been proposed that apomixis could be used to develop self-reproducing hybrids (Bashaw 1980; Duvick 1999). Apomixis has been observed in over 300 species (Hann and Bashaw 1987). Using apomixis to develop new hybrids would bypass multi-generation selfing for inbred development and eliminate the need for large-scale and expensive cross-pollination blocks to produce hybrid seed (Duvick 1999; Hanna and Bashaw 1987). Several apomictic mechanisms have been shown to be genetically controlled however, it would appear that to date this development has only been used commercially in the breeding of tropical grasses, such as Brachiaria (Miles 2007).

Selection of genotypes with desired traits using genetic and molecular markers

The advantages and disadvantages of different genetic markers used in identifying desired genotypes for further breeding have been recently summarised by Nadeem et al. (2018).

Marker assisted selection

Molecular DNA markers can increase the efficiency of breeding programmes. This has been used effectively for manipulating simple traits but has been slower for complex traits, such as salt tolerance (Ashraf and Foolad 2013). DNA markers that have been used in Marker Assisted Selection include Restriction Fragment Length Polymorphism (RFLP), Random Amplified Polymorphic DNA (RAPD), Amplified Fragment Length Polymorphism (AFLP), Simple Sequence Repeats (SSRs) or microsatellites, and Single Nucleotide Polymorphisms (SNPs) (Collard et al. 2005; Jiang 2013; Nadeem et al. 2018; Lamichhane and Thapa 2022). Marker Assisted Selection has played a prominent role in cereal crop breeding for resistance to obligate biotrophic pathogens such as leaf rust in wheat (Vida et al. 2009), However, while used in the commercial breeding of several crops, including wheat, rice, barley, and maize (Hasan et al. 2021), in general, it has not been fully integrated into all conventional plant breeding programmes (Young 1999; Gupta et al. 2010). Reasons for this low impact of Marker Assisted Selection include reliability and accuracy of quantitative trait loci mapping, association between marker and trait is poorly correlated, impacts of genetic background used to identify markers and a lack of transferability to other germplasm, and environmental interactions (Collard and Mackill 2008).

Genomic selection

Genomic selection (Lenaerts et al. 2019) is based on making genomic predictions from very large numbers of DNA markers rather than focusing on specific genes or quantitative trait loci defined as the section of DNA that correlates with a variation of a quantitative trait in the phenotype of a population (Desta and Ortiz 2014; Heslot et al. 2015). Molecular tools are therefore used to identify superior genotypes and track trait variation. Using a large amount of marker information genomic selection can be used to calculate genomic estimated breeding values (GEBVs) for complex traits that older Marker Assisted Selection systems were ineffective in achieving and were therefore less efficient in accelerating genetic gain (Heffner et al. 2010; Merrick et al. 2022a and 2022b; Sandhu et al. 2022a). Genomic tools have also enabled a better understanding of the complex interactions between plants and pathogens at a molecular level to better identify host defences in resistant and susceptible interactions to provide consideration for the introgression of these traits into breeding programmes (Kankanala et al. 2019). Genomic selection has been heralded as a means to accelerating the efficiency of molecular selection for complex traits associated with climate-resilient crops (Budhlakoti et al. 2022), seed yield (Hu et al. 2022), growth and wood property traits of forest trees (Grattapaglia 2022), and in plants such as sugarcane with the large complex genome, high levels of polyploidy and heterozygosity (Sandhu et al. 2022b). A recent review concluded that plant breeders can improve genetic gains through the use of genome selection by utilising multi-trait and, multi-environment models, high-throughput phenotyping, and machine learning (Merrick et al. 2022a). Another recent review likens genomic selection to ‘instead of trying to find the needle in a haystack, i.e. the “magic” gene in the complex and fluid genome, genomic selection more efficiently and humbly ‘buys the whole haystack’ of genomic effects to predict complex phenotypes’ (Grattapaglia 2022).

Genotyping-by-sequencing

Genotyping-by-sequencing has enabled the development of genome-wide markers in non-model polyploid plants (Baral et al. 2020). Sequence-based markers that have been used are Single-Nucleotide Polymorphism (SNP) and Diversity Arrays Technology (DArT) markers (Nadeem et al. 2018). The ability to use markers for traits of interest is enhanced through using genome-wide SNP markers which provide more accurate selection of parental genotypes (Meuwissen et al. 2001; Heffner et al. 2009; Crossa et al. 2011, 2017).

Genetic engineering and genetic modification (GM)

Genetic engineering and modification are here defined as the manipulation of an organism’s genes by introducing, eliminating or rearranging specific genes using the methods of molecular biology (USDA 2022). This process can allow the introduction of new traits and variation otherwise not available in the unmodified genome, such that ‘genetic material has been altered in a way that does not occur naturally by mating and/or natural recombination’ (EU Directive 2001). This broad all-encompassing definition from the European Union has almost certainly influenced decisions in parts of the world where the options are fewer and less attractive to alleviate human suffering wherever possible (Herring 2008). The benefits, risks, and intended and unintended consequences of this technology have been recently reviewed (Caradus 2022a) along with opportunities they present for the future (Bowerman et al. 2023). There is a range of methods used to modify plant genomes some or all of which are regulated depending on the country. In New Zealand, all are currently regulated under the Hazardous Substances and New Organisms (HSNO) Act 1996 administered by the Ministry for the Environment (New Zealand Legislation 1996) (Caradus 2022b). Adapting regulatory frameworks to accommodate rapidly changing genome editing technologies to balance risk against benefits is a challenge which some jurisdictions are adjusting to better than others (Rozas et al. 2022).

Transgenesis

Transgenesis involves using recombinant DNA technologies to introduce genes from other organisms to another. It may also involve genes artificially synthesised in the laboratory(Herrera-Estrella et al. 2004). This methodology has been extensively used over the last 25 years to deliver 525 different transgenic events, in 32 crops and flower species across 26 countries (Caradus 2022a). Transgenesis has predominantly used either an Agrobacterium-mediated process (Gelvin 2000), or biolistics (Sanford 1990) to introduce the transgene(s), along with a selectable marker (for ease of differentiating between genotypes with the transgene and those without) and a promoter. Less frequently used methods include the use of electroporation, viral vectors, polymers, and nanoparticles (Cunningham et al. 2018; Ahmar et al. 2021). Selectable markers might be either herbicide resistant gene, antibiotic resistant gene or inducible growth, or differentiation of transformed tissues (Miki and McHigh 2004). The promoters used might be one that is constitutive (drives expression of a gene in all plant tissues and throughout all developmental stages), tissue-specific (expresses the gene in only a specific type of tissue), and inducible (initiates gene expression under certain external conditions) (Liu 2009; Smirnova et al. 2012).

Using plant breeding to combine stacked genes into a single breeding line can be unwieldy with the population size needed to isolate a stacked F2 plant increasing exponentially with the increasing number of unlinked genes (Pathak and Srivastava 2020). Repeated recombinase-mediated DNA cassette exchanges linking the introduced genes and stacking them into a locus improve efficiency (Petolino and Kumar 2016; Que et al. 2010). For example, multi-gene integration in crops for agronomically relevant traits has been achieved in soybean were genes fatty acid ω−6 desaturase 2 and acyl-acyl carrier protein thioesterase 2 were silenced to improve oleic acid content (Li et al. 2010). Similarly, site-specific integration of a five-gene cassette has been demonstrated successfully in rice (Pathak and Srivastava 2020).

Cisgenesis and intragenesis

Cisgenesis is defined as transferring a gene from the same or a closely related species (Holme et al. 2013; Freddy et al. 2022; Hefferon 2022; Koul 2022) along with its own regulatory promoter and terminator (Kumar et al. 2020). This can lead to a change in the expression pattern of the desired gene driven by regulatory elements from other genes (Kumar et al. 2020). Similarly, intragenesis involves the construction of the entire vector region destined for transfer to a species from the genome of that species (Rommens 2004; Rommens et al. 2004; Conner et al. 2007; Rommens et al. 2007; Barrell et al. 2010; Barman et al. 2020). For well over a decade, cisgenesis and intragenesis have promised a ‘clean’ DNA delivery system for gene transfer. There is no published evidence of cisgenic or intragenic events having ever been commercialised and their promise remains to be fulfilled.

Targeted gene editing

Gene editing or sequence-specific nuclease technology (Schaart et al. 2016; Mohanta et al. 2017; Songstad et al. 2017) differs from previous techniques in manipulating the plant genome by being more precise and more versatile, and relatively rapid to undertake. Gene editing methods rely on specific DNA binding proteins which are found in all cells to ensure that genetic material is appropriately expressed, replicated, and transmitted from one generation to the next. These DNA binding sites are short DNA segments referred to as motifs or signatures that can be used to predict protein function (Das and Dai 2007). Early approaches to introduce site-specific modifications to plant genomes in an attempt to control gene expression used oligonucleotides, small molecules (Gottesfeld et al. 1997), or self-splicing (Yang et al. 1996). However, more recent technologies using principles of DNA–protein recognition resulted in the development of site-directed zinc finger nucleases (ZFNs) and TAL effector nucleases (TALENs) and then most recently clustered regularly interspaced palindromic repeats (CRISPR) and their associated (Cas) nucleases (Doudna and Charpentier 2014).

Oligonucleotide-directed mutagenesis (ODM) is a precise genome editing technology that uses oligonucleotides (short single strands of synthetic DNA or RNA that serve as the starting point for molecular biology applications) to make targeted edits in plasmid, episomal, and chromosomal DNA (Oh and May 2001; Dong et al. 2006; Gocal 2015; Gocal et al. 2015; Sauer et al. 2016). While the use of oligonucleotides to precisely direct, mutagenesis in preselected sequences has been demonstrated to be effective it has not been widely deployed in crops (Cardi and Neal Stewart 2016).

Chimeric zinc finger nucleases allow direct targeting of chromosome sites for gene insertion (Bibikova et al. 2003) by exploiting the natural recognition mechanism of cellular DNA repair machinery, to make sequence-specific double-stranded DNA breaks at a target locus (Townsend et al. 2009; Shukla et al. 2009; Kumar et al. 2020).

Transcription activator-like (TAL) effector proteins (TALENS) are an effective technology for genome editing that use targetable nucleases to induce double-strand breaks into specific DNA sites, which are then repaired by mechanisms that can be exploited to create sequence alterations at the cleavage site (Joung and Sander 2013). Transcription activator-like effectors (TALEs) are major virulence factors secreted by the plant pathogenic bacterial genus Xanthomonas, which causes disease in plants such as rice and cotton. TALEs are injected into host cells and interfere with cellular activities by activating the transcription of specific target genes (Bogdanove et al. 2010; Kumar et al. 2020). TALENS has been most effectively used to knock out genes expressing unwanted traits (Bezie et al. 2021).

Clustered regularly interspaced palindromic repeats (CRISPR) and their associated (Cas) nucleases allow a process of rapidly and efficiently targeting protein domains in areas of interest in a genome (Zhang et al. 2014; Ma et al. 2015; Doudna and Sternberg 2017; Eş et al. 2019). The CRISPR/CasX (i.e. Cas9 and Cas12 and its variants) system consists of two essential components–Cas nucleases which allow a process of rapidly and efficiently targeting areas of interest in a genome, and a single-guide RNA which targets the section of DNA to be edited (Hsu et al. 2014; Lee et al. 2019b). An extensive review has shown that CRISPR/Cas9 technology has been effectively used by researchers to bring efficient targeted transformation to a range of crop species (Bezie et al. 2021). Gene editing using CRISPR deliver three types of outcomes. CRISPR gene editing uses Site-Directed Nucleases (SDN) which uses different DNA-cutting enzymes (nucleases) that are directed to cut the DNA at a predetermined location by a range of different DNA binding systems (Hsu et al. 2014; Lee et al. 2019b). After the double strand, the cut is made, the cell’s own DNA repair mechanism recognises the break and repairs the damage, using one of two pathways that are naturally present in cells (1) non-homologous end-joining (NHEJ): The cut DNA is re-joined, but while doing this errors are made resulting in random small deletions (up to 20) or additions (a few base pairs) of nucleotides at the cut site or (2) homology-directed repair (HDR): a donor DNA that carries the desired change and has homology with the target site is used to introduce this change at the cut site (Bibikova et al. 2002; Cong et al. 2013). This leads to one of three outcomes–gene disruption or deletion (SDN1), gene correction or modification (SDN2), or DNA insertion (SDN3) (Doudna and Charpentier 2014; FAO 2022).

The European Food Safety Authority has recently declared that the documents for ‘Guidance for risk assessment of food and feed from genetically modified plants’ and the ‘Guidance on the environmental risk assessment of genetically modified plants’ are ‘not relevant for the risk assessment of plants developed via SDN-1, SDN-2 or oligonucleotide-directed mutagenesis approaches if the genome of the final product does not contain exogenous DNA’ (Naegeli et al. 2020). In reality, SDN-1 and SDN-2 type gene edits are cisgenic in nature (Cabrera-Ponce et al. 2022; Ghose et al. 2022), and SDN-1 type knockout gene edits have been confusingly referred to as null lines (Kim et al. 2023).

Alternatives to the commonly used Cas9 single-component effector protein have been identified and include Cas12/Cpf1 which is a single RNA-guided endonuclease which cleaves DNA via a staggered DNA double-stranded break motif (Zetsche et al. 2015; Yin et al. 2017; Bandyopadhyay et al. 2020; Kumar et al. 2020). Different Cas effector proteins can result in a range of on-target mutations in plants, ranging from 90 to 100% for Cas9 and 0%–60% for Cas12a (Lee et al. 2019b). Others have used CRISPR/Cas9 ribonucleoprotein complexes to achieve consistently no off-target mutations (Liang et al. 2017). Compared with other forms of mutagenesis (e.g. chemical mutagenesis using ethyl methanesulfonate or X-ray radiation, etc.) off-target effects of CRISPR-Cas9 are trivial. Even with low levels of off-target mutations, sequence analysis can be used to identify genotypes with off-target events to ensure they are not selected.

A comparison of the four main genome editing methods oligonucleotide-directed mutagenesis, zinc finger, TALEN, and CRISPR is provided by Gao et al. (2020). They conclude that compared with CRISPR, both zinc finger and TALEN-mediated gene editing techniques have some cumbersome protein engineering steps, high costs, and are difficult to multiplex.

RNA-directed DNA methylation and epigenesis

RNA-directed DNA methylation (i.e. where a methyl (CH3) group molecule gets added to DNA) induces transcriptional gene silencing via methylation of the target gene promoter sequence by RNA interference (RNAi) (Schaart et al. 2016), is unique to plants (Erdmann and Picard 2020). Epigenetic engineering technologies make it experimentally possible to modify individual chromatin marks (sites where DNA methylation, small interfering RNA, or histone variants are associated with an epigenetic event) at specific user-defined sites (Margueron and Reinberg 2010; Köferle et al. 2015). Epigenetic processes change the activity of genes without changing a DNA sequence, but still leads to modifications that can be transmitted to daughter cells (although some epigenetic changes can be reversed) (Weinhold 2006). Epigenetic processes can occur through DNA methylation, acetylation, phosphorylation, ubiquitylation, and sumolyation. Epigenetic changes occur naturally through environmental impacts, nutrition, and ageing (Huidobro et al. 2013). Initially, understanding the epigenetic regulation of gene expression in plants was undertaken in the lab-species Arabidopsis. Evidence would suggest heritable epiallelic variations associated with a trait of interest could be utilised for crop improvement, particularly in combination with targeted gene editing (Puchta 2016; Adli 2018; Kumar 2019). This includes the understanding of epigenetic regulation in crop plants in response to both biotic (Huang and Jin 2022) and abiotic stresses (Perrella et al. 2022; Singh and Prasad 2022; Verma et al. 2022).

Grafting with GM rootstocks

Grafting onto GM rootstock occurs when the top of a non-GM plant (scion) is grafted on a GM rootstock such that the scion may benefit from traits conferred by transgenes in the rootstock, but its products do not contain the transgene itself. (Schaart et al. 2016; Hu and Gao 2023). Using transgenic rootstock to protect the grafted non-transgenic scion of grapevine from Xylella fastidiosa (which causes Pierce’s disease) is an example of this technology (Dandekar et al. 2019). In this case, a plant polygalacturonase inhibitory protein (PGIP) that prevents Pierce’s disease by inhibiting the breakdown of pectin present in primary cell walls is expressed in the transgenic rootstock transiting into the xylem of the grafted scion.

Agroinfiltration

Agroinfiltration is a technique using Agrobacterium transformation as a tool to achieve temporary and local expression of genes in plant tissue. Agroinfiltration is applied for testing the reaction of target plants to transgenic proteins, or for functional gene analysis in plants (Schaart et al. 2016).

Synthetic biology

Synthetic biology in plants refers to the introduction of new synthetic genomic pathways that deliver natural products and is expected to provide significant opportunities in agriculture (Zhu et al. 2021). ‘All definitions encompass the notion that applications of synthetic biology involve the creation of novel living systems through synthesising and assembling artificial and/or natural components’ (Jin et al. 2019). In terms of crop improvement outcomes, all have been at laboratory scale with most commercial-scale applications occurring in microbes that can be fermented (Wehrs et al. 2019; Yilmaz et al. 2022; Kwan et al. 2023).

Null segregants

Null segregants are non-GM progeny derived from crossing a GM plant with a wild-type non-GM plant. The value of null segregants is primarily with horticultural plants where it can take several years for plants to reach maturity and the introduction of new traits by crossing plants over multiple generations can take many years. By using genetically modifying germplasm with a shortened flowering time in the breeding programme (Flachowsky et al. 2009), so that they mature more rapidly (e.g. mature in 1 year instead of 4 years), means that new desirable characteristic(s) can be incorporated into elite commercial cultivars in significantly less time. The genetic modification is simply used to speed up the breeding programme and the new trait is the result of a non-GM change in the genome. To produce the commercial cultivar the line with the now-stably incorporated new trait is crossed one or two times with a non-GM line. The genetic modification for speed breeding is therefore selected out and the new cultivar now has the desired trait(s), matures at the normal rate, and is not genetically modified (Royal Society Te Apārangi 2019).

Advantages of new breeding technologies (NBT)

Some of the methods used to provide increased variation and opportunities for trait selection have been termed New Breeding Technologies (NBT) or New Genomic Techniques–namely, genome or gene editing to modify DNA at one or more specific sites using CRISPR, Zinc Finger Nucleases, or TALENs; introducing targeted changes to a small number of bases of DNA using oligonucleotide-directed mutagenesis; cisgenesis; intragenesis (inserting a reorganised regulatory coding region of a gene from the same species); and using epigenetic processes to change the activity of genes without changing a DNA sequence (Parisi and Rodríguez-Cerezo 2021).

Many have stressed that differences between older transgenic technologies and New Breeding Technologies challenge the appropriateness of many existing regulatory systems. The benefits of New Breeding Technologies compared with transgenesis are described in Table 2.

Table 2. Comparison of benefits and challenges of New Breeding Technologies compared with transgenesis

Comparator	Transgenesis	New Breeding Technologies	Reference
Precision of mutation	Low; Relatively random and untargeted	High; CRISPR Cas can modify one or a few target sites with precision; some Cas effector proteins are more accurate than others	Ma et al. 2015; Lee et al. 2019b
Gene modification without introducing foreign DNA	Not possible	Possible through gene disruption or deletion (SDN1)	Veillet et al. 2019
Versatility, flexibility, adaptability	Moderate in requiring a gene construct to allow expression of the introduced gene	High–can be facilitated through gene disruption or deletion, but can also include gene modification or gene introduction	Pandey et al. 2019; Wheatley and Yang 2021
Efficiency and time to delivery	Moderate	Relatively fast	Zhang et al. 2014; Wolter et al. 2019
Inclusion of genetic diversity from wild relatives	Poor and time consuming	Effective option	Wolter et al. 2019
Agrobacterium and/or biolistics mediated gene insertion	Agrobacterium and biolistics	Agrobacterium and biolistics	Kuluev et al. 2019; Basso et al. 2020

Precisely targeted gene editing is becoming an increasingly important and dominating method which is being promoted as ‘an advance in technology that has the potential to transform some aspects of the world’s agrifood systems for the better’ (FAO 2022). The potential influence and power of genome editing have led to claims that science is now ‘not only able to “read” the “Book of Life”, but also to “write” it and “edit” it’ (European Commission 2021). This has led to the question of ethics and how metaphors used to describe new technologies can influence public understanding (O'Keefe et al. 2015).

As a market, genome editing across pharmacology, biotechnology, and academia is estimated to be worth US$5.1 billion and with a CAGR of 18.2% expected to grow to US$11.7 billion by 2021 (Markets and Markets 2022). Other estimates vary between $7.4 billion by 2031 (Allied Market Research 2022) and $18.5 billion by 2028 (Biospace 2023). This is being driven by growth in funding by governments, increased application areas, and the introduction of CRISPR CasX. The potential benefits of integrating gene editing into plant breeding programmes using targets and traits with application in New Zealand has been reviewed and concludes that New Zealand is torn between the cautious ‘wait-and-see-approach’ with regard to the regulation of gene editing and implementing innovative solutions which are essential for the industry to maintain its global competitiveness (Fritsche et al. 2018).

Research with New Breeding Technologies has been used in the most part to demonstrate proof of concept. Most of these have been aimed at improved disease resistance, improved crop quality, flowering traits, and herbicide resistance. However, several finished cultivars have been submitted for de-regulation and a few have been commercialised (Table 3). There is also the promise of other traits of value that New Breeding Technologies could provide (Jones et al. 2022), including abiotic stress tolerances (Zafar et al. 2020); environmental stress resilience (Kouhen et al. 2022); and improved yield (Gao et al. 2020). An extensive summary of the use of gene editing by plant species, traits, techniques, and applications is available through an interactive internet database (eusage 2023).

Table 3. Examples of cultivars or germplasm developed using different methods of genetic engineering and genetic modification–those in bold have been submitted for de-regulation or commercialised. Refer also to Penna and Jain 2023) for a summary of successful examples of genome editing in fruit crops.

Method	Germplasm	Trait	Status	Reference
Cisgenesis	Apple (Malus domestica)	Increased scab resistance	Demonstrates proof of concept	Vanblaere et al. 2011
	Barley	Improved grain phytase activity	Demonstrates proof of concept	Holme et al. 2012, 2020
	Durum wheat (Triticum durum)	Improved baking quality	Demonstrates proof of concept	Gadaleta et al. 2008
	Grape (Vitis vinifera)	Improved fungal disease resistance	Demonstrates proof of concept	Dhekney et al. 2011
	Potato	Expression of resistance genes for late blight (Phytophthora iInfestans) resistance	Demonstrates proof of concept	Haverkort et al. 2009; Jo et al. 2014
		Resistance to Potato virus Y	Demonstrates proof of concept	Miroshnichenko et al. 2020
Intragenesis	Potato	Gene silencing to levels of amylopectin	Demonstrates proof of concept	de Vetten et al. 2003
		Prevention of black spot bruising	Demonstrates proof of concept	Rommens et al. 2004
		Limited acrylamide production in French fries	Demonstrates proof of concept	Chawla et al. 2012
		Low acrylamide, resistance to bruising and discolouration	Commercially available	Waltz 2015; ISAAA database 2023
	Tomato	Resistance to tomato spotted wilt virus and caulifower mosaic virus	Has been de-regulated in USA	Kumar et al. 2020
Transgenesis	32 crop and flower species	525 different transgenic events providing 2063 GM foods and 1461 GM feeds; underpinned by 824 regulatory approvals	Commercially available	Reviewed by Caradus 2022
Oligonucleotide-directed mutagenesis	Canola	Mutations in canola AHAS homologues, to develop Cibus’ herbicide-tolerant SU Canola^TM	Commercially available	Gocal 2015
	Maize	Single-point mutation of genes encoding acetohydroxyacid synthase, at a position known to confer imidazolinone (herbicide) resistance.	Demonstrates proof of concept	Zhu et al. 2000
		Altered epicuticular wax composition	Demonstrates proof of concept	Char et al. 2015
	Rice	Chimeric RNA/DNA oligonucleotides targeting endogenous acetolactate synthase gene conferring resistance to some herbicides	Demonstrates proof of concept	Okuzaki and Toriyama 2004
Zinc finger	Maize	Insertional disruption of one target locus results in both herbicide tolerance and the expected alteration of the inositol phosphate profile in developing seeds	Demonstrates proof of concept	Shukla et al. 2009
TALENs	Potato	Reduced acrylamide formation	Demonstrates proof of concept	Clasen et al. 2016
		Reduced toxic steroidal glycoalkaloids formation	Demonstrates proof of concept	Sawai et al. 2014
	Rice	Improved rice bacterial blight resistance	Demonstrates proof of concept	Li et al. 2012
		Increase the synthesis of 2-acetyl-1-pyrroline which is a major fragrance compound	Demonstrates proof of concept	Shan et al. 2015
	Soybean	Oleic acid increased from 20% to 80% and linoleic acid decreased from 50% to under 4%; increased shelf life and oil stability	Submitted for de-regulation	Haun et al. 2014; Kumar et al. 2020
		High oleic acid (increased shelf life and trans-fat free)	Submitted for de-regulation	Waltz 2019; Kumar et al. 2020
	Sugarcane (Saccharum spp. hybrids)	Reduce lignin content to improve the conversion of lignocellulosic biomass into ethanol	Demonstrates proof of concept	Jung and Altpeter 2016; Kannan et al. 2018
	Wheat	Mutation of all three mildew resistance locus homeologs confers heritable broad-spectrum resistance to powdery mildew	Submitted for de-regulation	Wang et al. 2014; Kumar et al. 2020
CRISPR CasX (i.e. Cas9 and Cas12 and its variants)	Apple	Increase resistance to fire blight disease (Erwinia amylovora)	Demonstrates proof of concept	Malnoy et al. 2016
	Banana	Knockout the integrated the virus sequences of the integrated endogenous banana streak virus (eBSV) in the B genome of Plantain to allow breeding and dissemination of hybrids	Demonstrates proof of concept	Tripathi et al. 2019
	Cacao (Theobroma cacao)	increased resistance to the cacao pathogen Phytophthora tropicalis	Demonstrates proof of concept	Fister et al. 2018
	Canola (Brassica napus)	Improved fatty acid composition	Demonstrates proof of concept	Huang et al. 2020
		Single base editing of gene producing Acetolactate synthase to confer tribenuron-methyl herbicide resistance	Demonstrates proof of concept	Wu et al. 2020
	Cassava (Manihot esculenta)	Reduced cassava brown streak disease	Demonstrates proof of concept	Gomez et al. 2019
	Citrus (Citrus sinensis)	Improvement of citrus canker resistance through targeted modification of the susceptibility gene	Demonstrates proof of concept	Peng et al. 2017
	Cotton (Gossypium hirsutum)	Resistance to Verticillium wilt	Demonstrates proof of concept	Li et al. 2017
	Cucumber (Cucumis sativus)	Improved virus resistance	Demonstrates proof of concept	Chandrasekaran et al. 2016
	False flax (Camelina)	Increased oil content	Submitted for de-regulation	Waltz 2018
	Grape	Reduced tartaric acid synthesis	Demonstrates proof of concept	Ren et al. 2016
		Increased resistance to powdery mildew (Erysiphe necator)	Demonstrates proof of concept	Malnoy et al. 2016.
	Green bristlegrass (Setaria viridis)	Delayed flowering	Submitted for de-regulation	Waltz 2018; Kumar et al. 2020
	Maize	Decreased sensitivity to ethylene and concomitant improvement in yield under drought	Demonstrates proof of concept	Shi et al. 2017
		Reduced ligule development, imposed male sterility, and improved herbicide acetolactate synthase resistance	Demonstrates proof of concept	Svitashev et al. 2015
		Waxy corn with starch composed exclusively of amylopectin	Submitted for de-regulation	Waltz 2018
	Potato	Altered starch quality	Demonstrates proof of concept	Andersson et al. 2017
		Chlorsulfuron herbicide resistance	Demonstrates proof of concept	Veillet et al. 2019
	Rice	Potential for improved salinity tolerance	Demonstrates proof of concept	Duan et al. 2016; Farhat et al. 2019
		Improved resistance to rice blast	Demonstrates proof of concept	Wang et al. 2016
		Increased amylose content and resistant starch content to 25% and 9.8%, respectively; altering fine structure and nutritional properties of starch	Demonstrates proof of concept	Sun et al. 2017
		Regulation of stomatal density and patterning	Demonstrates proof of concept	Yin et al. 2017
	Sorghum	Reduced kafirin levels and improved protein quality and digestibility	Demonstrates proof of concept, but not in all germplasm (Massel et al. 2023)	Li et al. 2018
	Soybean	improved herbicide acetolactate synthase resistance	Demonstrates proof of concept	Li et al. 2015
		Early flowering under long day conditions	Demonstrates proof of concept	Han et al. 2019
		Drought and salt tolerance	Submitted for de-regulation	Curtin et al. 2018; Waltz 2018; Kumar et al. 2020
		Increased Oleic oil content	Approved for sale in USA	Calyxt 2019
		Low presence of raffinose and stachyose sugars	Approved in Colombia	GDM Seeds 2022
	Tomato	High amounts of γ-aminobutyric acid (GABA)	Commercialised	Waltz 2022
		Increased vitamin D3 expression, to provide a plant based source of Vitamin D3	Demonstrates proof of concept	Li et al. 2022
		Altering morphology of wild-type tomato with high antioxidant content	Demonstrates proof of concept	Zsögön et al. 2018
		Increased resistance to pathogens, Pseudomonas syringae, Phytophthora capsici and Xanthomonas spp.	Demonstrates proof of concept	De Thomazella et al. 2021
		Chlorsulfuron herbicide resistance	Demonstrates proof of concept	Veillet et al. 2019
	Wheat (SDN1)	Low gluten	Held up by European GM regulations	Sánchez-León et al. 2018; Jouanin et al. 2019, 2020
RNA-directed DNA methylation (Epigenesis)	Maize	High amylopectin starch content	Submitted for de-regulation	Waltz 2016; Kumar et al. 2020
	Petunia	Induced cytosine methylation that affected changes in flower colouration	Demonstrates proof of concept	Kanazawa et al. 2011
	Soybean	Increase in variance for pods per plant, seed weight and maturity time	Demonstrates proof of concept	Raju et al. 2018
	Tomato	Induced cytosine methylation that affected rate of fruit ripening	Demonstrates proof of concept	Kanazawa et al. 2011
Grafting on GM rootstock	Cherry (Prunus avium)	RNA interference (RNAi) vector used for developing viral resistance Prunus necrotic ringspot virus in rootstocks	At proof of concept stage	Song et al. 2013
Synthetic biology	Cereals	Nitrogen fixation symbiosis	Proposed opportunity concept	Rogers and Oldroyd 2014
	Crop plants	Production of health producing phytochemicals, e.g. carotenoids/isoprenoids	Proposed opportunity concept	Fraser et al. 2009

Perceptions–why are some methods causing concerns?

Of the wide range of methods used for creating and manipulating genetic variation of plants only those defined as genetic engineering, which in some countries includes New Breeding Technologies, are likely to be regulated (Table 4). The extent of this regulation does vary from country to country, and currently there is no internationally recognised or agreed standard for what should and should not be regulated. It has been argued that ‘current uncertain and complex global regulatory situation is stifling innovation and the realisation of the full potential of these technologies’ (Tagliabue 2016; Atanassova and Keiper 2018). Many countries, including New Zealand, regulate plant breeding outcomes based on the process used to produce them while others regulate the risk of the product with less concern being given to the process used to produce them (Table 4). Reasons for regulation are sound in that they are ensuring that products produced result in low levels of risk to human health and the environment. However, achieving this through regulating the process makes no logical sense when it is the end product that is released of which the process used to produce it is only the means to an end. Indeed the process-based regulatory systems are an overly precautionary approach which has the unintended consequence of stigmatising the use of some biotechnology methods (Tait 2001; Marchant and Stevens 2015; Aerni 2019; Smyth and Lassoued 2019; Anyshchenko and Yarnold 2021; Caradus 2022). Process-based regulatory systems also suffer from becoming increasingly obsolete as the scope of risks associated with processes become better understood (Atanassova and Keiper 2018). However, changing regulatory laws can be slow and contentious with polarised views reducing arguments to differences of opinion rather than pragmatically seeking a fact-based solution.

Table 4. Review of some country differences in regulation of genetic engineering techniques and in particular New Breeding Technologies (NBTs) (also refer to Atanassova and Keiper 2018; Menz et al. 2020; and Turnbull et al. 2021; Jones et al. 2022).

Country or region	Regulatory system	Regulation of NBTs	Reference
Argentina	Process based but with updated legislation to provide clarity on the regulatory status of products of plant breeding innovations; regulatory framework explicitly addresses products obtained from NBTs	Uses a case-by-case determination of whether a ‘new combination of genetic material’ was generated with an absence of events (transgenes) that may have been used transiently during the development process. Crops modified using SDN-1, SDN-2 and ODM would generally not be regulated	Whelan and Lema 2015; Lassoued et al. 2020; OGTR 2021
Australia	Process based, but reviewed in 2016; however GM derived food can be used if shown to be as safe as non-GM food (must be labelled)	NBTs that do not involve the application of nucleic acid repair template (e.g. SDN-1) are excluded	MPI 2021; OGTR 2021; FAO 2022; FSANZ 2022
Canada	Regulates products based on the expression of a novel trait, rather than the method used in its development	Uses ‘novelty’ as a trigger to assess the environmental safety of plants with novel traits or novel feed prior to product release	CFIA 2016; Lassoued et al. 2020
European Union	Both process- and product- focused elements, but predominantly interpreted as a strict process-focused system–reviewed in 2016 but remains unchanged	All NBTs regulated as genetically modified	Lassoued et al. 2020; OGTR 2021
New Zealand	Processed based–not reviewed in 25 years; however GM derived food can be used if shown to be as safe as non-GM food (must be labelled)	All NBTs regulated as genetically modified as a result of a High Court ruling	Fritsche et al. 2018; Royal Society Te Apārangi 2019; FSANZ 2021, 2022; MPI 2021; Caradus 2022; FAO 2022
USA	US implemented a Coordinated Framework for the Regulation of Biotechnology to assess and regulate the products of biotechnology. Regulation of specific risks (i.e. product characteristics) rather than the application of default regulation to all products of biotechnology	Does not regulate plants with deletions, single-nucleotide substitutions, insertions from compatible plant relatives or null segregants, unless plant pest species are involved. Therefore the majority of SDN-1 and ODM outcomes and some SDN-2 and −3 outcomes in plants are outside USDA’s regulatory scope	FDA 2020; Lassoued et al. 2020; OGTR 2021

The driving forces behind these concerns about the processes used to deliver new and unique variations are general concerns by some consumers resulting from suspicions of corporate multinationals, that GM crops are deemed to be un-natural and ‘playing with nature’, the so-called ‘precautionary approach’, absence of economic or environmental benefit, a significant level of misinformation, controversy surrounding negative results, and distrust of science.

Consumer concerns

Consumer acceptance is crucial for food products to be successfully marketed and recognised (MacFie 2007; Krishna and Qaim 2008; Maghari and Ardekani 2011). Concerns include impacts on human health, the environment through the creation of super weeds, and adverse effects of beneficial species and biodiversity (Gaharwar et al. 2021). Consumer attitudes are influenced by the perception of risks and benefits, knowledge and trust, and personal values (Lucht 2015). Some anxiety about the use of GM crops is driven by scare stories and pseudo-science provided by anti-GM opinion (McHughen 2013). The New Zealand economy is heavily reliant on export income from primary produce where the countrywide market perspective is to portray and clean green image (Tourism NZ 2009). Using GM crops and forages has been construed as the antithesis of a ‘clean green image’ (Edwards 2017). A recent review on consumer attitudes towards the use of GM crops and forages in New Zealand concluded that

while there will always be a proportion of consumers against the use of GM in food production, the published evidence would suggest that the use of GM plants in New Zealand for food production will have no long-term deleterious effects in overseas markets. (Caradus 2022b)

Indeed the consequences of GM crops grown for food and feed over the last 25 years have shown that ‘GM technologies like many non-GM technologies can bring risks, but these can and have been monitored and quantified, allowing decisions balancing commercial, societal and environmental benefits against measurable risks’ (Caradus 2022a). Another recent review (Bowerman et al. 2023) concluded that

the limited data available suggests that there may be some willingness to consider novel crops modified to be more resilient if they indeed benefit farmers (particularly those seen as family farmers as opposed to multinationals or corporates), and if the crop fulfils other beneficial desiderata such as reduction of agrochemical use or documentable environmental benefits, and if it is produced using gene editing rather than conventional GM.

This review also confirms that often in promoting the value of new gene manipulation technologies the social aspects and impacts are overlooked. Involving public engagement more in the process of delivering new technologies rather than simply seeking consent is a priority.

Suspicion of corporate multinationals

The majority of commercialised GM crops have been developed by a few multinational companies (Kumar et al. 2020). The attitude, real or perceived, of some of these large corporate multinationals in commercialising the first GM crops in the mid-1990s has tarnished the technology and has been a major driver of antagonism towards the use of GM crops for food and feed (Lewontin 2000; Patel et al. 2005; Amann et al. 2007; Clancy 2016). Those first GM crops offered either herbicide resistance or resistance to some economically important insect pests (Qaim 2015). Criticisms included:

• GM crops with herbicide resistance were developed so that corporates could continue to make a return on sales of the herbicides (Bonny 2008, 2011).

• The traits were all input traits that benefited the farmer with little concern for providing benefits to the consumer. Indeed ‘a new food technology may become an issue when consumers are convinced that this technology provides no additional value to them or to society and may only have advantages for producers and the industry’ (Siegrist 2008).

• Intellectual property protection on the technology would make it more expensive than conventional cultivars, and this cost would be passed onto the consumer. However, a European Parliament study has concluded that there is no monopolistic pricing power present in the delivery of GM crop technologies (Wesseler et al. 2015).

• Corporations developing and marketing GM technologies acted with arrogance and over-confidence (Sandler 2004).

• GM crop seed would be too expensive for poorer farmers (ETC Group 2013).

• Farmers were held hostage to the ‘agro-industrial machine’ due to their control over the use of GM technologies (Tirado and Johnston 2010).

• The argument that GM technologies are needed to secure future food production was viewed simply as a reflection of corporate interests (Jacobsen et al. 2013).

• Concern about corporate ownership of seed and threats to the purity of indigenous crops (Schmidt 2005).

• Poor communication, both in substance and means, on the balance of benefits and risks of this new technology was poor (Cook et al. 2006;Nicolia et al. 2014; Quemada 2022).

It has been determined that many of these concerns are still real with control of New Breeding Technologies being largely with a small number of influential companies (Clapp and Ruder 2020).

‘Playing with nature’

Risk perception can be partly driven by notions of what is seen as unnatural and immoral activities of a modern technology (Sjöberg 2000). This is largely a moral position related to the level of understanding of consequences and whether GM crops in agriculture are beneficial or harmful to those consuming the food or to the environment and therefore deemed acceptable (Gregorowius et al. 2012). GM crops and food are also viewed by some as not being ‘natural’ (Rozin 2005, 2006; Siegrist 2008). However, when considering the visibly significant changes made through phenotype selection breeding in some domesticated crops, for example in the development of modern maize non-GM cultivars, why is that now considered more natural (or certainly not criticised as unnatural) than comparatively small changes in genes and gene function made by genetic modification technologies? While that is the case now for hybrid maize, it was not always the case. Despite the superior performance of maize hybrids and their rapid acceptance by farmers (Sprague 1946) in the 1970s were blamed for social and environmental problems arising from industrialised agriculture (Curry 2022).

The precautionary principle and precautionary approaches

An early accepted definition of the Precautionary Principle was that it ‘seeks to impose early preventive measures to ward off even those risks for which we have little or no basis on which to predict the future probability of harm’ (Wiener 2001). It is based on the adage of ‘better safe than sorry’ (Margolis 1997). Alternatively, the precautionary approach is more about how a country, organisation or individual determines how they may evaluate the likelihood of specific risks in relation to the Precautionary Principle. As pointed out by Conko (2003), often the two terms are not differentiated as has occurred in the European Commission’s communication on the precautionary principle (European Commission 2000), and the Canadian governments environmental regulatory agency (Vanderzwaag et al. 2002). There is a view that precautionary principles should be motivated by ‘decision making under conditions of uncertainty (which) necessitate taking a precautionary approach to decision making’ on a case-by-case basis (Hartzell-Nichols 2013). Either way ‘the best elements of a precautionary approach demand good science and challenge the scientific community to improve methods used for risk assessment’ (Hayes 2005).

The Cartagena Protocol on Biosafety (Cartagena Protocol 2000) outlines the risk assessments required for the trans-boundary movement of modified organisms, and specifically places the precautionary approach (note the term Precautionary Principle is not used) as a fundamental concept when using genetically modified organisms. The precautionary approach in environmental law is outlined in Principle 15 of the Rio Declaration on Environment and Development152 (1992):

In order to protect the environment, the precautionary approach shall be widely applied by States according to their capabilities. Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation. (Rio Declaration 1992)

This ‘precautionary approach’ has had a significant impact on how genetically modified plants are regulated in a number of countries including the European Union and New Zealand, amongst others, ‘even though there is little evidence of serious or irreversible damage to the widespread use of these crops in the rest of the world’ (Conko 2003; NASEM 2016; The Royal Society 2016; Adenle 2017; European Commission 2021). Indeed the Royal Commission on Genetic Modification (Eichelbaum et al. 2001) concluded that New Zealand should keep it options open and ‘it would be unwise to turn our back on the potential advantages on offer, but we should proceed carefully, minimising and managing risks’. This is simply the precautionary principle explained. However, 12 years after the Royal Commission, ‘little progress has been made towards developing the public policy capacity necessary to make effective strategic decisions on GM crops in New Zealand’ (McGuinness Institute 2013), and even since then nothing has changed. The McGuinness Institute concluded that ‘politicians and public policy analysts are faced with an uneven and incomplete policy landscape, leaving them ill equipped to make sound decisions on New Zealand’s position regarding the release of a GM organism’.

In Europe, where the precautionary principle is also adhered to for all GM technologies (European Commission 2000), there has been a plea to use the ‘innovation principle’ where relevant risk assessment would be designed on a case-per-case base, to enable benefiting of gene-edited products while complying with relevant risks management (Jouanin et al. 2018). An innovation principle means ensuring that whenever policy is developed, the impact on innovation is fully assessed. The principle should provide guidance to ensure that the choice, design, and regulatory tools foster innovation, rather than hamper it (European Political Strategy Centre 2016).

Perceived absence of social, environmental, and economic benefits

While the economic benefits of GM crops have demonstrated considerable benefits to farmers, consumers, and the environment where they have been used (Caradus 2022a), in New Zealand, it has been stated that ‘there has been a clear absence of any commercial or other benefit to the New Zealand public’ of the investment to date, largely by the government, in developing genetically modified crops and forages (McGuinness Institute 2013). They went further on the basis that

New Zealand is not equipped to make a decision on the release of a GMO in the outdoors; however, we do consider there is sufficient evidence to make a decision on New Zealand becoming a dedicated GM-free food and fibre producer in the short to medium term.

The likely positive consideration of any application to EPA for a conditional release of a GM or gene-edited technology in New Zealand requires a convincing cost–benefit analysis. The onus is on science here to deliver that convincing analysis.

The same applies to ensuring that environmental and societal benefits are also espoused. Societal benefits will ensure that both the quantity and quality of food produced are sufficient for a growing world population (Hickey et al. 2019; Zaidi et al. 2019), while environmental benefits will ensure that quality of air, land, and water is improved. Many of the benefits resulting from the use of GM crops (Qaim 2020; Caradus 2022a) can also be attributed to future opportunities provided by New Breeding Technologies.

In two surveys, one of 555 articles from journals, company web pages, and web pages of governmental agencies which summarised 1328 studies/applications of genome-editing in model plants and agricultural crops in the period January 1996 to May 2018 (Modrzejewski et al. 2019), and the second of 62 plant breeding companies taken between January and May 2020 (Jorasch 2020) identified the types of traits being sought through using New Breeding Technologies (Table 5). Proponents of New Breeding Technologies propose that environmental benefits will result from improving crop performance in more hostile climatic conditions, while also mitigating carbon emissions; reducing the use of toxic synthetic chemistry; and increasing efficiency and productivity on-farm and in so doing reducing pressures on natural resources and benefiting farmer incomes, supporting both environmental and economic sustainability goals (Clapp and Ruder 2020).

Table 5. Percentage of traits sought using New Breeding Technologies by researchers (Modrzejewski et al. 2019) and crop breeding companies (Jorasch 2020).

Trait category	Survey
	Researchers and web pages	Crop breeding companies
Agronomic value	36	26
Food and feed quality	29	18
Biotic stress tolerance	16	23
Abiotic stress tolerance	5	15
Herbicide tolerance	8	5
Industrial applications	6	9
Other*	–	5

* Other includes –flavour, shelf-life, digestibility, ornamental value (flower colour), and post-harvest quality.

Misinformation

Misinformation defined as ‘false or inaccurate information, especially that which is deliberately intended to deceive’ abounds in society on a range of topics including GM crops and food. A review of mainstream and online news media on genetically modified crops and food, with a potential readership of 256 million, over a two-year period (2019–2021) found an overall falsehood rate of 9%, none of which was positive in sentiment (Lynas et al. 2022). They concluded ‘that misinformation about GMOs in the mainstream media is still a significant problem and outranks the proportion of misinformation in other comparable debates such as COVID-19 and vaccines’. The confusion created by misinformation not only impacts public attitudes but also affects government agencies, professional organisations, and the scientific community (Kabat 2017; Lelieveld and Andersen 2020). In a survey, it was found that ‘public attitudes to public confusion about food safety and health risks’ was the dominating factor (38% of respondents) related to the use of New Breeding Technologies in crops (Lassoued et al. 2020)

Misinformation in and about science is a real issue. This includes the predisposition to publish positive results and leave negative or non-significant results unpublished, to stray into unrealistic exaggeration, extrapolation and speculation, use citation bias (preferentially citing papers that support a claim over those that undermine it) or simply using citations incorrectly, sometimes unwittingly, to justify claims they do not in fact support, and inappropriate statistical analysis and/or interpretation (West and Bergstrom 2021). A very popular strategy for misinforming society is to spread doubts by referring to the uncertainty of scientific conclusions about pertinent issues of the day such as the health impacts of tobacco and the impact of carbon emissions on climate change (Oreskes and Conway 2010). The same could apply to attitudes towards GM crops and food. For example, it has been argued that an altered plant DNA sequence that does not exist in nature could be a health issue due to ‘many GM foods having some common toxic effects’ and some GM food can compromise immune systems (Dona and Arvanitoyannis 2009). Yet, the overwhelming evidence based on over 25 years of reporting and debate would indicate that this conclusion is unfounded (Caradus 2022a). However, this will not stop this statement being re-stated elsewhere.

Negative results

A number of publications have shown negative effects of GM crops and food, notably the ‘Monarch butterfly controversy’ (Losey et al. 1999), the Pusztai study (Ewen and Pusztai 1999), Seralini study (Séralini et al. 2014), and the more recent Shen study (Shen et al. 2022). There has been vigorous debate on each of these (refer to Caradus 2022a) and in some cases has resulted in retraction (e.g. the case of the Séralini study) and then defence of the original study sighting collusion and corruption (Qaim 2016; Novotny 2018). For the lay person, these results are confusing and possibly alarming, and certainly are difficult to explain in the knowledge that for more than 25 years GM crops have been produced and marketed on a commercial basis and have been consumed by millions upon millions of people and livestock, having gone through pre-market regulatory approval, and yet have shown no evidence of adverse health or nutritional effects in the general population (Kour et al. 2022).

Distrust of science

In an online survey, conducted in 2018, of just under 20,000 adults across 23 countries about the level of trust by the public in 18 professions, scientists were seen as the most trustworthy profession globally, followed by doctors and teachers, while politicians and government ministers were the least trustworthy (Ipsos 2019). However, despite that encouraging statistic scientists cannot be complacent. While scientists had the lowest level of untrustworthiness at 11%, that still indicates that about 10% of the population does not trust scientists or believe in science. But clearly, ‘science relies on public trust for its funding and opportunities to interface with the world’ (West and Bergstrom 2021). The widespread confusion about issues such as GM crops and food ‘contributes to an increasing public distrust of science since scientists are seemingly incapable of resolving these controversies’ (Kabat 2017). Distrust is driven by a number of factors including religious belief, level of education, political affiliation, socioeconomic status, and simply by the large number of controversies being publicly debated. But not all scientific controversies are of the same type. ‘On some issues, such as vaccines and GM crops, we have solid experimental and epidemiological data based on rigorous research, which enables us to make strong inferences about causality and the rate of adverse effects’ (Kabat 2017).

Some, but not all, of this misinformation and distrust of scientists and science can be corrected by reassessing career-related incentives associated with publishing, ensuring productivity metrics are a measure and not a target, addressing better ways of undertaking peer review and evaluating reference lists to reduce citation errors, and increasing the number of science reporters both inside and outside of science institutions to improve public engagement and understanding of science (West and Bergstrom 2021). Scientists have an important role in addressing misinformation by not ‘pretending that there is a scientific consensus on controversial issues when there is not’ (Kabat 2017). This requires those representing science to engage in transparent, open, and honest debate, examining without bias all available information, resisting exaggeration, and speculation so that a balanced and fact-based case is presented.

Engagement with European consumer experts and societal stakeholders to consider their perceptions, expectations, and acceptability of improving crops and New Plant Breeding Technologies for future-proofing the agri-food systems indicated that ‘governments to take a proactive role in regulating them, ensure openness and transparency in breeding new crop cultivars, and inform consumers about the effects of these breeding programmes and the risks and benefits of the new crop varieties developed’ (Nair et al. 2023).

GM crops and food concerns–real and justified?

Systematic reviews of intended and unintended consequences of 25 years of GM crops have concluded that perceived risks associated with GM crops are low to non-existent and that GM crops provide considerable benefits and will provide tangible solutions for many of the current challenges facing humanity, improving societal, economic, and environmental outcomes (Herman and Price 2013; Qaim 2016; Carzoli et al. 2018; Delaney et al. 2018; Ladics 2019; Louwaars 2019; Herman et al. 2020; Caradus 2022a; Vega Rodríguez et al. 2022). In addition, GM crops that provide food and feed are the most highly regulated biological technology in the world (DeFrancesco 2013; Baulcombe et al. 2014; Brune et al. 2021), which would indicate that risks associated with them are low and are significantly outweighed by the benefits.

Many of the New Breeding Technologies used to provide new trait opportunities for crop and forage plants result in products that are indistinguishable from those developed using alternative non-regulated techniques (Glenn et al. 2017) and, as a result, some will argue that are safe for use (Schouten et al. 2006; Schaart and Visser 2009), although others argue ‘that precision, cannot be considered an indication of safety per se, especially in relation to novel traits created by such modifications’ (Eckerstorfer et al. 2019). Criteria for categorising and evaluating cultivars produced from New Breeding Technologies have been proposed and include the rationale for plant breeding including the effect achieved, criteria concerning the method of application, the process, and the product (Lusser and Davies 2013). A multi-country (USA, Canada, Belgium, France, and Australia) assessment of consumers’ willingness-to-consume CRISPR-produced food compared to GM and non-GM derived foods, indicated that 56, 47, 46, 30, and 51% of respondents, respectively would consume both GM and CRISPR derived foods (Shew et al. 2018). However, an additional 15, 13, 10, 30, and 9% respectively would consume CRISPR-produced foods, compared with 8, 5, 6, 3, and 7% for GM-produced foods.

The question to debate is whether using a biotechnology tool is more unnatural or artificial and therefore riskier than earlier methods of manipulating the plant genome? Ethically for some, it is considered wrong to mix genetic material ‘unnaturally’ in a way that cannot occur in nature. But is it also wrong to create ‘unnatural’ genetic mixes not known to exist in nature but by a process where the genetic mix is within the same organism (Bruce 2017)? Non-GM breeding, where there is complete freedom to operate, results in this outcome as does some gene editing technologies, which in some countries is tightly regulated (Table 6). As others have stated ‘scientifically, it is illogical to regulate systems that produce single, very precise genomic changes, while others, such as chemical and radiation mutagenesis, which produces thousands of random mutations, remain unregulated’ (Mao et al. 2019). However, this is not an encouragement for regulators to now include chemical and radiation mutagenesis within their purview. Over the many decades that these forms of mutagenesis have been used, there have been no recorded issues of concern in commercialised cultivars (Oladosu et al. 2016). Further, the plant breeding process irrespective of whether it involves conventionally bred, genetically modified, or gene-edited crops provides multiple opportunities to eliminate adverse unintended effects resulting from any of those processes used to develop and select new or improved traits (Glenn et al. 2017; Louwaars 2019; Brune et al. 2021).

Table 6. Comparison of using chemical or radiation mutagenesis with CRISPR site-directed nucleases gene editing (derived from Lema 2021).

Comparator	Technique
	Ionising radiation or chemical mutagenesis	CRISPR site-directed nucleases
Control over gene sequences targeted	None	Tight. Breakage directed to specific sequences of interest
Knowledge on the trait-associated mutation	None–regulators not required to ask	Known from the start, part of the breeding strategy and the regulatory analysis
Unintended consequences–off target mutations	Significant –regulators not required to ask	Unlikely but regulators will ask
Possibility of finding off-target mutations	Limited –regulators not required to ask	Easy, and can be provided to reduce regulatory burden
Versatility, flexibility, and adaptability	Random and most often resulting in gene disruption	High –can be facilitated through gene disruption or deletion, but can also include gene modification or gene introduction

Concluding comment

Some have argued that with world population projected to be 10 billion by 2050 ‘that postponing technologies that can accelerate breeding makes no economic sense and should justify immediate adoption of accelerated breeding practices’ (Lenaerts et al. 2019). However, there are still misgivings due to an unwillingness to differentiate GM crops (generated by transgenesis) from genome-edited crops. Nine out of the top 10 countries by population (i.e. except Pakistan) having either paved a way or stated intentions to open up for easy use of gene-edited plants in commercial agriculture (Sprink et al. 2022). Many countries are now either not regulating or seriously considering not regulating SDN1 and, in some cases, SDN2 gene edits as genetically modified organisms except for the European Community and New Zealand (Buchholzer and Frommer 2023). A willingness by political decision and lawmakers is an essential element for any progress to be achieved in using the New Breeding Technologies for solving problems (causes and impacts of changing climate, malnourishment, and environmental degradation) currently facing society.

For millennia, human beings have manipulated the genetics of crops and forages, mostly without precision or knowledge of genetic systems, and without legal regulation. However, over the last 25 years with the advent of capability to introduce genes of known effect and with genetic methods that are increasingly precise and targeted regulatory systems has slowed delivery in many territories. Having said that balancing the risk and benefit of any new technological advance is important and required. But regulators must dispense with process-driven regulatory systems for more informed and nuanced product-based regulatory processes. The disconnect between the permitted consumption of foods derived from genetically modified crops but the inability to use genetically modified crops for animal feed (as occurs in New Zealand) is illogical and needs resolution. Similar inconsistencies exist in some territories (e.g. Europe) where importing genetically modified crops for animal feed is permitted but farmers are unable to grow these crops. Societal perspectives and public engagement about emerging technologies are crucial for their successful implementation (Bowerman et al. 2023). The science community has a responsibility to work with government and industry decision-makers to counter misinformation, and outline the social, environmental, and economic benefits of New Breeding Technologies through both listening and engaging in a constructive and empathetic manner.

Disclosure statement

The author is employed by Grasslanz Technology Ltd. which has a R&D investment portfolio that includes both genetic modification and gene editing of forages and microbes to provide mitigating solutions to current environmental and animal welfare issues facing both New Zealand and other pastoral economies.