Innovative pulses for European temperate regions: a review

: In the Global North, there is an increasing interest in pulses both for their beneficial effects in cropping systems and for human health. However, despite these advantages, the acreage dedicated to pulses has been declining and their diversity reduced, particularly in European temperate regions, due to several social and economic factors. This decline has stimulated a political debate in the EU on the development of plant proteins. By contrast, in the Global South, a large panel of minor pulses is still cropped in regional patterns of production and consumption. The aim of this paper is to investigate the for cultivation of potential minor pulses in European temperate regions as a com-plement to common pulses. Our assumption is that some of these crops could adapt to different pedo-climatic conditions, given their physiological adaptation capacity, and that these pulses might be of interest for the development of innovative local food chains in an EU policy context targeting protein autonomy. The research is based on a systematic review of 269 papers retrieved in the Scopus database (1974 – 2019), which allowed us to identify 41 pulses as candidate species with a protein content higher than 20% that are already consumed as food. For each species, the main agronomic (e.g. temperature or water requirements) and nutritional characteristics (e.g. proteins or antinutritional contents) were identified in their growing regions. Following their agronomic characteristics, the candidate crops were confronted with variability in the annual growing conditions for spring crops in European temperate areas to determine the earliest potential sowing and latest harvest dates. Subsequently, the potential sum of temperatures was calculated with the Agri4cast database to establish the potential climatic suitability. For the first time, 21 minor pulses were selected to be grown in these temperate areas and appear worthy of investigation in terms of yield potential, nutritional characteristics or best management practices.


Introduction
There is an increasing debate about the importance of plant-based proteins and the diversification of protein sources for food or feed [1]. Plant-based proteins originate in several botanical families but are mainly concentrated in legumes. Despite the importance of grain legume, their production and consumption are declining worldwide [2,3] whereas the acreage dedicated to soybean is continuously increasing due to the intensification of livestock production [4]. Although grain legume diversity is very high from a worldwide perspective [5], these species often have local patterns of production and consumption in several developed and developing countries [6]. Worldwide grain legumes production increased by 15% between 2016 and 2017. This increase in production occurred in each continent, except America where production remained stable. Soybean production also increased during this period. This increase was mainly due to the largest producers North and South America. Based on the most recent available FAOSTAT data [7], there are four main pulses cultivated worldwide. Beans (all Phaseolus species included) were Table 1: World production of pulses in tonnes (T) and contribution percentage (%) of each continent (Faostat 2019) Minor pulses are produced and consumed locally worldwide, and there is no exchange between countries unless they are already cultivating and consuming these minor pulses [8]. This is more a consequence of food habits than because of pedoclimatic nor agronomic constraints.
However, the production of pulses has decreased during the last five decades [9]. Several factors have driven this decrease. First, the reduction of human consumption of vegetal proteins was negatively correlated to the increase in the standard of living of the population and the substitution by animal proteins [10]. Besides, less time is devoted to cooking, which is essential to reduce the impact of anti-nutritional factors of pulses, leading to a reduced consumption [11]. Second, the increasing import of cheap soybean meal to feed livestock from North and South America (2017) induced a reduction in the price of meat. The availability inexpensive proteins has supported the intensification of livestock production which, at the same time, has dissociated livestock breeding from crop production for feeding [12]. The separation of protein and livestock productions led to an intensification and a reduction of the number of cultivated crops, chosen for their economical performances. Moreover, as Watson et al. [9] argued, the yield increase of cereals in Europe since the 70s was not followed by a similar yield increase of protein crops, leading to an expansion of cereal acreage while protein crops remained less attractive for farmers. However, since 2015 this trend has ended in Europe probably because of both the increase in organic farming and the different policy instruments supporting protein crops [13].
European farmers, particularly those in arable systems look to diversify the crop rotations to more sustainable and resilient cropping systems [14]. One of the major groups of crops targeted for diversification is grain legumes [14, 15]. Several scholars have been interested in grain legumes because of their potential uses for food, feed and bioenergy/biomaterials but also because of the agro-ecosystem services they provide [2,4]. However, there is a general lack of knowledge and references regarding these grain legumes in terms of production potential in Northern countries and also about the diversity of their possible usage. Moreover, some scholars argue that the low adoption by European farmers of grain legumes compared with other crops is caused by low productivity and low profit margins, technological lock-in, and a low temporal yield stability [16][17][18]. Reckling et al. [19] have underlined that in Northern Europe, the grain legume yield is as stable that of other major spring crops e.g. spring cereals or rape seed. Yield stability is also enhanced in the other crops of the cropping system by grain legumes, as demonstrated by Petrova Chimonyo et al. [20] in maize-based cropping systems. Mawois et al. [21] have studied the trajectories of farms that increased the proportion of grain legumes in their crop rotations, highlighting three levels for their successful introduction: the stability of the supply (both as onfarm consumption and in the food supply chain), the benefits of grain legumes as the preceding crop and/or the involvement of farmers in peer' groups.
In this context, the increase in the use of grain legumes in European cropping systems can rely on two strategies. The first is to increase the current rate of cultivation of common pulses, e.g. lentils, chickpeas and common beans, in these countries, attempting to overcome the problems of lock-in at the supply chain level. This is the strategy that is currently analysed in the mainstream literature regarding European temperate regions and in some European research and innovation projects, e.g. PROTEIN2FOOD or LEGVALUE. A complementary strategy could be to provide alternatives in the panel of grain legumes potentially cropped or processed by identifying new species from worldwide grain legume biodiversity and creating new niches for these species and a renewed interested in grain legumes. This strategy can rely on the one hand, on the adaptability of grain legumes to new environmental conditions [22,23] and on the other hand, on the interest of the supply chain in new plant-based products, as occurred in the case of quinoa [24,25]. In this paper, our goal is to explore the latter strategy. Information on tropical grain legumes or minor pulses already exists in literature, however it is often limited either to one species [26,27] or to a comparison of several species on a national/continental level, e. . The aim of this review is twofold: on the one hand, it will provide insights about the agronomical and nutritional characteristics of a large panel of grain legume crops which are currently unknown and cultivated poorly or not at all in European temperate regions and on the other hand, it will explore the potential matching of these pulses in temperate European climatic conditions. As we are interested in grain legumes as food, in this paper we will use the terms 'pulses' and 'grain legumes' as synonyms and distinguish them when the legume destination is not intended as food.
The paper will be organized as follows: in the materials and methods section the first part will present a systematic review and the different variables considered for each species and the second part will describe how the main climatic characteristics of European temperate region were extracted from the Agri4cast database. In the results section we will summarize the main agronomic and nutritional (or antinutritional) characteristics of a set of pulses and some candidate species to be grown in temperate conditions, additional material are provided in Appendices A to D. In the discussion and conclusion section, we will discuss both the barriers to the introduction of these crops and the research and development needed to better evaluate their potential in European temperate regions.

Literature review and data collection
A systematic literature review was performed according to the 7 steps method formalized by the Centre for Evidence Based Conservation [37]. The method adopted for this study is illustrated in Figure 2. The first step was the identification of the research question. We wondered about the global diversity of protein-based crops, namely of pulses, that could support a crop diversification in European temperate regions. To answer this question, we looked for a general overview of worldwide pulses diversity and characteristics. To this end, we firstly carried out a general literature survey in the Scopus database according to the Scopus reference Guide [38] by selecting the advanced searches TITLE-ABS-KEY ((pulses AND 'grain legumes') AND crops) AND PUBYEAR > 1974 AND PUBYEAR < 2020 AND DOCTYPE (re) AND LANGUAGE (English) AND SUBJAREA (agri)) in the title, abstract and keywords with a focus on review articles, in the 1974-2019 time span in the Biological Sciences sub-database. We obtained a preliminary list of 231 reviews. Based on these reviews, we compiled a list of species to be checked. The eligibility criteria were to be cropped (not wild), used for human consumption and having a protein content in the grain equal or superior to 20. The species identified were double checked through a complementary search in three international databases: FoodData Central from the United State Department of Agriculture (formerly USDA National Nutrient Database for Standard Reference), Prota4U hosted by Wageningen University, and the International Network of Food Data System from FAO. A list of 47 species (common and scientific names) was finally established (Table 2). At this stage, since the information collected was too general on the agronomic and nutritional characteristics of grain legumes, we performed in a second step an in-depth literature review on the 47 previously selected species (Table 1). This second review was also performed in the Scopus database, focusing on scientific articles (therefore excluding reviews), in the 1974-2019 time span in the Biological Sciences sub-database. A grouped search including all the species was also performed in order to build a common database of minor pulses and deleting duplicates. We considered minor and major pulses as separate searches.  Table 1.
Example of the Boolean string for this second search for one species is TITLE-ABS-KEY ((ervil OR 'Vicia ervilia' OR 'Bitter vetch') AND crop*) AND PUBYEAR > 1974 AND PUBYEAR < 2020 AND DOCTYPE (ar) AND LANGUAGE (English) AND SUBJAREA (agri) AND NOT ('cover crop*' OR 'green manure' OR intercrop*). By selecting the subject area 'AGRI' we tried to avoid the number of papers in ecology or natural sciences focusing on medical or beneficial properties of some of these species. Moreover, limiting the search to 'crop' helped us to select those papers that refers to cultivated species (not wild) and grouping both the common(s) and scientific names guaranteed to catch a maximum of the papers referring to a specie. Finally, after a preliminary analysis of the database, we identified several papers dealing with the use of minor pulses as cover crops or green manure, along with their use in intercropping. As we were not interested in these topics, we excluded the papers including such words in the abstract, title or keywords. Even though we are aware that searching among the English literature limited the number of papers, we believed that this choice could ensure a common standard among the selected papers. Table 2 Retrieved papers in the Scopus database for the period 1974-2019 for a sample of cultivated minor and major pulses (in grey) with a protein seed content above 20%, individual researches. In grey, a comparison with the same search repeated for major pulses. FoodData Central (n = 1 ) Prota 4U-PROSEA (n = 12 ) World Agroforestry (n = 1) Technical institutions (n=12 ) Specific research on data missing Google scholar (n = 110 ) Studies included in review (Scopus n =132) + (Other sources n=137 )= (n=269) agroforestry potential or looking only for a specific characteristic in growth, agronomy, nutritional or anti-nutritional composition. For minor pulses with a limited or local use, nutritional characteristics were obtained from research articles, of which a mean was calculated (protein, oil, carbohydrates and fibre content) or an interval was indicated. The papers obtained from the Scopus searches for each pulse species were stored in a Zotero database, which contained a total of 17628 articles and 10730 of them were discarded as duplicates. Among the duplicates, there were often papers comparing two or several pulses. An exhaustive screening on titles was performed in order to eliminate outof-scope papers that did not provide useful information for our research (6661 papers). The most common reasons for deleting a paper from the list were the following: the name of a pulse was mentioned only as a comparison of one characteristic to another crop [40], the trial or analysis was performed in intercropping, thus evaluating the impact of the cover crop on the staple food yield [41], or there was no relation to the topic, e.g. papers dealing with biochemical [42] or genetic characteristics [43] or the use of pulse extracts as green fertilizers [44] or seed treatments [45] without any relation to pulse yield or seed quality. Papers with an historical or archaeobotanical approach [46] and those stating consumer preferences [47] or the impact of the feed on the animal's health [48] were also excluded. For some crops such as cowpea, or Bambara bean not all the articles found are cited in the present review, as for one topic such as resistance to a fungus or an insect dozens of articles could be found. Some cultivated legumes showed an interesting seed composition but they are not used as food. Brebra seed (Millettia ferruginea) was excluded from our analyses, even though the seed is edible. Other species from Lathyrus and Vicia with a protein content above 20% were used as food but are either no longer cultivated [49] or are cultivated on a few hectares of a single Greek Island for traditional dishes like Lathyrus clymenum [50]. Finally, papers dealing with consumers preferences for different varieties or for different cooking techniques were also excluded.

Common name
In the end, the final number of papers used for the review was 269.

Bibliographic corpus analysis
The file containing the Scopus extraction with information from 4082 documents assessed for eligibility was first analysed to describe the papers in terms of sources (journals), years of publications and origin. For these analyses, we calculated the absolute, relative and cumulative frequencies. Then, the corpus was uploaded and analysed using the CorTexT platform, already used for agronomic literature reviews [51,52], throughout the CorTexT Manager (http://manager.cortext.net/). A term-extraction algorithm was first performed using the title, abstract and keywords of each document to identify the 100 most cited terms. With those terms several functions have been tested in the default proposed form such as the contingency matrix, geospatial exploration and the network mapping to see the relationships between species, countries, journals and other agronomical terms.

Descriptive analysis
The aim of the descriptive analysis was to characterize both the agronomic and nutritional characteristics of the selected species. Both are essential to select candidate species diversifying cropping systems in temperate regions to be used in the food industry. On the one hand, for the agronomic characteristics, we analysed the environmental requirements (optimal, minimal and maximal temperature, average water requirement, duration of crop cycle), fertilizer inputs, potential yield, fixed nitrogen and symbiotic bacteria. On the other hand, for the nutritional characteristics, we analysed the nutritional profile (calories, protein content, carbohydrates, vitamins, minerals, micro-nutrients) and the most common antinutritional characteristics (phytic acid, total tannins, saponins, trypsin inhibitors, lectins and other antinutritional molecules).

Species selection for temperate regions
The most promising crops for arable systems in temperate regions were selected from the database mainly according to the environmental requirements. Focusing on arable systems, only annual crops that could rotate with other crops were considered. For a preliminary evaluation, we also excluded crops that require special harvesting techniques such as those that produce grains underground. For annual species we considered those having varieties with a maximum crop duration of 120 days, thus corresponding to summer crops in temperate regions. We also excluded species with optimal temperatures higher than 25°C or where those data were absent. We hypothesize for the annual crops a growing period from 1 March to 15 October. We could not consider the crop heat units to reach maturity, the sensitivity to day length and the precocity because of a lack of data in the literature. We considered as candidate species those that can potentially grow in continental European temperate regions under oceanic influence (by excluding temperate areas in Ireland) according to the European environmental stratification proposed by Metzger et al. [53]. Climatic data (1998-2018) were extracted on the Agri4cast database provided by the European Joint Research Centre (https://agri4cast.jrc.ec.europa.eu/DataPortal/Index.aspx) in the area between Latitudes 47.56° to 53.44° and Longitudes -0.017° to 7.28°. The spring sowing date (date to start accumulating Growing Degree Days [GDD]) was considered as the last of three consecutive days with daily mean air temperatures equal to or greater than 12.8 °C based on the Corn Heat Units Index used to characterize climatic regions [54]. The potential season-ending date (date to stop accumulating GDD) was considered as the earliest date when the daily mean air temperature dropped to 12 °C, excluding dates included in the 90 days following the sowing date. The sum of temperatures was calculated within those dates considering base temperatures of 6 °C and 10 °C. Given the lack of data on the grain legumes investigated, those climatic data were compared to the requirements of already cultivated spring crops such as corn or soybean, to establish a preliminary conformity of the criteria.

Descriptive analysis of the bibliographic corpus
The selected corpus on pulses was composed of 4081 documents. More than 50 % of the research was published since 2010 (data not shown). The observed bibliography shows that the interest in pulses has been increasing especially the last 15 years reaching more than 250 documents per year (data not shown). An analysis of the country of origin of the first authors show that affiliations from 116 countries are represented, of them more than 50 % are from two countries, Brazil and India. Figure 2 shows the countries of origin of the first authors of 75 % of the articles, and papers from European temperate areas are poorly represented. Finally, more than 600 individual journals are represented in the corpus, and 55 account for 50 % of the articles. Of these latter journals, mostly are in the Agronomy and crop science area, e.g. Crop Science, Field Crop Research or Experimental Agriculture, and have a partially a regional coverage, e.g. Indian Journal of Agricultural Sciences or the Australian Journal of Crop Science (data not shown). The matrix of contingence obtained from the CorTexT platform (data not shown) also confirmed that almost 40% of the retrieved papers were published in two Indian Journals (Indian Journal of Agricultural Sciences and Indian Journal of Agronomy) and more than 40 % on Phaseolus vulgaris (or common bean, data not shown). This was confirmed by the geospatial exploration which identified India as the main producer of references on pulses, ahead of for the other countries in the world. Sowing the large disparity that exists between main pulses and minor pulses in the litterature. Those terms are also entered into evidence on the network map of the 100 most cited terms in the abstract ( Figure 3). We can also identify other terms associated with the yield (height of the plant, seeds dry weight or leaf area), management practices (P fertilization) associated with the impact on other crops (like wheat or mustard) or on the pulse studied and resistance to biotic or abiotic stresses (drought, moisture, weeds, diseases and insects). These data also confirm the existing gaps in knowledge and research interests between the common bean, pea, other major pulses like cowpea, mung bean or urad and the others minor pulses.

Types of pulses and agronomical needs
Pulses are found worldwide, and some of them are adapted to temperate climates (kidney bean, lupins) ( Table 3, Appendix A) whilst a larger number can be found in tropical or subtropical areas (adzuki bean, cowpea) and a few can be found in both areas (soybean, broad beans). Crops such as the common bean that can grow under both temperate and tropical climate conditions are grown during spring-summer period and in cold (winter) periods in temperate areas, or at high altitudes in tropical areas [53][54][55]. In soybean, on the other hand, several maturity groups adapted to different pedoclimatic conditions have been developed [56,57]. Most of them have a crop length between 2 and 6 months, but it is highly variable even within a species. However, those crop lengths have been observed in tropical conditions allowing a high GDD and the exact GDD is not precisely defined in the literature. Some crops have longer cycles that can reach nearly a year (i.e. cowpea and sword bean [58]), and others can be cultivated as an annual plant (i.e. urad, tepary bean [59]) or as a perennial (i.e. pigeon pea [60], morama bean, Mucuna atropurpurea [58], and yam bean [61]). The perennial pulses could also be cultivated as annuals depending on the agronomical practices of the country. Despite the in-depth research conducted it was difficult to identify the intra species genetic diversity available that may be conductive to a larger potential for one species to spread to other climatic conditions or agronomical practices. In temperate climates pulses are mainly cultivated as main crops or in association with cereals [64], whereas in tropical countries, we can find associations with corn [65], coffee [66], cassava [67] and other diverse crops [68,69]. In the latter crop, little or no fertilization or pest management is required. Two pulses are cultivated not for theirs grains but for their tuber production, making the grains a co-product (yam bean and African yam bean [70,71]). Among the studied species, minimal temperature has been retrieved in 25 species over 57. Among them few minor pulses resist cold temperatures (Tarwi bean, pigeon pea, kidney bean and fenugreek). Among the major pulses five over six tolerate negative temperatures. The other pulses require positive temperatures during their entire life cycle, and in most cases, above 10°C. The optimal growing temperature is characterized in the literature only in 33 out of 47 species, but the temperature intervals are often wide. The maximal temperature is retrieved in 23 species over 47 but it will not be a constraint in temperate climates of Western Europe.

Pulses
Legume plants can fix atmospheric nitrogen (up to 550 kgN/ha in exceptional examples [60]. Table 3 and Appendix A describe crops requirements and potential yield. Most pulses are grown without adding nitrogen to the crop, even though a small quantity at sowing time can aid in proper development of the plant before the symbiotic relationship with nitrogen-fixing bacteria becomes efficient. It is also recommended that poor soil be fertilized with phosphorus and potassium. The requirements of common pulses are equivalent to those of minor pulses, the main difference is their adaptation to a temperate climate, as they are adapted to low temperatures during a part of their growth cycle. The water needs and fertilization requirements of both are similar. The data collected on the fixed nitrogen cannot be used to make a comparison between species because they were obtained in most of the cases in one single pedo-climatic condition or the methodology to assess it was not identical which may lead to big differences not related to the real capacity of a species to catch up nitrogen [72,73].

Nutritional characteristics
Pulses have an interesting nutritional profile (Table 4, Appendix B), because they have a high protein content with a mean value in our sample of 23 % (rice bean has the lowest content with 19.7 % [74] and tarwi [75] has the highest content, 51 %), a low fat content with a mean value of 3.8%, and a small percentage of saturated fatty acids. Pulses also provide fibre, vitamins and minerals which are important for human health. Table  and Appendix B show the nutritional characteristics of pulses, not taking into account genetic or pedoclimatic variability between species. Vitamins and minerals, when present they were included in Appendix B. It is interesting to note that the values of protein content in minor pulses were comparable to or higher than those of the common pulses already cultivated in North-West Europe. Other current and largely cultivated crops are also rich in proteins (Appendix C) but often they also have a high fat content making them less nutritionally valuable. Appendix C describes in the same way as Appendix B the nutritional characteristics of other protein rich crops cultivated or not in temperate regions. For cotton and rapeseed there are nearly no data concerning their nutritional composition as they are mainly used to obtain fibre and oil or oil and proteins respectively, with no direct use as food or feed. A common protein supplement, spirulina (Arthrospira maxima) was also included because of its interesting nutritional profile, highest content of proteins and third lowest level of oil content by comparison.

Antinutritional characteristics
All pulses contain one or more antinutritional compounds ( Appendix ) and although the concentration is under the lethal dose a regular or almost exclusive consumption may induce certain medical problems, such as the lathyrism caused by grass pea in European populations after the wars [76] that led to a law forbidding the consumption of grass pea grain or its derivatives in Spain [77]. Efforts have been made to breed new varieties with a reduced content of the neurotoxin [78].
Appendix reports data from research articles, that analysed one or several compounds. The other columns indicate the presence of other toxic molecules that are only present in some crops, even in a single crop, such as β-ODAP (beta-oxalyl-diamino-propionic acid) in grass pea or gossypol in cotton. Itching and velvet bean present the highest content of phytic acid with some varieties reaching between 53 and 57 mg/g [29] whereas the other pulses varied from 0.5 to 41 mg/g. In the case of tannins some varieties of African yam bean contained up to 18.1 mg/g [79], while other pulses contained from 0.01 [80] to 96 [81] mg/g. Saponins reached the highest values in some varieties of navy bean, but have lower values than some varieties of lentils or chickpea [82]. Trypsin inhibitors were found in high quantities in sesban seeds reaching 140 mg/100g [83]. Finally, lectins were less abundant in a few species of pulses, mung bean seeds contained the highest quantity with 15,8 mg per 100 g [84].
No pulse exceeded the lethal dose estimated at 50-60 mg/kg for phytates, 30 mg/kg for tannins, 2.5 g/kg for trypsin inhibitors, 50 mg/kg for lectins, 50-60 g/kg for hydrogen cyanide or 20 mg/100 g and 2-5 g/kg for oxalates [85]. Nevertheless, the presence of these antinutritional compounds may cause some troubles. Phytic acid chelates with several minerals limiting their bioavailability for the organism. Tannins reduce the absorption of nutriments and vitamin B12, reducing the efficiency of energy conversion [86]. Cooking or transforming these pulses inactivates the negative effects of these antinutritional factors on health or wellbeing [87].

Potential for minor pulse cultivation in European temperate regions
Analyses of the meteorological data showed that in the considered European temperate regions between 1998 and 2018, the total number of observed situations was 7456. The minimal duration of good cropping conditions was 121 days, and the average maximum was 140 days. The most frequent duration was 120-125 days (18.  By applying the criteria of crop needs and the possibility of reaching the GDD necessary to attain crop maturity, compared to other spring crops already cultivated in those regions, almost half the selected grain legume crops (21) appeared to be of interest for temperate cropping systems. Those species are the one listed in tables 2 and 3.

Conclusions
The present systematic review has shown the agronomic and nutritional interests of almost 20 pulses beyond the major ones which could be cultivated in Western European temperate regions. At this stage this possibility is theorical as there are several gaps in the literature. Firstly, there is a lack of physiological data, especially for minor pulses whereas for major pulses information could be retrieved but only for the current cultivation area. Particularly, data concerning the optimal and minimal growing temperatures, the GDD requirements to reach physiological maturity as well as the minimal water requirements are missing. Secondly, there is little information about soil constraints such as the soil structure, soil depth, pH or soil water availability. Pests and diseases are described for most pulses, even for those species with few references. Thirdly, beyond the species level, there is a lack of knowledge about the genetic diversity that could be useful to extend the cultivation area, as the characterisation that is retrieved in scientific literature is not sufficient to the best of our knowledge. Probably, in seedbanks, the information available will be more exhaustive for assessing certain traits such as crop length or disease resistance but this information will not be enough to establish the potential crop length in climatic areas other than those where genetic material has been collected. Among the 20 pulses with the highest cultivation potential in temperate areas, we have found that both their nutritional and anti-nutritional characteristics are comparable to those of the major pulses already cropped and consumed in these areas. Technological characteristics, processing potential and consumer preferences were outside of the scope of this review; however, all the retrieved species are already consumed as raw food or food components in different parts of the world. The potential of these species for other uses than as food, i.e. feed (grain or forage) or no food (bio-plastic) were also outside of the scope of this literature review but they can be considered worthy of investigation when contemplating the development of a new supply chain.
From these conclusions, we can highlight three main research perspectives when addressing the introduction of these pulses in Western European temperate areas. These are traditional perspectives when considering the introduction of a new crop to different environments, as occurred for example with quinoa. Firstly, a characterisation of the requirements of the pulses in a controlled environment is necessary to identify their growing needs in order to adapt the crop management to obtain the maximum potential yield. Then, multi-site and pluriannual field trials are required to establish the optimal pedoclimatic conditions and management practices. In parallel, a second perspective concerns the great effort that must be undertaken to study the existing genetic diversity in order to find the varieties that are most adapted to temperate conditions. This effort will require a large collaboration with seedbanks in order to characterise this diversity. Finally, a third perspective will concern a breeding program to improve the existing varieties. These three perspectives are a necessary first step before considering the adoption by local farmers and the introduction to cropping systems in temperate areas.

Data Availability Statement:
The datasets analyzed during the current study are available from the corresponding author on reasonable request.