Case study 1: improving phosphorus acquisition efficiency in common bean in South East Africa
Common bean (Phaseolus vulgaris ) has critical food security, economic and agroecosystem value throughout South East Africa (Wortmann, Kirkby, Eledu & Allen 1998; Beebe 2012). It is fundamental for food security at the household level, as well as economically important at the household, village and regional level. Furthermore, common bean has a key agroecosystem role as a nitrogen (N) fixer. Bean in South East Africa is grown either as a sole crop or intercropped with maize, in both cases usually in rotation with the latter. The primary constraints to production in typical low-input small-holder and low-external input cropping systems in SE Africa include a) low phosphorus (P) availability in highly N fixing soils, b) terminal and/or intermittent drought, and c) limited root growth and elevated root mortality due to acid soil complexes, pathogens and insects (Lynch 2019). These challenges are further complicated by irrigation systems, fertilizers and other soil amendments being inaccessible due to cost and transportation limitations.
In Mozambique, researchers from the Mozambican Agricultural Research Institute (IIAM) along with collaborators at the International Centre for Tropical Agriculture (CIAT) and The Pennsylvania State University targeted root traits to develop novel varieties that substantially improved small-holder crop production and benefits to local agroecosystems. In this context, increasing P acquisition efficiency, defined by the ratio of carbon investment to P recovered, was identified as a key cross-cutting element because it impacts multiple aspects of plant productivity and agroecosystem function (Lynch & Ho 2005). For instance, shallow root growth has an advantage for the acquisition of soil resources with greater availability in shallow soil zones, particularly P. Greater P acquisition is in turn related to increased shoot growth, biological nitrogen fixation and yield. With a larger shoot biomass shielding the soil, erosion is also reduced (Henry, Kleinman & Lynch 2009; Henry, Chaves, Kleinman & Lynch 2010).
Towards breeding of novel common beans for P acquisition efficiency in Mozambique, greater basal root whorl number and longer and denser root hairs traits were specifically targeted. These traits affect the zone of soil explored (Fig. 1a), they can be easily and rapidly phenotyped in the field with resources readily available in less developed countries (Burridge, Jochua, Bucksch & Lynch 2016), modifications of these traits do not require any change to the cropping system or farmer practices, and they have multi-scale impact. Basal roots are a class of roots specific to dicots with epigeal germination and emerge from the transition zone between radicle and hypocotyl (Burridge, Rangarajan & Lynch 2020). In common bean, basal roots emerge in whorls composed of four individual roots, due to its tetrarch vascular patterning (Fig. 1b; Miguel, Widrig, Vieira, Brown & Lynch 2013). Basal root whorl number (BRWN) is under genetic control and ranges from one to five, with two being the mode (Miguel et al. 2013). It can be phenotyped using low cost root roll-ups five to nine days after germination (Fig. 1b) and is associated with field phenotypes and field performance (Strock et al. 2019; Jochua, Strock & Lynch 2020). Greater BRWN increases the vertical zone of soil explored, enabling the efficient exploration of both deep and shallow soil. It may also afford a degree of root redundancy that ensures at least a few roots will survive pathogen or insect attack to contribute to both deep and shallow resource acquisition (Strock et al. 2019).
Root hairs are subcellular protrusions emerging from specialized epidermal cells on all root classes (Bhosale et al. 2018). Root hair length and density are under genetic control and have been related to greater P acquisition in multiple crops and experimental systems (Ma, Walk, Marcus & Lynch 2001; Yan, Liao, Beebe, Blair & Lynch 2004; Hanlon et al. 2018). Root hairs have minimal construction and maintenance costs making them an extremely efficient means to increase root volume and acquire immobile soil resources such as P (Lynch 2019). They are relatively easy to phenotype at the seedling stage using a low-cost root roll-up protocol (Vieira, Jochua & Lynch 2007). Substantial synergy between greater root hairs and basal root growth angle have been observed in common bean in which the advantage of longer and denser root hairs is magnified when paired with shallow basal root growth (Miguel, Postma & Lynch 2015). Taken together, the shallower basal root growth angle and longer and increased density of root hairs traits makes ideal selection targets for improving P acquisition efficiency (Fig 2A).
The selection strategy for improved P acquisition efficiency varieties involved extensive phenotyping of varieties and breeding lines in order to identify accessions possessing high BRWN as well as long and dense root hairs (Burridgeet al. 2019). These superior lines were then crossed with common varieties either already grown in Mozambique or to varieties with similar and desirable agronomic and market characteristics, such as seed colour and size. Several early generations of these lines were grown at the experimental stations with non-limiting conditions to select them for adaptation, grain quality, shoot architecture, disease and pest tolerance. Subsequently, seeds from F4 plants were evaluated for key root hair phenes with those possessing longer and denser root hairs being advanced. Targeted field-based selection of varieties took place first on multiple stations across Mozambique with single constraints, such as low soil phosphorus or terminal drought. Simultaneously, selections were made on more remote stations and farmer’s fields with multiple, interacting constraints. A final set of 40 varieties were included in a round of on-farm participatory variety selection using a tricot model in which each of many individual farmers received a random pack of three varieties (van Ettenet al. 2019). As a result, three new varieties of common bean, named Kufuna, Tiyela and Matina, have been released in Mozambique. Trials on station and on farmer’s field suggests these lines have yields roughly double the yield of local varieties when grown with or without P fertilizer (Table 1). This is a noteworthy achievement, in part because no new varieties had been released in Mozambique for over 12 years, and none of those were bred specifically for Mozambique.
To further promote and study how new varieties are shared and impact yield, food security and local economies, a pilot promotional campaign was organized in three different regions and compared to three regions that received the improved seed but not the promotional campaign (Fig. 1c&d). These improved varieties of local adapted and accepted material are available directly from IIAM, as well as from multiple private companies. Stocks of breeder and pre-basic seed are maintained by IIAM and are distributed to companies to produce basic seed. Village level farmer organizations receive pre-basic seeds to produce certified seed lots. In the 2019-2020 season, approximately 11 tons of seed was produced, 6.1 tons of which went to seed companies and the rest to farmers either directly or via various programs, including farmer organizations producing certified seed. In the 2020-2021 season, approximately 45 tons of seed is expected to be produced by IIAM. At typical planting rate of 80 kg/ha, 45 tons of seed converts to 3,600 hectares of improved P-efficient seed that will be planted by farmers in the 2021-2022 season, up from zero in the 2018-2019 season. This mixed public, private and farmer organization model promotes the maximum penetration of the new varieties and promotes long-term sustainability for both companies and farmer organizations. Other varieties of common bean are in various stages of the pipeline with some nearly ready for release and others in on-farm trials.
The accomplishments in Mozambique demonstrate the utility of a selection strategy involving trait-based selection of seedling root phenes, which was followed by on-station trials targeting specific abiotic constraints and soil types and then on-farm trials with multiple biotic and abiotic constraints and the use of local practices. Similar systems have proved successful and continue to offer great potential (Humphrieset al. 2015; Ryan et al. 2018; van Etten et al.2019). Other key legumes, namely cowpea, groundnut and chickpea have very similar food security, economic and agro-ecosystem roles across broad swaths of Africa. Lessons from common bean can be in large applied to these other legumes due to similar root system, phenology as well as seed saving, sharing and buying systems.