Conserving useful plants for a sustainable future: species coverage, spatial distribution, and conservation status within the Millennium Seed Bank collection

A substantially rich diversity of the world’s recorded useful plants (UPs) is captured within the Millennium Seed Bank collection hosted by the Royal Botanic Gardens, Kew, at Wakehurst, UK with 13,598 species (34%) belonging to 3696 genera and 325 families. This constitutes just over half of the total accessions and one third of the species and covers all 9 continents, 8 realms, 14 biomes, 34 biodiversity hotspots and 175 countries. The most common beneficial use category within the captured diversity is medicinal, then species with environmental, material, and human food value. About 86% of conserved UP species have a ‘Least Concern’ conservation status but 8% are identified as globally threatened. The advantages of mutual, continued, and long-term partnership (e.g., Mexico) are showcased when conserving plants important for local communities and addressing conservation challenges beyond seed banking. However, the geographic coverage suffers from a lack of partnerships with some parts of the world. Also, a low number of accessions contributed from many countries means that insufficient native range is yet to be captured for many species. This is particularly the case for restricted or narrow distribution species from families or genera with a high incidence of recalcitrant or short-lived seeds. Future planning must tailor better to cover the spatial distribution patterns for individual species, thereby improving the number of accessions and geographical coverage for those captured. Moreover, seed biology research should characterise desiccation tolerance and develop complementary, novel conservation methods, such as cryopreservation, to strengthen conservation options for UP species.


Introduction
Natural capital, both living and non-living, includes biodiversity (plants, fungi, animals, etc.), ecosystems (forests, grasslands, mangroves, etc.), landscape (rivers, mountains, valleys, etc.) and resources (water, soil, minerals, etc.). Those elements with value to society can be designated as "Nature Contributions to People" (NCPs; sensu Diaz et al. 2018). Plants, the primary producers upon which most of our biodiversity depends, provide humankind with food, shelter, medicine, materials, fuel, and contribute to economies and livelihoods, through use of plant material in various industries (food and agriculture, medicine, pharmacology, pesticide, etc.). They play a critical role in ecosystem services by maintaining the structure of ecological communities, soil regeneration, regulating water and nutrient cycles, removing carbon dioxide-a greenhouse gas-from the atmosphere (Millennium Ecosystem Assessment 2005;Tellez et al. 2020;Dablin et al. 2021). They also have a religious, cultural, aesthetic and/or ornamental value for the wellbeing of our society (Milliken 1994;Martins et al. 2021).
The term 'ethnobotany' has been used historically for the study of plants used by Indigenous people to fulfil their basic needs, but its modern concept has roots in botany, chemistry, pharmacology, etc. In a broad sense, ethnobotany can be summed up by four words: humans, plants, interactions, and uses (Rahman et al. 2019). Useful plants (UPs) are those identified with at least one documented use to humans with an economic, ecological, social, cultural, or scientific value. At Level 1 of the economic botany data collection standards, beneficial uses of plants to humankind are allocated to 13 categories (Cook 1995). The World Checklist of Vascular Plants (WCVP) lists 350,968 species of tracheophytes with accepted taxonomic status (WCVP 2022). Approximately 40,239 of those species (11%), from 403 families and 6707 genera, are identified as UPs in the World 1 3 Major threats to biodiversity are escalating due to anthropogenic climate change, habitat loss, fragmentation, degradation, overexploitation, invasive species, unsustainable livestock ranching, and pollution (Corlett 2016;Dinerstein et al. 2017;Borelli et al. 2020;Tellez et al. 2020;Dablin et al. 2021;RBG Kew 2021a). More than 20% of the assessed vascular plants are threatened with extinction and it is estimated that one in five plant species are threatened (Brummitt et al. 2015;RBG Kew 2016;Bachman et al. 2017;Antonelli et al. 2020). Consequently, biodiversity conservation has become a global and national priority (DEFRA UK 2007;CBD 2012) and ex situ conservation facilities for wild plants have grown considerably (Li et al. 2010;CHABG 2011;León-Lobos et al. 2012;Hay and Probert 2013;Corlett 2016). Whilst Target 8 of the Global Strategy for Plant Conservation addresses the ex situ conservation of threatened species, Target 9 focuses on conserving the genetic diversity of crops including their wild relatives and other socio-economically valuable plant species, while respecting, preserving, and maintaining associated Indigenous and local knowledge (CBD 2012). Thus, the global need is to ensure adequate conservation in ex situ seed banks of both wild species and crop relatives.
Nature-based solutions (NbS; sensu Nesshöver et al. 2017) are actions taken by humans to protect, sustainably manage, and restore ecosystems, with simultaneously benefits to both nature and society. Plants or seeds constitute the main assets for NbS that support human wellbeing directly and indirectly through their ecosystem services (Mattana et al. 2022). Understanding how they can provide and support NbS is essential as this approach provides a better framework to manage the natural capital in nature sustainably with amenable human interaction (RBG Kew 2021b). This includes the role of humans in the stewardship of the planet's natural resources and the means of investing in NbS to restore, conserve, and rebuild a better natural environment to address the current climate and biodiversity crises. In situ conservation is known to be the optimal strategy for NbS to conserve biodiversity but conventional seed banking is identified as a valuable and complementary ex situ tool for integrated plant conservation (CBD 2012;Rodríguez-Arévalo et al. 2017).
However, conventional seed banking is limited to orthodox seeds, i.e., seeds that can be dried to low moisture content without damage and subsequently cooled and stored c. − 20 °C (Ellis et al. 1985;Dickie and Pritchard 2002;Probert 2003). For every 10% decrease in equilibrium relative humidity, or 1% moisture content or 5 °C drop in temperature the lifespan of orthodox seeds approximately doubles (Harrington 1960;Pritchard and Dickie 2003). In contrast, recalcitrant seeds cannot survive more than a very limited amount of drying and cannot be stored in a conventional seed bank. Most spermatophytes produce orthodox seeds, while 8% of the total taxa are predicted to bear recalcitrant seeds, which are predominantly of tropical affinity and mainly phanerophytes (Wyse and Dickie 2017;Subbiah et al. 2019). While seed banking appears reasonable for conserving most crops and CWR, most threatened plants and tropical moist forest species are recalcitrant and cannot be conserved in conventional seed banks (Wyse et al. 2018b). Therefore, cryopreservation may be the only resource to ensure their effective ex situ conservation (Li and Pritchard 2009;Hay and Probert 2013). Other factors can hinder the efficient and effective application of seed bank processes (collecting, processing, storage, and recovery), and these so-called "exceptional" plant (sensu Pence et al. 2022a, b) are providing an increasing focus for conservation efforts.
The Millennium Seed Bank (MSB), Royal Botanic Garden, Kew (RBG Kew) is a long-term seed storage facility for seeds of wild plants (Dickie 2018). Conceived in the early 1990s, the main aim of the MSB conservation mission is to safeguard worldwide wild plant diversity, prioritising useful and threatened species. The MSB project led by RBG Kew (now MSB Partnership-MSBP), initially focussed on the UK and Europe, 1 3 and developed into the largest ex situ seed conservation effort in the world for wild species (Smith et al. 1998). Through a global network of partnerships and portfolio of numerous projects with national and international focus, orthodox seeds of wild plants are collected and conserved in the country of origin and where possible duplicated at the − 20 °C long-term storage in the MSB vault.
The international partnership agreements follow best practice under the Convention on Biological Diversity (CBD; Cheyne 2003). Article 15 of CBD recognises and declares the sovereign rights of States over their natural resources and the access to genetic resources by other parties should be on mutually agreed terms with fair and equitable benefit sharing. Each MSB partnership is associated with a legally negotiated access and benefit sharing agreement which complies with local, national, and international legislations while respecting Indigenous and traditional knowledge. The agreements forbid any commercial use of the germplasm, their progeny, or derivatives without a separate agreement.
The conservation effort has mainly focused on dryland habitats as species adapted to hot, dry environments may have evolved longer lifespans in the dry state and are suitable for preservation in seed banks (Li and Pritchard 2009). As the conservation effort evolved, new collaborative projects emerged to deliver innovative solutions for not only conservation but also for agriculture, horticulture, forestry, and habitat restoration. Projects to support the sustainable use of plants by local communities to enhance their livelihoods have been of increasing interest. In this review, we do not intend to list all partnership projects that have been or are currently being implemented, as many are already summarised in RBG Kew website (https:// www. kew. org/ scien ce/ our-scien ce/ proje cts/ banki ng-the-worldsseeds). Rather, we focus here on the two major projects that were specifically designed to conserve UPs and/or CWR and that have been successfully completed.
(i) The 'MGU-Useful Plants Project' (MGU-UPP) led by RBG Kew during 2007 promoted the conservation and sustainable use of wild multipurpose plant species that are important for rural communities in sub-Saharan Africa (Botswana, Kenya, Mali, and South Africa), and Mesoamerica (Mexico) (Ulian et al. 2016). In each of these countries, the overall MSBP project was operated alongside MGU-UPP, but not specifically focused on UPs. In addition to various other UP species, the MGU-UPP specifically focused on the conservation and sustainable use of 110 multipurpose species suitable for agriculture and forestry activities (Ulian et al. 2016(Ulian et al. , 2019a Eastwood et al. 2022). An estimate of the potential value of the benefits derived from CWR traits of these priority crops alone could amount to about $120 billion (PwC 2013). Although global food security is being supported through crop seed banking, greater focus is needed on conserving wild species, such as CWR and underutilized UPs, as they harbour valuable functional plant traits and gene pools of great importance for the development of more productive, nutritious, and resilient crop varieties through research and plant breeding programs (Ulian et al. 2019c;Dempewolf et al. 2023). As demonstrated through the MGU-UPP, seed accessions of underutilized UPs at seed banks are a valuable resource for local communities, as these species have the potential to improve people's livelihoods, through the sustainable use of plants and their products.
There are various published books, articles and reports based on the germplasm of particular groupings of UP species captured at the MSB at project, country, or region level (Ulian et al. 2016(Ulian et al. , 2019a(Ulian et al. , c, 2020Rodríguez-Arévalo et al. 2017;Liu et al. 2018;Tellez et al. 2020;Crop Wild Relatives, Crop Trust, RBG Kew 2020;Müller et al. 2021;Eastwood et al. 2022). Liu et al. (2018) highlighted a rich natural capital value of MSB germplasm, as 49% of accessions and 32% of taxa conserved at that time were found to be useful to humankind. The objectives of our study here are to: (1) understand gaps in captured UPs in comparison to world's recorded UP diversity, their native ranges, beneficial uses to livelihoods, and vulnerability to extinction; (2) examine for uncaptured UP species whether their seeds possess traits which preclude storage in conventional seed banks; and (3) showcase the importance of building a long-term collaborative partnership to address challenges beyond seed banking.

Materials and methods
Data for seed accessions were extracted from the MSB Seed Bank Database (SBD, RBG Kew 2022b) on 16 May 2022. Data includes current taxonomic name with authorities and provenance from where seeds were originally harvested from (continent, country, and georeferences). Cultivated accessions inherited the geographic origin of the wild plants from which they were regenerated. To identify UPs, taxonomic names assigned to accessions were matched against WCUP.

Taxonomic diversity
To establish gaps, captured UP diversity was compared against WCUP families which were ranked according to their species totals (rank one has the highest and equal ranks have the same numbers). The plant lifeforms (Raunkiaer 1934) were compiled from Plants of the World Online (POWO 2022) and used to examine whether there were any biases in conservation that resulted in over-or under-representing certain growth forms. Seven main categories, based on the way plants protect their perennating buds during unfavourable seasons, were used: phanerophytes (mega, meso, micro and nano); epiphytes; chamaephytes; hemicryptophytes; cryptophytes (geo, helo and hydro); therophytes; and aerophytes. Subdivisions were attached to their main groups and descriptive information such as tuberous, succulent, parasitic, climbers, liana, biennial, etc. were removed. We retained the lifeforms described as bamboo and lithophytes.
To establish gaps, localities described as 'native' at TDWG Level 3 were extracted from POWO. The natural 'native' range was compared against accessions. Accessions with localities absent in POWO were assumed to represent an 'introduced' range. For each species, the coverage of 'native' range was calculated as a percentage. As conservation planning is usually undertaken within geopolitical units, we have provided a summary at country level for the total number of 'native' species captured against their naturally occurring totals. As it is not possible to describe all localities in POWO to boundaries of countries, some UP species were grouped at TDWG level 3. To understand the spatial distribution pattern for those not captured, we used a similar approach.

Beneficial uses
In WCUP, beneficial uses of plants are allocated into 10 categories instead of 13 . First, the number of species captured for each use category was compared against WCUP. Subsequently, species were grouped into different size classes by the number of different uses (1, 2-4, 5-7, 8-10) and their percentage representation was explored. The analysis in Liu et al. (2018) for CWR was repeated but using 94 of 99 crop genera in Maxted et al. (2013), as six of the genera are synonyms in WCVP (Fortunella, Lens, Lycopersicon, Pennisetum, Pisum and Vavilovia) and five of them have had their accepted names already listed.

Vulnerability to extinction
To verify species are vulnerable to extinction, their taxonomic names were matched against IUCN (2022) to extract species level global Red List assessments. Assessments described for sub-specific epithet levels were excluded.

Seed traits preclude conventional seed banking
As families with a high incidence of recalcitrant species are less likely to be conserved in conventional seed banks, for absent species, we inspected whether it is from a family described towards being recalcitrant in Dickie and Pritchard (2002). The incidence of recalcitrant can be predicted (Wyse and Dickie 2017;Wyse et al. 2018a) but for this instance we followed evidence-based approach using seed storage data from Seed Information Database (SID-RBG Kew 2022a). We also examined whether any missing species had been identified as "Exceptional Plants" in Pence et al. (2022b), as those require strategic ex situ conservation actions beyond conventional seed banking.

Challenges beyond seed banking
Mexico, the fourth most floristically rich country in the world with over 23,000 vascular plant taxa, has a tremendous biocultural diversity with many traditional uses of plants in 1 3 domestication, livelihoods, and conservation (Ulian et al. 2016(Ulian et al. , 2021Tellez et al. 2020). Between 5000 and 7000 species are used by humans. The vast majority (90%) of these are 'native' and often associated with a traditional use (Casas et al. 2001). We used the RBG Kew partnership with the Facultad de Estudios Superiores Iztacala (FESI) of the National Autonomous University of Mexico (UNAM), also MGU-UPP partner, to illustrate the benefits of a mutual collaboration to address challenges beyond seed banking. To understand the diversity and status of Mexican native UP flora, we compiled a list of 'native' species in Mexico by using distribution data from POWO and examined their beneficial uses using WCUP and global Red List assessments using IUCN (2022). For the captured diversity of Mexican native flora at the MSB, we used a similar approach for beneficial uses and IUCN assessments. In addition, we reviewed the current viability status of seed accessions by extracting seed viability data from the SBD for the most recent monitoring test round as a part of routine quality assessments. We used the highest viability recorded as an indication of current viability status of an accession. We compiled a list of relevant collaborative works to date where we can identify contributions (co-authorship, financial support, etc.) to the conservation or research programmes both between RBG Kew and FESI-UNAM and within Mexico. This is presented as supplementary material.

Results
Overall, 97,995 seed accessions capturing 39,989 species across 364 families and 6084 genera are conserved in the MSB. Taxonomic identities of 3339 accessions are either unknown, incomplete, have unpublished names or contain inaccuracy.

Taxonomic diversity
From world's recorded UP diversity, 13,598 species (34%) from 3696 genera (55%) and 325 families (80%) are captured in the MSB vault. Such UP species constitute 52% of accessions and 34% of species across 89% of families and 61% of genera. That is, one in two accessions conserved and one in three species captured are UPs. Of the 78 families and 3011 genera for which UPs are absent, many families (61) and genera (2539) are not represented at all at the MSB. A comprehensive summary of UPs conserved by each WCUP family is given in "Appendix". The highest number of UP accessions and species are found in the Fabaceae, Asteraceae, and Poaceae. Percentage of species captured within a family does not show a particular pattern (Fig. 1). Among the 10 highest ranked families, 40-49% of species are captured (high to low) for Asteraceae, Fabaceae, Poaceae, and Lamiaceae, 19-34% for Malvaceae, Orchidaceae, Rubiaceae, Apocynaceae, and Euphorbiaceae, and 8% for Arecaceae. Conservation efforts for five 'species rich' UP genera resembled the order of their species numbers, but Syzygium, which has recalcitrant seeds, is necessarily underrepresented (Table 1).

Geographic representativeness
Geographic origin is unknown for 1330 UP accessions from 689 species. The others represented all 9 continents and spanned 175 countries. Accessions come from 15 other countries, but none are UPs. The most UP diversity (species) or accessions contributed from Africa, Asia-Temperate, North America and Europe (Table 2). A comprehensive summary of UPs conserved by country is given in Table Sa   The realm and biome from where the seeds are harvested were mapped for 43,886 UP accessions containing 12,622 species. Accessions represent all eight world's realms (Fig. 5a), but most arose from (percentage of accessions and species, respectively) Palearctic (44%, 47%) and Afrotropic (31%, 52%), then Nearctic (8%, 23%) and Neotropic (7%, 14%) realms. The other four realms were represented by < 1-5% accessions and < 1-13% The gaps in the 'native' range at TDWG Level 3 for captured UP diversity were explored for 13,447 species (Table 3a). The main findings indicated that 50% of captured UPs had a relatively narrow range with 10% of being possible endemics. In contrast, 1% had a wider range, and the rest fall in between. Their captured range revealed ( Table 3b) that 10% of species came from a non-native range, 9% had cover over their full range and 59% covered only < 25% of their range. The rest covered 25-99% of range. An unrefined synopsis for the number of 'native' UP species captured against their naturally occurring totals at locality level is presented in Table Sc    Gulf of Guinea Islands, and South European Russia (lowest) (Table Sc-filter column M where 'native' UP species yet to be conserved is 100%).
The 'native' range for uncaptured diversity was explored for 25,959 species (Table 3c). We found that 30% are possible endemics, 42% have a narrow range, 27% have a moderate range and only 1% have a wider range.

Beneficial uses
The WCUP included 1040 to 26,638 species for each use category. For captured diversity, it varied from 636 to 10,005 species and most have a medicinal value, then an environmental use, are a source of material and have human food values (Fig. 6). Accessions followed a similar order but were more associated with a social use than a source of fuel. Within each use category, the percentages of species captured was relatively higher for invertebrate food, animal food, poison, and environment use (53-61%) than for fuel, human food, social use, gene source, medicine, and material (34-46%).
The Tropical and Subtropical Grasslands, Savannas and Shrublands biome ( Fig. 7) was found to have the highest number of species across seven use categories (animal food, fuel,

Vulnerability to extinction
Among UPs in WCUP (Fig. 9), 74% were found to be of LC-Least Concern. Five species are known to already be EX-Extinct, and 17% vulnerable to extinction (EW-Extinct in the Wild, CR-Critically Endangered, EN-Endangered or VU-Vulnerable). A similar pattern was observed for the captured diversity, with 86% of LC and 8% vulnerable to extinction including Brugmansia arborea, Franklinia alatamaha and Trochetiopsis (Melhania) erythroxylon which are designated as EW.

Seed traits preclude conventional seed banking
About 3987 (15%) of uncaptured UPs, represent a family with a high incidence of species bearing recalcitrant seeds as cited in Dickie and Pritchard (2002 bear recalcitrant seeds (Pritchard et al. 1999). There were two UP accessions for this genus capturing two species. We also identified 288 UP species that fit the category of 'exceptional plants' as described in Pence et al. (2022b) Challenges beyond seed banking The seed conservation partnership between RBG Kew and FESI-UNAM began in February 2002 in the Framework of RBG Kew's MSB Project and under an Access and Benefit Sharing Agreement. It initially focused on the conservation of seeds from wild plant from the arid and semiarid areas of Mexico, especially species that are endemic, narrowly distributed, rare, threatened, or economically important. The special focus on UPs began in 2007 through the "MGU-UPP". Integrated with applied scientific principles, this time the approach was holistic and participatory reaching a range of stakeholders including local communities and schools. In Mexico, the targeted communities were Guadalupe Victoria, Pueblo Nuevo (437 inhabitants The plant species were prioritised according to their importance (e.g., food, medicine, material) by involving local people who were then trained in seed conservation and plant propagation techniques during workshops (87) held in the three communities. These workshops were focused on seed management and plant propagation, as well in the preparation of some natural products and food conserves to help generate income at the local level. Research activities supported the training and the conservation and sustainable use of plant species. Thus, continuously students (bachelor, master's, and doctorate) were incorporated into the different research areas of the project, including plant physiology, phytochemistry, and ethnoecology studies. For instance, biochemical composition was analysed for Lippia graveolens, the 'Mexican oregano' to explore its medicinal properties; for example, hexane compounds (Hernández et al. 2009) and flavonoids (Moreno-Rodríguez et al. 2014). For Stenocereus pruinosus (Cactaceae), population genetics was compared in wild, managed in situ and cultivated plants to inform its conservation (Parra et al. 2008). The germination of Polaskia species (Cactaceae) was studied to quantify the impact of climate change on plant developmental processes (Ordoñez-Salanueva et al. 2015).
In 2015, in the framework of RBG Kew's "Global Tree Seed Conservation Program" (2015-2019) funded by the Garfield Weston Foundation, the focus moved to the conservation of endemic, protected and useful tree species of Mexico through the collaborative project "Science-based conservation of native tree species in Mexico" to support reforestation activities. It produced the most comprehensive database and catalogue of native trees of Mexico with information on species diversity, geographic distribution, conservation status and human uses (Tellez et al. 2020). It included the conservation of seeds of priority trees at the national level and the study of seed desiccation tolerance for seed banking and germination requirements of species, such as Cedrela odorata and Pinus spp., which are economically important native trees Sampayo-Maldonado et al. 2019).
Through the projects "Conserving native useful trees of Mexico to maintain its natural capital" (2019-2021) funded by the UK Newton Fund and "Using native trees important for local communities to enhance reforestation in Mexico" (2020-2027) funded by the Garfield Weston Foundation and the Aldama Foundation, the collaboration was enhanced and included the Mexican non-government organization (NGO) PRONATURA Veracruz A.C. (PNV). The work focused on the Veracruz State and developed an integrated conservation programme. The main components of the work included the selection of native tree species important for local communities, seed conservation of targeted native tree species, research on seed biology and ecology, tree propagation and distribution of seedlings, and reforestation trials to support community-based reforestation activities in the region (Rodríguez-Zúñiga et al. 2022;Sampayo-Maldonado et al. 2021).
There were 3793 accessions from Mexico in the MSB collection, capturing 8% of Mexican flora with 1916 species. UPs account for 58% of accessions and 50% species. This included 882 'native' Mexican UP species (20%) across 121 families and 478 genera. Contribution from FESI-UNAM Partnership included 3406 accessions, of which 57% are UPs, and 1713 species, of which 52% are UPs. Like the overall 'native' flora, the global conservation status for 382 species of captured UPs was 89% of LC or LR/lc-Lower Risk/least concern, 8% threatened (CR-1, EN-6, VU-23), 2% NT and 1% DD. The most frequent use was medicine (82%), then material (42%), environmental use (38%), human food (27%), animal food (21%), poisons (16%), fuel (11%), gene source (10%), social use (7%), and invertebrate food (6%). The viability of seeds was assessed at the MSB through germination tests and subsequent cut testing for non-germinated seeds for 1681 of these 'native' UP accessions (747 species). The viability levels recorded during the most recent round of tests was > 69% for 1404 accessions. This means that there are suitable propagation protocols for many of the species. For another 124 accessions, germination was 50-69%, but < 50% for the other 153 accessions. Comprehensive data for all these accessions are captured in SBD and is accessible through RBG Kew website. Over the past 20 years, in addition to seed banking as NbS to halt biodiversity loss, a multitude of other collaborative work has been conducted through the partnership, as summarised in Table Sd, e.g., conservation, research, etc.

Species capture
A substantially rich diversity of the world's UPs covering a wide geographic range is already held in the MSB collection. Overall, just over half of the total seed accessions and one third of species captured are represented by UPs. Nevertheless, UPs are under conserved in seed banks worldwide (Fielder et al. 2015;Castañeda-Álvarez et al. 2016;Khoury et al. 2019;Ulian et al. 2021). Moreover, about 81% UP species are represented by less than five accessions, below the threshold recommended for conservation programs 1 3 to capture sufficient genetic diversity. This trend is also observed for globally threatened species captured at the MSB and seedbanks elsewhere (Godefroid et al. 2011;Rivière and Müller 2017;Rivière et al. 2018;Liu et al. 2020), although the threshold for the number of accessions can be less for threatened or rare species (Brown and Briggs 1991;ENSCONET 2009). Although many UPs have a narrow distribution or are endemics, many others are more widespread and yet the native coverage is underrepresented. Thus, 59% of the species captured are represented from < 25% of their native range and further 10% are likely from a non-native range. Most edible plant species have a 'Not Threatened' conservation status ) and this is also true for UPs as only 17% of those assessed by IUCN are categorised as threatened. Nonetheless, of the threatened species captured, three species are already extinct in the wild. Taken together, future collecting programmes should target threatened UPs and seek overcome the currently limited intraspecific coverage across their ecological ranges.
As a result of the acquisition of new accessions, improved taxonomy leading to the renaming of species, the loss of accessions when their seeds are found to be non-viable, and the likelihood of capturing more UPs now than before due to WCUP, the current study captured 10,579 accessions and over 1000 species more than previously reported as UPs in Liu et al. (2018). Also, because WCUP is compiled from 13 databases that are widely known and/or accessible, any spatial bias in the WCUP has already been identified. For example, edible plants in tropical America are underrepresented compared to Africa ). Nonetheless, it is possible that more UPs could be captured in the MSB collections than our estimates.
The UPs identified so far are more likely the species that grow in close proximity to human habitation. The natural occurrence of UPs and their characteristics and beneficial uses we describe here do reflect a natural biological phenomenon. However, this phenomenon is impacted by the interaction and behaviour of humans with nature, because of the long history of use of plants for human survival and wellbeing. Starting from the need for food, shelter, protection, and remedies for injuries and diseases, humans eventually domesticated plants around their settlements. This lead later to the development of large-scale cultivation alongside human settlements. Ethnobotanical studies, including on specimens deposited in herbaria and museums, have revealed the relative human interests in a range of UPs (Nesbitt 2014). The same spread of interest in UPs is reflected in the overall conservation effort in the MSB partnership countries.

Native flora
Percentages of UPs captured by family, beneficial uses and lifeforms align with the species diversity described in WCUP and/or the incidence of species bearing orthodox seeds (Wyse and Dickie 2017;Subbiah et al. 2019; RBG Kew 2022a). Most families with a rich diversity of UPs contribute more to the conserved germplasm, but their individual percentages remain relatively low. Moreover, of the 'species rich' UP genera described in Diazgranados et al. (2020), some genera, particularly Syzygium (in the family Myraceae), are underrepresented. Syzygium tree species produce recalcitrant seeds (Pritchard et al. 1999), as do many phanerophytes (Wyse and Dickie 2017). Phanerophytes are the most frequent lifeform among UPs, and the greatest contributor of multipurpose species in this and earlier analyses (Ulian et al. 2019a, b, c). Although many phanerophytes are represented in captured UP diversity, 72% of species are absent, possibly because of the seed desiccation sensitivity trait. Of the uncaptured UP species or genera, at least 15% of species represent a family with a high incidence of bearing recalcitrant seeds. Moreover, 355 species certainly bear recalcitrant seeds and a further 43 species produce seeds for which there is evidence of a rapid decline in viability in dry cold storage. At the genus level, 113 of uncaptured UP genera have a high frequency of species bearing recalcitrant seeds, seven CWR genera bear recalcitrant seeds and three more have species with rapidly declining viability in seed banks. When you add in six multipurpose species bear recalcitrant seeds, at least 288 identified UP species can be described as 'exceptional' plants that have seeds that either are desiccation sensitive or that store poorly when dry in a conventional seed bank. Not only is there a need for multiple and diverse approaches to the preservation of seeds, as conventional seed banking does not suit all species (Li and Pritchard 2009;Dickie 2018;Pence et al. 2020), but also a requirement to characterise the seed biology of a wide range of species, particularly from the tropics (Visscher et al. 2022).

Crop wild relatives
Plant breeding is an ancient practice described as a coevolutionary process between human and edible plants (Breseghello and Coelho 2013). It dates to the early periods of agriculture when humans began to understand how yields could be improved and seeds from plants with the most valuable characters could be saved between growing seasons. As the continuation of traditional breeding tends to narrow the species' gene pool, over time crops become more vulnerable to biotic and abiotic stress (Breseghello and Coelho 2013). To counteract this risk, molecular tools for precision breeding by selecting specific genes have been developed to enable the breeding of more resilient crops. Undoubtedly, the UPs conserved in the MSB collections will be a valuable genetic resource in plant breeding activities to address future food security. This is particularly the case for CWR as many of their useful traits (disease or drought resistance) may have been inadvertently bred out of commercial crop varieties. The CWR project was a coordinated global effort to conserve the CWR gene pool as a basis of reequipping crops with resilient traits to cope with climate change, pest outbreaks and diseases. Overall, 4587 seed accessions from 355 CWR taxa containing 28 crop gene pools were collected, conserved ex situ, and duplicated in national and international germplasm banks, including the MSB, in more than 25 countries (Eastwood et al. 2022). The high number of accessions for CWR genera is explained by the focus of CWR project across many countries in five continents. Nonetheless, a recent analysis of the CWR project revealed that some gene pools are insufficiently captured (e.g., banana/plantain-Musa, potato-Solanum tuberosum, rice-Oryza, eggplant-Solanum and Sorghum), and some important centres of plant diversity are underrepresented, especially the Indian subcontinent (Müller et al. 2021). Plugging these gaps in CWR conservation remains an urgent need (Dempewolf et al. 2023) and the sharing of conserved CWR diversity for use in breeding programs, research, agriculture, etc., an ongoing priority (Crop Wild Relatives, Crop Trust, RBG Kew 2020; Eastwood et al. 2022).
The native flora and naturalised plants have also been valuable resources for sustainable use. Naturalised plants make a significant contribution to ecosystem services, economies, and livelihoods. This process started in prehistoric times, when human-mediated dispersal events contributed to the distribution of UP species (Roullier et al. 2013). More recently, many economically valuable species or 'cash crops' such as tea, coffee, rubber, cotton, cocoa have been introduced to non-native localities and have become primary contributors to country economies (Saraiva 2016;Rahman et al. 2019). Also, there are many instances, where 'introduced' species are widely used among communities for their livelihoods (Roullier et al. 2013;Rahman et al. 2019). With increasing concerns about the loss of plant diversity due to anthropogenic causes, ongoing conservation of a widest range of UPs remains important.

Geographic and country representation
The geographic representation by continent and realm is influenced by several factors: different phases of seed conservation projects led by RBG Kew (initially UK and later the rest of the Europe, especially the Mediterranean region); by regional foci (e.g., Africa); and by a preference for dryland habitats which more likely harbour species with desiccation tolerant seeds. Therefore, captured UP diversity or accessions are mostly represented from Africa and Europe or Palearctic and Afrotropic. The Tropical and Subtropical Grasslands, Savannas and Shrublands biome and Temperate Broadleaf and Mixed Forests biome found to be enriched with many UP species. As the conserved germplasm has shown a bias towards Africa and Europe, we do not imply that the natural incidence of UPs across realms and biomes remains the same.
The UK national program started in 1995 and by 1999 had conserved c. 98% of the native flora. Thus, most of its UPs was captured by that time. In 2013, the UK National Tree Seed Project was launched, aiming to conserve the genetic diversity of over 70 native trees and shrubs using a sampling strategy to account for ecological factors and biogeographic zones (Kallow and Trivedi 2016). Thus, UK native tree species that are also UPs are represented by high numbers of accessions (> 50-96 accessions each). The other countries that rank highly are where RBG Kew has continuing and long-term conservation partnerships (Australia, China, Israel, Madagascar, Tanzania, and USA) and/or with specific independent in-country programmes for conserving UPs (Israel for CWR). Federal countries such as Australia and USA have multiple partnerships with RBG Kew at the state level, and this is reflected by the conservation efforts. Syria also ranks highly, due to the receipt of accessions from other long-term storage facilities elsewhere in the UK and overseas.
The partner organisations decide on priority species suitable for conservation, some choose to collect widespread species for use in habitat restoration projects while others concentrate on specific species such as endemic, threatened or species with economic value or useful for livelihoods. Consequently, the number of accessions collected varies with country with about 60% of the countries contributing relatively low numbers of accessions (1-97 each). It is also the case that some geographic coverage in parts of the world with high UP diversity is insufficiently represented because of the lack of partnership countries (e.g., Côte d'Ivoire, Congo, Gabon, Benin, Paraguay, French Guiana, Albania, Togo, Suriname, El Salvador, Eritrea, etc.). This also means that the taxonomic representation of captured diversity is variable. Whilst this could be addressed in future collaborations, the prioritisation of taxa inevitably depends on the terms and conditions set out in the legal partnership agreements, that cover the political, legal, and administrative requirements, the level of implementation of the CBD, diversity of national flora, infrastructure needs, skills enabling collection, handling and storage of seed, funding, etc. (Cheyne 2003;Smith 2007;CBD 2012).
The top five countries contributing to UP conservation (accessions or species) are Burkina Faso, Botswana, Mali, Kenya (also by CWR), and South Africa, but the reasons for their position vary. Burkina Faso ranked highest for conserving the percentage of accessions and species as a result of a concerted conservation programme prior to 2010, which was then strengthen by further funding. A similar UP conservation focus applies to the MGU-UPP countries Botswana, Mali, Kenya (also by CWR), and South Africa. Mexico, another MGU-UPP partner, ranked seventh with similar numbers of accessions and species but much less UP coverage by percentage. Megadiverse and floristically rich large country such as Mexico are more likely to be underrepresented than a small country with less diverse flora. The knowledge on Mexican flora (diversity, distribution, and ethnobotanical uses) is still found to be largely fragmented (Tellez et al. 2020). Overall, seeds of 2598 wild plant species (> 10% of the Mexican flora) have so far been conserved in Mexico (Ulian et al. 2021) and this included 19% of 2885 native tree species (Tellez et al. 2020). Various projects continue to build in-country capacity for research and conservation in relation to global challenges. This demonstrates the importance of partnerships to achieve common goals in seed banking and beyond, such as tackling the wider Sustainable Development Goals.

Participatory approach
The conserved germplasm represents an asset for NbS at species and ecosystem levels. However, it is crucial that information, knowledge, and research outputs generated from conserved germplasm as well as from other strategic outputs related to plant conservation need to be made accessible to a wider scientific community and the public, including policy makers, through public presentations, scientific articles, data portals, social media and the websites to reach different audiences and raise awareness about the importance of biodiversity conservation and how it is essential for human wellbeing.
A key step in achieving this ambition is to start by working more equitably in partnership, which leads to a better understanding of the necessities of plant conservation and development needs around the world to build a sustainable future. Much of the success of the MGU-UPP emanates from the involvement of rural communities in prioritising native UPs for plant conservation, propagation, and planting activities. This included the co-design and co-delivery of conservation and research activities and the valorisation of native UPs and plant products for revenue generation at the local level (Ulian et al. 2016(Ulian et al. , 2019a. Overall, 1598 seed accessions have been made, of which 1271 have been banked in-country and 910 duplicated at the MSB. About 504 seed accessions have been evaluated at the MSB for quality, through research on their germination requirements. A further 357 species have been propagated in 26 local communities (38 community groups) and 39 schools, with 76,389 seedlings of 267 species planted and maintained in community gardens for conservation and use. Research activities were carried out on 297 UPs to generate knowledge and support their conservation and sustainable use. The project was not limited to plant conservation but delivered extensive training and supported the development of NbS to address development challenges in the rural communities (Ulian et al. 2016). For example, in Botswana, the focus was on conserving edible plants to support food security and developing food products to help generate income at the local level; in Kenya, the project trained small-scale farmers in tree seed management and propagation techniques for seedling production to support reforestation activities at the national level; and in Mali, worked with traditional healers to grow important medicinal plants to support local health and with organic cotton farmers to produce native pesticide plants to improve their yields without using chemicals. Finally, the project raised awareness on the importance of indigenous plants species both at the national and local level and UPs gardens were used also as a tool for education purposes and public engagement.

Recommendations
The principles of working more equitably within partnerships has led to a better understanding of the necessities of plant conservation around the world. The participatory approach involving local communities in plant conservation focusing on plants important to their livelihoods has been a success and has enable to address local needs. Whilst a considerable number of accessions of UPs have been introduced to the MSB collection over the last five years, the overall increase in species diversity captured is relatively less than during the earlier period of the MSB project. The most uncaptured UP diversity tends to have restricted or narrow distribution pattern, come from families or genera with a high incidence of bearing recalcitrant or short-lived seeds. A more robust, and tailored seed collecting strategy is needed based on filling gaps in spatial distribution patterns and native ranges of individual species with greater understanding of seed desiccation tolerance. Also, an important aspect is the improvement in the number of accessions and geographical coverage for those UPs already captured so as to widen the genetic diversity of the species. Beyond strengthening existing partnerships, the drive to establish new partnerships from parts of the world where UP conservation has been limited should be accelerated. Finally, alongside complementary conservation methods, we must establish novel seed banking techniques, including widening the application of cryopreservation, and explore gene banking of non-seed material for the world's most threatened UPs to halt their loss.

Table 4
A comprehensive summary of UPs captured at the MSB by family: the representativeness of each family was estimated as the percentage of UP genera and species captured against their totals summarised from WCUP (           Percentage conserved: Doubtful data were excluded from the analysis. The families were ranked according to the total number of UP species in WCUP. The families with the same number of UP species were received equal ranks and rank one has the highest number of species