Transitions to a legume-based food and agriculture

A summary with diagrams and photographs of an invited talk at the recent Conference on Advances in Legume Science and Practice organised by the Association of Applied Biologists in Glasgow 21-22 March 2018. Topics at the meeting covered a wide range of experience and disciplines from crop physiology, nutrition, molecular and traditional breeding, symbioses, landscape processes and food security.

Background – summary

Our invited presentation on Transitions to a legume based food and agriculture [1] introduced the aims and approach of the EU TRUE project, notably its central matrix consisting, first, of the quality chain from production through to markets and consumption, and second, sustainability, assessed through  economic, societal and environmental indicators.

The argument runs as follows. (A) Crops and their management alter the flows of energy and matter to various functions in the managed ecosystem. (B) Legume crops and forages have unique roles in channelling energy and matter to crucial functions related to soil quality, the nitrogen economy, pollinators and the production of plant protein. (C) To achieve a balanced and sustainable system, different types of crop, including legumes, need to be grown in planned configurations, whether within fields as mixtures, in  sequences or rotations and in spatial mosaics in the landscape. Practical designs need to consider those configurations that achieve the desired combination of functions.


Fig. 1 Faba beans Vicia faba: young crop, plant in flower, pods and fresh beans, and (small squares l to r) dried beans, flour, shelled split beans and bran.

Increasing legume production and output can be designed and managed in three stages. First, the area grown with existing legume crops such as field bean (Fig. 1)  can be increased with no change to the existing system. Second, the existing system can be modified – but not fundamentally changed – through (for example) mixed cropping of legumes and cereals, rhizobial inoculation of legume seed and  new legume products. Third, the system can be changed completely, with new crops, biotechnology and methods and untried configurations.

The presentation concentrated on  stage 1, but related work in Agroecology at the Hutton is already advancing in stage 2 though experimentation with crop mixtures, rhizobia and new products such as bread, beer and tofu made from beans [2].

Diversifying agriculture using grain and forage legumes

The flows of energy in production systems are investigated through a chain of effect linking interventions, such as agronomic management and choice of crop, through biota, including crops, to ecological processes which in combination satisfy (or not) desired higher level outputs [3].

The main crops in temperate Europe today are managed so that most of the energy is channelled to grain, oilseed or tuber yield (Fig. 2). In consequence,  other channels have been closed, or at least severely restricted, leading to long term declines in farmland wildlife and soil quality. Ultimately, such losses will feed back to limit economic output itself.

Fig. 2 The flow of energy in a winter cereal is concentrated into resource capture by the crop, then formation of yield and product, at the expense of trophic functions and soil.

The solution is to diversify the production systems of the region, in effect opening and regulating channels to other functions.  The diagram in Fig. 3 offers a highly simplified depiction of the wider balance of flows that should be realised in a forage legume.

Fig. 3 The flow of energy in a legume forage is distributed across a range of functions, notably N-fixation and trophic activity, e.g. through invertebrates.

The scope for diversification is being examined in this way for the case study of lowland Scotland. Grain legumes, mainly peas Pisum sativum and beans Vicia faba have been present from the neolithic and bronze periods and a wide range of forage species have been tried and grown over the millennia. Grains and forages are therefore being quantified as to their effect on flows such as represented in Fig. 2 and Fig. 3. Species are then modelling alone and in various spatial and temporal combinations to find optimum states.

Much can be learned from the way legumes and other crops have been grow in in the past, including in-field mixtures, often broadcast from a single ‘bag’ of mixed seed, such as mashlum, and temporal sequences, in some of which the legume and non-legume overlap (Fig. 4). Fields and sequences then combine to give additional properties at the scale of the landscape.

Fig. 4 Examples of crop diversification used traditionally in the region: B is a legume and A another crop (e.g. a cereal, root, oilseed, grass); C is an in-field mixture, such as mashlum, D a blocked, in-field mixture where the crops are separated, Ea a sequence or rotation, Eb a sequence in which some crops overlap in time (e.g. nurse crops and undersowings) and F a spatial configuration in a landscape.

The main problem facing the study was uncertainty in the locations within the region in which legumes appeared historically. However, crop census data, beginning in the mid-1800s is being examined to get the missing information.

First census of the mid-1800s

Legumes became integral to both crop sequences and forage mixtures in the Improvements era after 1700, but while some records suggest legumes occurred  1 in 4 years [4], there is little hard data on the areas grown with them compared with cereals such as oat and barley.

The 1700s and 1800s witnessed a phase of innovation and trialling of both grain and forage legumes, but for reasons that will be explored elsewhere on these pages, most forage legumes dropped out of mainstream usage with the exceptions of clovers and vetches, while grain legumes were reduced to various forms of pea Pisum sativum and bean Vicia faba.

The census of crops and grass in 1854 carried out by the Highland Society, covered most of Scotland and initiated a period of regular crop censuses which have proved invaluable in charting the phasing in and out of different crops. Data on the main crops [5], summarised for each of the old counties of Scotland (current up to to 1890), were transcribed from the 1854 records. Data were available for peas, beans and vetches: as an example, that for vetches is shown in Fig. 5, where the area of the circles, each representing an old county, indicate the relative area occupied by each crop. The circle out to the north-east represents Orkney and Shetland.

Fig. 5 Distribution of the vetches crop in 1854, sown areas represented by circles centred on the old counties of Scotland, superimposed on current administrative areas.

Beans occupied the largest area, followed by vetches and peas which covered similar areas. Most of the crops would have been grown for animal feed. Their combined areas were small, about 5% of that grown with cereals. Other sources specify that mixed forages, such as red clover, ryegrass and plantain, were also grown extensively, but no records are available of their composition and coverage. One of the recurring deficiencies of agricultural census is the classification of mixed forages as ‘grass’.

The distributions of vetches (Fig. 5) and peas were similar, both concentrated in the east and today’s central belt, but extending both south-west and north to Orkney and Shetland. That of beans was more concentrated in the east and centre.

Various crop census after the 1880s continued to show a similar distribution. When peas and beans were distinguished as to whether they were intended for human and animal consumption, those for human occupied a more restricted area in the east of centre.

Grain legume coverage today

Going forward 160 years, IACS data – from the EU’s Integrated Administration and Control System [6] – allows more precise definition of the current area grown with grain legumes.  There is still no data for grass-legume forages which must all classified under one or other of the forms of ‘grass’. Four types of grain legume are reported,  in decreasing order of area – beans for animal consumption, peas for human and for animal consumption, and least, beans for human consumption.

The total areas grown today are even smaller than the combined area of legumes in the 1850s. Maps of legume distribution after 2000 are in preparation. Examples  can be seen at the Living Field post Can we grow more vegetables? and further analysis of changes over time will be given later in these pages. However, the combined areas of the four legume types recorded tend to remain <2% and in some years near 1% of the total cereal acreage.

Low inclusion of legumes in a dynamic production ecosystem

The main conclusion so far is that grain legumes (pulses) were minor components of agriculture in the mid-1800s and have remained minor. Yet many aspects of the the crop and grass production systems in the region have been far from stable. For example, ‘root’ crops, mainly swede and turnip, covered large expanses in the late 1800s, but are relatively minor now, while of the cereals, oat was dominant in the 1800s and early 1900s but  supplanted by barley and now occupies less than 10% of the cereal area.

Recently, other crops have risen to much greater coverage than the legumes, notably winter wheat and oilseed rape in the later part of the 1900s. During all these changes, grain legume areas remained small or decreased.

One of the questions being examined is why legumes have occupied such low acreage in the region and whether and where they could be increased. Investigations of the phenomenon are continuing but one reported contribution is a greater reliance historically on clover and other legume forages for soil fertility.

There seems no particular reason, however, due to limitations of soil or climate, for the restricted area grown with legumes today within the eastern and central ranges shown in Fig. 5. Nor should it be assumed that increases will come only from existing crops. In response to CAP Greening measures, small fields of assorted legumes have appeared in the region.

That in Fig. 6 comprised three species of clover, the well know red Trifolium pratense and white Trifolium repens species, but also an unusual one, crimson clover Trifolium incarnatum which was once tried as a forage at these latitudes. A few plants of sainfoin were seen near the edge of the field, but it was not certain they were sown as part of the mixture.

Fig. 6 Legume forage, mainly of white, red and crimson clover: (top l c’wise) the field, young and older flowering head of crimson clover, sainfoin (image from plant in the Living Field garden) and  plants in an approx 0.5 m width of field (images by curvedflatlands).

Assessment by multi-attribute decision modelling

Opportunities for expanding the area of existing grain legumes are now being examined. It should also be possible to quantify potential savings of mineral nitrogen fertiliser and pesticide as the legume area is increased. The IACS data again provides the wherewithal, allowing us to assess not only which fields contained grain legumes in any year, but also which other crops were grown in the same fields in years before and after the legume. With knowledge of the crops grown in each field, nominal attributes can be assigned based on the pesticide and fertiliser applied to each crop as quantified from national surveys.

Each field can then be given a nominal agronomic ‘intensity’.  The reduction of intensity due to the substitution of an existing crop with a grain legume can then be calculated, as can the trade offs in the areas and output of other crops and products such cereal grain. The four current grain legumes offer plenty of scope for substitution, since some are grown with high-input crops, mainly winter wheat and potato, while others are grown with short-term grass and spring cereals.

Placing a value on each system, and then comparing systems, is facilitated by multi-attribute decision models (MADM) built in DEXi software [7]. A  part of the ‘tree’ structure of the current MADM is shown in Fig. 7. The interventions are shown to the right. They affect in turn the biota and ecosystem processes that determine a higher-level attribute, in this case the N loss in water leaving a field. The full MADM will include the wide range of attributes determining the economic, environmental and societal contributions of production systems.

Fig. 7 Part of a decision tree or multi-attrbute decision model built in DEXi software showing the way interventions combine in effect to influence field-scale attributes, in this case loss of nitrogen (N) in water.

Sites for expansion of legumes are therefore being selected on the basis that (a) they lie within an area, soil and climate in which grain legumes are or have been grown, and (b) they have a balance of crops very close to those fields that already include legumes in the crop sequence.

The aim is to quantify the benefits of legume expansion for the purpose of informing government policy and encouraging food and agriculture to use and grow more legumes. While concentrating on the grains at this stage, there is no reason why the approach cannot be extended to forages such as those in Fig. 8. More widely, the results will form a comprehensive study in the EU TRUE project of an approach to define long term trajectories of legume-based production systems and to extend those trajectories across Europe and farther afield.

Fig. 8 Legume species historically trialled in the region as forages: (top l, c’wise) sainfoin, milk vetch, kidney vetch, tufted vetch and white melilot,all grown in the Living Field garden (images by www.livingfield.co.uk)

Acknowledgement of funding

The main work summarised here and presented at Glasgow was funded as part of the EU TRUE project.  Background knowledge of the maritime production system of lowland Scotland was acquired with funding from Scottish Government (Rural and Environment Science and Analytical Services Division). The authors are based at The James Hutton Institute, Dundee UK.

Sources, references

[1] Squire GR, Iannetta PPM. 2018. Transition paths to sustainable legume production. Aspects of Applied Biology 138, 121-130.  Squire GR, Quesada N, Begg G, Iannetta P. 2018. Transition paths to sustainable legume production. Invited presentation at Advances in Legume Science and Practice, Association of Applied Biologists Glasgow UK, 21-22 March 2018.

[2] Links to bean beer on the Hutton website – Feed the world, help the environment and make great beer. Link to ‘tofu’ make from beans on the Living Field web site – Scofu: the quest for an indigenous Scottish tofu. See also Feel the Pulse.

[3] Squire GR 2017. Defining sustainable limits before and after intensification in a maritime agricultural ecosystem. Ecosystem Health and Sustainability, 3:8, DOI: 10.1080/20964129.2017.1368873

[4] Wight, A. 1778-1784. Present State of Husbandry in Scotland. Extracted from Reports made to the Commissioners of the Annexed Estates, and published by their authority. Edinburgh: William Creesh. Vol I, Vol II, Vol III Part I, Vol III Part II, Vol IV part I, Volume IV Part II. All available online via Google Books. Note from GS: Wight’s journals of his travels through the agricultural regions of Scotland present an unrivalled account by a farmer of the state of agriculture in the Improvements era.

[5] The agricultural census of the Highland Society, 1854, summarised by:  Thorburn T 1855. Diagrams, Agricultural Statistics of Scotland for 1854. London: Effingham Wilson. The Living Field web site has more on Thorburn  at Thorburn’s Diagrams.

[6] Integrated Administration and Control System, IACS on the EU web site.

[7] Decision trees, multi-attribute modelling and DEXi software at Marko Bohanec’s web site DEXI: a programme for Multi-Attribute Decision Making.

Contact: geoff.squire@hutton.ac.uk