Genet Resour Crop Evol https://doi.org/10.1007/s10722-020-00886-8 (0123456789().,-volV) ( 01234567 89().,-volV) RESEARCH ARTICLE Diversity of grasses (Poaceae) in southern Africa, with emphasis on the conservation of pasture genetic resources M. Trytsman . F. L. Müller . A. E. van Wyk Received: 12 June 2019 / Accepted: 16 January 2020 Ó Springer Nature B.V. 2020 Abstract A renewed interest in the present state of genebanks conserving pasture genetic resources worldwide motivated this study to quantify the wealth of grass (Poaceae) diversity indigenous to southern Africa, here defined as South Africa, Lesotho and Eswatini (previously Swaziland). Botanical occurrence records were extracted from BODATSA and PHYTOBAS datasets to generate a list of grass species indigenous to the study area. The phylogenetic classification, growth form, photosynthetic pathway, grazing status, endemism and conservational status attributes were added to the 43,889 species level records, sourced from published literature. Results from the current study indicate that the subcontinent is represented by eight subfamilies, 25 tribes, 151 genera Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10722-020-00886-8) contains supplementary material, which is available to authorized users. M. Trytsman (&)
F. L. Müller Agricultural Research Council-Animal Production Institute, Lynn East, PB X05, Pretoria 0039, South Africa e-mail: mtrytsman@arc.agric.za A. E. van Wyk Department of Plant and Soil Sciences, University of Pretoria, Pretoria 0002, South Africa A. E. van Wyk National Herbarium, South African National Biodiversity Institute, Pretoria 0001, South Africa and 685 species, inferring that only 20% of the world’s grass genera and 6% of world’s grass species are found in the study area with Panicoideae the most speciose subfamily. Paniceae is the only tribe with large numbers of both C3 and C4 species and with several species of high grazing value, therefore, was suggested as a priority lineage in the collection and conservation efforts of the South African National Forage Genebank. This genebank conserves at present 73 genera and 162 indigenous grass species, i.e. 48% and 24% of the total number of taxa respectively, denoting the current vulnerable status of grass genetic resources in southern Africa. A need to therefore collect and conserve grass genetic resources is emphasised, with greater focus on the conservation of seed of wellknown pasture genera classified as endangered or possibly extinct (mainly Panicum L. and Secale L.). Keywords Endemism
Eswatini
Gramineae
Lesotho
South Africa Introduction Poaceae (Gramineae), the members of which are commonly referred to as grasses and bamboos, is considered as probably the most valuable plant family to humankind (Bouchenak-Khelladi et al. 2010). The family includes the economically important cereals, 123 Genet Resour Crop Evol sugar crops, reeds, bamboos, forages and lawn grasses (Hodkinson 2018). The success of this family can be attributed to, amongst others, its adaptability to most ecosystems, including arctic regions and at high elevations uninhabitable by flowering plants (Tzvelev 1989), ecological dominance in many ecosystems, and high species richness (Linder et al. 2018). The important role of the Poaceae in sustainable livestock production is well known, with several genera housing important pasture species (Truter et al. 2015; Capstaff and Miller 2018). The exploration of the potential of southern African grasses for pastures began as early as the 1900s, with the past 50 years described as a period where the function and value of southern African grasses were studied by several pasture researchers (Truter et al. 2015). An important initiative towards the conservation of grass genetic resources in southern Africa was a collection excursion specially arranged for this purpose, to the Kruger National Park in South Africa, during the early 1990s. This ensured that selected ecotypes of important pasture species such as Anthephora pubescens Nees, Cenchrus ciliaris L., Chloris gayana Kunth, Cynodon dactylon (L.) Pers., Digitaria eriantha Steud., Eragrostis curvula (Schrad.) Nees and Panicum maximum Jacq. were conserved in the South African National Forage Genebank (SA-NFG) (Kruger et al. 1993). However, similar to trends in the international genebank community, especially those housing tropical and sub-tropical plant genetic resources (Maass and Pengelly 2019; Pengelly and Maass 2019), the conservation of these plant genetic resources have been under threat for the last 20 years. Generally, the funding to manage and maintain forage genebanks around the globe is on the decrease, resulting in many important plant genetic resources potentially being lost (Maass and Pengelly 2019). In the case of the SANFG, the plant genetic resources maintained at the facility was mainly threatened by the lack of funding which resulted in unreliable storage and testing facilities, coupled with a decline in trained personnel capable of maintaining and evaluating the valuable collection of plant genetic resources housed at the facility. Also, the SA-NFG houses a large number and diversity of plant genetic resources, many of which currently are perceived to have minimal to no agronomic potential due to the lack of information regarding their pasture potential. These species were collected with future breeding in mind. 123 The decline in the pasture breeding capacity in South Africa (Truter et al. 2015) has resulted in many of these species not being characterised and evaluated for their pasture potential. Therefore, due to the lack of information regarding their pasture potential these species have not yet been included in breeding and evaluation programs. These plant genetic resources are, however, still important sources of genetic material that could, under future bioclimatic conditions become valuable resources for breeding of future pasture species adapted to specific agro-ecological conditions. As a result, these plant genetic resources are maintained in perpetuity at the SA-NFG as a means to maintain the plant genetic diversity that could potentially be beneficial under future bioclimatic conditions. Maintaining and conserving these plant genetic resources however, puts significant financial pressure on the SA-NFG. Recently, Pengelly and Maass (2019) and Maass and Pengelly (2019) called for the prioritization of germplasm as well as to improve the efficiency in conserving current collections of plant genetic resources housed at genebanks across the globe. This, in turn, is believed to reduce the financial burdens on genebanks, by shifting efforts to only priority species, and locally adapted varieties and ecotypes within important pasture species with known agronomic potential (Pengelly and Maass 2019; Maass and Pengelly 2019). This will allow for the conservation of a larger genetic diversity within species with known agronomic potential, which, in turn, will benefit future breeding programs of these species. An example of this can be found in domesticated African grass species such as Sorghum bicolor (L.) Moench and Pennisetum glaucum (L.) R.Br. that are currently maintained and conserved for at least twothirds of their diversity of their wild relatives (Buckler et al. 2001). Van Wyk (1995) emphasized that the recognition and interpretation of genetic variation in organisms is at the heart of taxonomy and called on plant taxonomists in Africa for the urgent naming of infraspecific units. A striking example that highlighted the importance of this appeal was the revision of Agrostis eriantha Hack. var. planifolia Goossens & Papendorf being reduced to synonymy under A. eriantha Hack. (Victor et al. 2012). A method to prioritize taxonomic revision of South African plant genera was developed by Victor et al. (2015) to reduce the taxonomy-conservation disorder where indicators Genet Resour Crop Evol such as revision dates, insufficient data (including taxonomic uncertainty) and endemism were proposed as indicators. Furthermore, a literature review done in the early 1990s showed that only a few key or important indigenous grass species had complete autecological studies with the exception of Themeda triandra Forssk., Eragrostis curvula and Digitaria eriantha (Shackleton 1991). A need for a coordinated, systematic approach in basic ecological research of grass species in different biomes was therefore suggested. The most recent taxonomic review of genera in southern African grasses was within the temperate genus Helictotrichon Besser s.l. (Mashau et al. 2010) where two new species were identified. Germishuizen and Meyer (2003) recorded major name revisions within the genera Aristida L. [= Stipagrostis Nees], Danthonia DC [= Merxmuellera Conert; = Karroochloa Conert & Türpe; = Chaetobromus Nees; = Centropodia Rchb.; = Dregeochloa Conert], Phragmites Adans. [= Tribolium Desv.] and Rhynchelytrum Nees [= Melinis P.Beauv.] whilst Pentaschistis (Nees) Spach and Setaria P.Beauv. were revised at species level. However, Fish et al. (2015) indicated that the genera Agrostis L., Anthoxanthum L., Cymbopogon Spreng., Cynodon Rich., Echinochloa P.Beauv. and Puccinellia Parl. are also in need of revision. Remarkably, a new species of Enneapogon Desv. ex P.Beauv. was recently identified by Mashau and Coetzee (2019) i.e. Enneapogon limpopoensis Mashau, being possibly endemic to South Africa and Zimbabwe. The classification system followed in earlier contributions on the identification of southern African grasses (Gibbs Russell 1986; Ellis 1988; Gibbs Russell et al. 1990) differ greatly from the present-day system (Fish et al. 2015). Gibbs Russell et al. (1990) recognised five subfamilies and 21 tribes whereas Fish et al. (2015) recognized eight subfamilies and 24 tribes. Hence, one of the aims of the present study is to provide fresh insights into southern African’s valuable grass genetic resources within a modern classification framework. For the present study, the most recent worldwide phylogenetic classification of Soreng et al. (2017) will be used for taxa found in the study area. The identification manuals of southern African grasses (Gibbs Russell et al. 1990; Fish et al. 2015) listed all indigenous grasses and also described the photosynthetic pathway and growth form, highlighting the diversity of indigenous grasses at tribe and species level. Linking attributes such as photosynthetic pathway, growth form and grazing status with phylogenetic classification could assist in distinguishing taxa with pasture potential. To ensure a similar outcome as for the pasture potential appraisal of legumes (Leguminosae/Fabaceae) indigenous to South Africa, Lesotho and Eswatini (previously Swaziland) (Trytsman et al. 2016, 2019), this paper takes stock of the wealth of indigenous grass genetic resources with added references to grass species currently used in pasture systems. This, in turn, will help with efforts to prioritise conservation of important grass genetic resources at the SA-NFG, in line with the call by Maass and Pengelly (2019) and Pengelly and Maass (2019). Methods The Botanical Database of Southern Africa (BODATSA) maintained by the South African National Biodiversity Institute’s (SANBI) and stored in the BRAHMS platform (Le Roux et al. 2017) was accessed on 2017/03/24 to extract the occurrence records for Poaceae. The taxon and quarter degree grid cell (QDGC) records were extracted and refined, i.e. alien and naturalized species, species with no QDGC reference, species present outside the study area, namely South Africa, Lesotho and Eswatini, henceforth referred to as southern Africa [SA], invalid botanical names, synonyms, as well as incomplete taxa were removed from the dataset. Genus and species names were verified using Fish et al. (2015) to ensure that only species indigenous to SA were included and naturalized or species recorded from countries bordering the study area, were excluded. The database does not reflect all herbarium records from southern Africa, but mainly those housed in the National Herbarium (PRE) in Pretoria and some of its satellite herbaria, notably the KwaZulu-Natal Herbarium (NH) in Durban and the Compton Herbarium (NBG) in Cape Town. Despite its inherent limitations, the meaningful results generated justify the use of this database, the only one of its kind for the study area (Trytsman et al. 2016). In addition, the botanical survey records contained in the PHYTOBAS database were accessed and Poaceae records with GPS locations were extracted. 123 Genet Resour Crop Evol Table 1 Botanical records of the Poaceae indigenous to South Africa, Lesotho and Eswatini contained in the Botanical Database of southern Africa (BODATSA; maintained by SANBI) and PHYTOBAS (South African National Vegetation Data Archive) datasets Database Date #QDGCs/GPS #Taxa #Records BODATSA: Total (QDGCs) 2017/03/24 1803 740 40,865 BODATSA: Species level (QDGCs) 2017/04/05 1803 685 40,139 PHYTOBAS: Species level (GPS) 2018/09/18 3114 373 29,589 BODATSA & PHYTOBAS: Total (QDGCs) 2018/10/26 1811 765 47,652 BODATSA & PHYTOBAS: Species level (QDGCs) 2018/10/26 1803 678 43,889 QDGCs quarter degree grid cells, GPS records with global positioning system localities PHYTOBAS is a National Vegetation Data Archive (2003–2009), designed and administered by the late Dr Bobby Westfall. This database is currently inactive and the proposal by Specht et al. (2018) that ecological data should be curated, proposing online open access of historical data, is supported. Since PHYTOBAS seldom contain infraspecific taxa, only species level records were used for further analyses. Introduced grass species and incomplete taxon records were also removed. A summary of the extent of the two datasets following the editing and merging processes is shown in Table 1. With the exception of references made to grass pasture species conserved in the SA-NFG, no other information was sourced from this genebank database. The merging of the two datasets resulted in many duplicated records within a QDGC. After duplicates were removed, the contributing number of QDGCs for the two datasets were determined and are shown in Table 2. BODATSA contributed to 86.7% of the total dataset and PHYTOBAS 0.7%. Grass species listed by Fish et al. (2015), but not recorded in either BODATSA or PHYTOBAS, are documented and thus the only species not represented in this study. There were eight indigenous grass species not recorded in either BODATSA or PHYTOBAS and these are listed in Table 3. The collection and/or survey intensity (expressed as the number of grass species per QDGC collected as herbarium specimens), henceforth referred to as collection intensity, was calculated and mapped. This is also used as a reference map to ascertain the presence or absence of grass species within QDGCs of the study area when e.g. distribution maps are compared. The taxonomy (subfamily and tribe) and descriptive attributes, specifically endemism, photosynthetic 123 Table 2 The post-editing contribution of BODATSA and PHYTOBAS datasets #QDGC BODATSA PHYTOBAS BODATSA & PHYTOBAS merged Total 1564 % of Total 86.7 13 0.7 226 12.5 1803 100.0 pathway and growth form, were added to each record. For the purpose of this study, growth forms indicated as ‘‘climber’’, ‘‘decumbent’’ and ‘‘scrambler’’ were grouped under the term ‘‘trailing’’. The information for the various attributes was sourced from, amongst others, Gibbs Russell et al. (1990), Germishuizen and Meyer (2003), Van Oudtshoorn (2012), Fish et al. (2015) and SANBI (2017). The phylogenetic classification of Soreng et al. (2017) was followed to compile the evolutionary relationships of subfamilies and tribes. The bioregions vegetation map of Rutherford et al. (2006) was used as a base layer for generating maps with ArcView GIS 3.2, ESRI Inc. 2002. This vegetation map shows 35 bioregions where a bioregion is defined as a composite special terrestrial unit based on similar biotic (vegetation and floristic) and physical features (landscapes and rock types) and processes at the regional scale (Rutherford et al. 2006). The biomes map of southern Africa (Rutherford et al. 2006) and the Köppen-Geiger climate classification map of southern Africa (Beck et al. 2018) are Online Resources 1 and 2 respectively, to be used as reference maps. The ‘Red List of South African Plants’ (SANBI 2017), was consulted for data on the conservational status of indigenous grass species. A list of grasses Genet Resour Crop Evol Table 3 List of Poaceae species indigenous to South Africa, Lesotho and Eswatini not recorded in the Botanical Database of southern Africa (BODATSA; maintained by SANBI) and PHYTOBAS (South African National Vegetation Data Archive) datasets Species Recorded in a Lesotho Catabrosa drakensbergense (Hedberg & I.Hedberg) Soreng & Fish a Ellisochloa papposa (Nees) P.M.Peterson & N.P.Barker Eastern Cape Helictotrichon rogerellisii Mashau, Fish & A.E.van Wyk Western Cape Helictotrichon roggeveldense Mashau, Fish & A.E.van Wyk Northern Cape Melinis scabrida (K.Schum.) Hack. Limpopo Pentameris praecox (H.P.Linder) Galley & H.P.Linder KwaZulu-Natal Poa leptoclada Hochst. ex A.Rich. KwaZulu-Natal, Lesotho Tribolium pleuropogon (Stapf) Verboom & H.P.Linder Southern Cape a Status: Vulnerable with categories ‘Rare’ (not exposed to any direct or potential threat), ‘Near Threatened’ (likely to become at risk of extinction), ‘Vulnerable’ (high risk of extinction), ‘Endangered’ (very high risk of extinction) and ‘Critically Endangered’ (extremely high risk of extinction) will be presented. The conservation status, however, applies only to species within South Africa’s borders and is thus not a global assessment. Results and discussion Fig. 1 Collection intensities of Poaceae indigenous to South Africa, Lesotho and Eswatini as recorded in the Botanical Database of southern Africa (BODATSA; maintained by SANBI) and PHYTOBAS (South African National Vegetation Data Archive) datasets mapped on the bioregions of Rutherford et al. (2006) Collection intensity The collection intensity for indigenous grass species is shown in Fig. 1. Collection intensity records, as presented here, are not linked to species diversity as the latter is determined by vegetation surveys of plant 123 Genet Resour Crop Evol communities. For this study, BODATSA included records documented since 1802 and PHYTOBAS records from 2003 to 2009. It is evident that the very high collection intensities ([ 151 spp./QDGC) were made in the Central Bushveld bioregion (Rutherford et al. (2006) as shown in Fig. 2). Moreover, collection efforts were high in the 2528CA QDGC (red square with [ 201 spp./ QDGC), indicating a possible biased collection approach close to Pretoria, Gauteng. Collection intensities ranging from 51 to 150 spp./QDGC are recorded in the Central Bushveld-, Lowveld-, Indian Ocean Coastal Belt- and the Fynbos bioregions. The lower intensities (\ 100 spp./QDGC) are relatively well distributed over the study area, with no recordings, however, in some of the central parts of the arid region. When the SANBI’s grass herbarium collections are compared to the legume collections (Trytsman et al. 2011), it is evident that the former are larger (± 40 139 records) than the latter (± 27 618 records) despite the larger number of indigenous legume (± 1455 intraspecific taxa of which ± 12% are trees) versus grass species (± 685 species). Phylogenetic classification Worldwide the Poaceae is here treated as containing 12 subfamilies, 52 tribes, 768 genera and 11,506 species (Soreng et al. 2017). The statistics of the grasses as a whole is compared with the indigenous grass diversity of SA in Table 4. Grasses, indigenous to SA, consist of eight subfamilies, 25 tribes, 151 genera and 685 species. The basal lineage subfamilies Anomochlooideae, Pharoideae and Puelioideae are not represented in SA. The subfamilies of the BOP clade (Bambusoideae, Oryzoideae and Pooideae) are all present in SA, but not Micrairoideae of the PACMAD clade. The PACMAD clade consist of subfamilies Panicoideae, Aristidoideae, Chloridoideae, Micrairoideae, Arundinoideae and Danthonioideae. Differences in the formal classification of Poaceae as followed by Gibbs Russell et al. (1990) and Fish et al. (2015), compared to that of Soreng et al. (2017), is outlined in Table 5. Bambusoideae has the poorest representation of species in SA (Table 4). Given that Bambusoideae is the only grass lineage to have diversified in forests (Sungkaew et al. 2009), this is expected with forests being the smallest biome in the study area (Online 123 Resource 1). Pooideae contains the largest number of world species (3968 spp.) but Panicoideae the largest number of species indigenous to SA (256 spp.). Furthermore, only 20% of the world genera (151 genera) and 6% of world species (685 spp.) are found in SA. Even though this is a low number of species, many have proven economic forage value worldwide, e.g. Anthephora pubescens, Chloris gayana, Digitaria eriantha and Eragrostis curvula, highlighting the possible wealth of genetic resources contained in especially the tribe Paniceae. As mentioned previously, the largest number of grass species is contained in Panicoideae resulting mainly from the high number of species in Paniceae (150 spp.). Pentameris P.Beauv. is the largest genus with 75 species followed by Eragrostis Wolf with 65 species (Table 6). Pentameris is found in more temperate regions, i.e. mainly in the Cape Floristic Region (Barker 1993), whereas Eragrostis is found in tropical and subtropical regions (Truter et al. 2015) (also refer to Online Resource 2 for tropical, subtropical and temperate regions of SA). The importance of the Panicoideae is highlighted by the fact that both Zea mays L. and Sorghum Moench are recognized species within the Andropogoneae. Table 6 denotes the genera currently conserved in the SA-NFG. This genebank conserves at present 73 genera and 162 indigenous grass species, i.e. 48% and 24% of the total number of taxa respectively. The low number of grass species presently conserved is a particular concern if the high risk of continued loss of genetic material as described by Pengelly and Maass (2018) is taken into account. However, the renewed interest from South African policy makers in restoring this genebank as a centre of excellence is encouraging and strengthen the aim to improve the collection and conservation efforts for potentially important indigenous grass species. Subfamily distribution The distribution patterns, i.e. the presence or absence of subfamilies in QDGCs of the study area, for the eight subfamilies represented in SA are shown in Fig. 3. Aristidoideae, Panicoideae and Chloridoideae covers most of the bioregions with the latter having more coverage in especially the central regions of South Africa. The presence of Panicoideae is infrequent in Namaqualand, Bushmanland and Karoo Genet Resour Crop Evol 123 Genet Resour Crop Evol b Fig. 2 Bioregions of South Africa, Lesotho and Swaziland Growth form (Rutherford et al. 2006) to be used as reference map regions (refer to Fig. 2) and Aristidoideae has a low presence in the central parts of the arid regions. According to Fish et al. (2015), Panicoideae occurs in more mesic habitats whereas Chloridoideae occurs in drier saline habitats. Figure 3 further shows that Oryzoideae and Pooideae have similar distribution patterns, covering the Cape Floristic Region, Succulent Karoo and a large part of the Grassland biome (Online Resource 1). Bambusoideae occurs only in the Drakensberg Grassland bioregion whereas Arundinoideae occurs mostly in Central Bushveld, Lowveld and Indian Ocean Coastal Belt bioregions (refer to Fig. 2). The high presence of Arundinoideae in the winter rainfall zone described by Gibbs Russell (1986) has been invalidated since the addition of the Oryzoideae and Danthonioideae resulted in the removal of Ehrharta Thunb. and Pentameris from Arundinoideae. Lastly, Danthonioideae is found in the Cape Floristic Region, Succulent Karroo biome (Online Resource 1) and the Drakensberg Grassland bioregion. The majority of grass species in SA are described as tufted (401 spp.) followed by rhizomatous/tufted (105 spp.) and rhizomatous (59 spp.) (Fig. 4). Linder et al. (2018) states that the tufted growth form, stemming from tillering, probably evolved within Poaceae. This growth form was found, however, to be less tolerant to heavy grazing compared to rhizomatous or stoloniferous grasses by Reynolds (1995). The tufted growth form is nonetheless an attribute of the majority of grass pasture species, particularly those mentioned earlier as key pasture species. Rhizomes (underground stems) are also a valuable survival trait, in not only protecting the plant from trampling (Rechenthin 1956) or being correlated with drought resistance (Zhou et al. 2014) but also having a regenerative ability (Zwerts et al. 2015). Da Silva et al. (2015) argues that complementary growth habits of C4 tropical grasses, i.e. rhizomatous grasses that occupy a lower horizontal stratum and tufted grasses a higher vertical stratum, result in coexistence and allows for greater grass diversity. Little information on the hemicryptophyte growth form (buds present at or near the soil surface) exists for these indigenous species, although it is accepted that it Table 4 Statistics of tribes, genera and species of Poaceae indigenous to South Africa, Lesotho and Eswatini (SA) compared to the world Sourced from Fish et al. (2015) and Soreng et al. (2017) Subfamily Tribe World Tribe SA Genera World Genera SA Species World Species SA Anomochlooideae 2 – 2 – 4 – Pharoideae 1 – 3 – 12 – Puelioideae 2 – 2 – 11 – BOP clade Oryzoideae 4 2 19 4 115 Bambusoideae 3 2 125 2 1670 2 15 6 202 18 3968 58 Pooideae 28 PACMAD clade Aristidoideae 1 1 3 3 367 42 Panicoideae 13 6 247 65 3241 256 Arundinoideae Micrairoideae 2 3 2 14 8 4 40 184 6 – Danthonioideae 1 1 19 10 292 114 Chloridoideae 5 5 124 45 1602 179 52 25 768 151 11,506 685 Total 123 – – Genet Resour Crop Evol Table 5 Comparison of the classification of Poaceae indigenous to South Africa, Lesotho and Eswatini followed by Gibbs Russell et al. (1990) and Fish et al. (2015) compared to that of Soreng et al. (2017) a Gibbs-Russell et al. (1990) b a Fish et al. (2015) Soreng et al. (2017) Subfamily Tribe Subfamily Tribe Subfamily Pooideae Triticeae Bambusoideae Bambuseae Oryzoideae Brachypodieae Bromeae Olyreae Ehrhartoideae Aveneae Meliceae Bambusoideae Ehrharteae Brachypodieae Ehrharteae Oryzeae Bambusoideae Oryzeae Pooideae Tribe Arundinarieae Bambuseae Pooideae Meliceae Poeae Bromeae Oryzeae Meliceae Stipeae Brachypodieae Olyreae Poeae Bromeae Centotheceae Stipeae Triticeae Ehrharteae Triticeae Poeae Bambuseae Stipeae Aristidoideae Arundinoideae Aristideae Arundineae Arundineae Danthonioideae Danthonieae Danthonieae Panicoideae Andropogoneae Paniceae Aristideae Arundinelleae Paspaleae Chloridoideae Pappophoreae Paniceae Arundinelleae Chlorideae Paspaleae Panicoideae Paniceae Tristachyideae Arundinoideae Arundinelleae Andropogoneae Aristidoideae Panicoideae Centotheceae Andropogoneae Arundinoideae Centotheceae Chloridoideae Maydeae Aristideae Tristachyideae Molinieae Arundineae Centropodieae Danthonioideae Danthonieae Cynodonteae Chloridoideae Centropodieae Eragrostideae Triraphideae Triraphideae Eragrostideae Zoysieae Zoysieae Cynodonteae a Tribes in phylogenetical order b Tribes in alphabetical order is common in grasses (Linder et al. 2018). This trait can contribute to survival during unfavourable conditions such as seasonal drought, cold, fire or heavy grazing when plants can regrow from the base after conditions improves (Linder et al. 2018). In general, most grass growth forms have a relatively wide distribution throughout the study area. The exceptions are the cushion growth form (low growing, mat forming at high altitudes) of the genus Pentameris (eight of the 75 spp.), the trailing form of especially eight Panicum L. spp. and the three trailing Oplismenus P.Beauv. spp. Pentameris is found mainly in the Fynbos biome, Panicum spp. in the Nama- Karoo, Grassland, Savanna and Indian Ocean Coastal Belt biomes and Oplismenus spp. in the Lowveld and Indian Ocean Coastal Belt bioregions (refer to Fig. 2 and Online Resource 1). A review on the concept of grazing lawns, i.e. a short-stature grassland community type, persisting and spreading under heavy grazing (Hempson et al. 2014) affirms that certain ecotypes of Themeda triandra Forssk. and Digitaria eriantha forms small cushion-like plants and dense stoloniferous or rhizomatous clonal mats under frequent grazing. This finding strengthens the call of Van Wyk (1995) to formally label infraspecific genetic 123 Genet Resour Crop Evol Table 6 Subfamilies, tribes and genera (number of species in brackets sourced from Fish et al. (2015)) of Poaceae indigenous to South Africa, Lesotho and Eswatini following the Subfamily phylogenetic classification of Soreng et al. (2017) for subfamilies and tribes. Taxa in bold indicate the largest (most speciose) group/s within the classification Tribe Genera Ehrharteae Ehrhartaab (23) Oryzeae Leersiab (2) Oryzab (2) Prosphytochloaa (1) Arundinarieae Thamnocalamusa (1) Bambuseae Oxytenanthera (1) Meliceae Melicaa (2) Streblochaete (1) Stipeae Stipaa (2) Brachypodieae Brachypodiuma (2) Bromeae Bromusa (4) Triticeae Hordeuma (1) Secalea (1) Poeae Agrostisa (9) Anthoxanthuma (4) Calamagrostis (1) Catabrosa (1) Festucaa (8) Holcusa (1) Helictotrichona (14) Koeleria (1) Poab (3) Polypogona (1) Puccinelliaa (2) Aristideae Aristidaab (23) Sartidia (2) Stipagrostisab (17) Tristachyideae Danthoniopsisb (5) Loudetiab (5) Trichopteryx (1) Tristachyab (3) BOP clade Oryzoideae Bambusoideae Pooideae i PACMAD clade Aristidoideae Panicoideae Arundinoideae Centotheceae Megastachya (1) Paniceaecef Acrocerasb (1) Alloteropsisb (2) Anthephoraab (3) Brachiariab (16) Cenchrusb (1) Digitariaab (25) Echinochloab (9) Entolasia (1) Eriochloab (4) Leucophrys (1) Megaloprotachneb (1) Melinisab (8) Odontelytrum (1) Oplismenus (3) Panicumab(34) Paspalidiumb (2) Pennisetum (7) Pseudechinolaena (1) Sacciolepis (6) Setariaab (12) Stenotaphrumb (2) Stereochlaena (1) Tarigidiaa (1) Tricholaenaab (2) Urochloab (6) Paspaleae Paspalumb (3) Arundinelleae Arundinella (1) Andropogoneaeeg Andropogonb (14) Arthraxon (1) Bothriochloab (3) Chrysopogonb (1) Cleistachneb (1) Coelorachis (1) Cymbopogonab (6) Dichanthiumb (1) Diheteropogonb (2) Elionurusb (1) Elymandra (1) Eriochrysis (2) Eulalia (2) Hackelochloa (1) Hemarthriab (1) Heteropogonb (2) Hyparrheniab (20) Hypertheliab (1) Imperatab (1) Ischaemumb (2) Miscanthusb (2) Monocymbiumb (1) Oxyrhachis (1) Phacelurus (1) Rhytachne (2) Rottboellia (1) Schizachyriumb (6) Sehima (2) Sorghastrumab (2) Sorghumb (2) Themedab (1) Trachypogonb (1) Urelytrum (1) Molinieae Elytrophorus (1) Phragmites (2) Styppeiochloa (1) Arundineae Dregeochloaa (2) Danthonioideae Danthonieae Capeochloaa (3) Chaetobromusab (1) Geochloaa (3) Merxmuelleraa (4) Pentamerisa (75) Pentaschistisa (3) Pseudopentamerisa (3) Schismusab (4) Tenaxiaa (5) Triboliuma (13) Chloridoideae Centropodieae Triraphideae Centropodiaab (1) Ellisochloa (1) Triraphisa (4) Eragrostideaedh Catalepis (1) Cladoraphisa (2) Diandrochloa (2) Enneapogonab (6) Eragrostisab (65) Fingerhuthiaab (2) Pogonarthriab (1) Schmidtiab (2) Stiburus (2) Tetrachneb (1) Zoysieae Spartina (1) Sporobolusab (34) 123 Genet Resour Crop Evol Table 6 continued Subfamily Tribe Genera Cynodonteaec Acrachneb (1) Bewsiab (1) Brachychloa (2) Chlorisb (6) Coelachyrum (1) Ctenium (1) Cynodonab (6) Dactylocteniumb (4) Dinebra (1) Eleusineb (2) Enteropogonb (2) Eustachysb (1) Harpochloa (1) Leptocarydion (1) Leptochloab (5) Lepturusb (1) Lintoniab (1) Lophacmea (1) Microchloab (2) Mosdeniaa (1) Odyssea (1) Oropetiumb (1) Perotis (1) Polevansia (1) Rendlia (1) Schoenefeldiab (1) Tetrapogonb (1) Tragusab (3) Trichoneurab (2) Tripogon (1) a Include endemic species b Genera conserved in the South African National Forage Genebank Tribes with majority of species with cPioneer, dSubclimax, eClimax, fDecreaser, gIncreaser I, hIncreaser II i Temperate grass lineage variants within especially species of potential economic significance. Photosynthetic pathway and grazing status The photosynthetic pathway distinguished in the subfamilies and tribes of Poaceae is shown in Table 7. The early-branching subfamilies (Oryzoideae, Bambusoideae and Pooideae) are shown to have a C3 photosynthetic pathway, whereas the later-branched Chloridoideae has a C4 photosynthetic pathway. Interestingly, Bouchenak-Khelladi et al. (2010) found molecular evidence that C4 photosynthesis (of at least the subfamily Chloridoideae) may well have originated in Africa. The majority of species of Aristidoideae and Panicoideae are C4 whereas all species of Arundinoideae and Danthonioideae are C3. Tribes with well-known pasture species i.e. Andropogoneae, Eragrostideae and Cynodonteae contains only C4 species whereas Paniceae have a mixture of C3 (i.e. genus Panicum) and C4 species. The importance of C4 grasses in pasture production could lie in their ability to transfer larger proportions of plant nitrogen to roots in infertile environments (Long 1999), as well as larger leaf area production in fertile and disturbed environments. This supports Linder et al. (2018) finding that grasses has the ability to colonize, persist and transform environments, properties that are the key to success. Furthermore, evidence presented by Linder et al. (2018) indicated that C4 plants have a higher carbon fixing efficiency over a range of habitats when soil resources are limited compared to C3 types. However, the higher forage value in terms of crude protein content and digestibility in C3 compared to C4 grasses (Gibson 2009) could in effect be an important grass collection objective, where the focus is on pasture development for temperate regions, especially for dairy farming. The distribution of C3 and C4 grass species is shown in Fig. 5. Species with the C3 pathway are widespread in the Fynbos, Succulent Karoo biomes and the Drakensberg Grassland, Sub-Escarpment Grassland bioregions (refer to Fig. 2 and Online Resource 1). The central arid region has low occurrences of C3 grass species whereas the Kalahari Duneveld bioregion has none. Species with a C4 photosynthetic pathway is found in all the bioregions of SA. Vogel et al. (1978) investigated the geographical distribution of C3 and C4 grasses in southern Africa and concluded that low temperatures during seasonal growth favour the C3 grasses in the regions mentioned above. The exclusive presence of C4 in the tropical region (Online Resource 2) as described by Vogel et al. (1978) is, however, not evident in Fig. 5. C3 species such as Agrostis L., Festuca L., Helictotrichon Besser (Poeae) and Pentameris (Danthonieae) were recorded in the tropical regions of the study area (Online Resource 2). The only known grass species in SA with both a C3 and C4 subspecies, Alloteropsis semialata (R.Br.) Hitchc. need special mentioning (Ellis 1974; Gibbs Russell 1983). Alloteropsis semialata (R.Br.) Hitchc. subsp. eckloniana (Nees) Gibbs Russ. has a C3 photosynthetic pathway and A. semialata (R.Br.) Hitchc. subsp. semialata (R.Br.) Hitchc. a C4, suggesting an evolutionary reversion from C4 to C3 (Ibrahim et al. 2009). Within Panicum, a genus containing important forage crops, Ellis (1988) distinguished 11 Panicum spp. having a C3 and 23 with C4 photosynthetic pathway. 123 Genet Resour Crop Evol Fig. 3 The distribution patterns for Poaceae subfamilies indigenous to South Africa, Lesotho and Eswatini in phylogenetic order according to Soreng et al. (2017), mapped on the bioregions of Rutherford et al. (2006) 123 Genet Resour Crop Evol 450 400 401 Number of grass species 350 300 250 200 150 105 100 59 6 6 6 4 3 3 3 3 3 2 Tufted/trailing Hydrophyte Hydrophyte/rhizomatous Hydrophyte/rhizomatous/stoloniferous/tufted Hydrophyte/rhizomatous/tufted Hydrophyte/trailing Cushion/tufted Stoloniferous/tufted Stoloniferous Rhizomatous/stoloniferous Trailing Rhizomatous Rhizomatous/tufted Tufted 1 1 Rhizomatous/woody 8 0 Geophytic 8 Rhizomatous/stoloniferous/tufted 12 Hydrophyte/tufted 13 Hydrophyte/stoloniferous/tufted 1… Hydrophyte/rhizomatous/stoloniferous 18 Cushion 50 Fig. 4 Growth forms of Poaceae indigenous to South Africa, Lesotho and Eswatini. Sourced mainly from Gibbs Russell et al. (1990) Pau et al. (2012) pointed out that the evolutionary history of Poaceae is important for understanding the C3 and C4 functional diversity of grasses, as this will affect their responses to global change. A comparison of the recorded successional status of tribes (Van Oudtshoorn 2012) shows that Paniceae and Cynodonteae contain the largest number of species with pioneer status, whereas Paniceae and Andropogoneae contain the largest number with climax status (Table 7). Pioneer species are usually annuals, growing in disturbed habitats or unfavourable conditions whereas climax species are perennials, growing only when normal, optimal growth conditions prevail (Van Oudtshoorn 2012). Panicoideae distinctly contains most climax species compared to other subfamilies and thus should be the focus for further assessing grass genetic resources with pasture potential, especially within the tribes Paniceae and Andropogoneae. Grass species belonging to Paniceae and Andropogoneae are well presented in all biomes but narrowly presented in the drier areas namely the Nama-Karoo and Succulent Karoo biomes (Online Resource 1). The ecological status as defined by Forani et al. (1978) and Van Oudtshoorn (2012) indicate that the high number of Decreaser species present in Paniceae further underlines this tribe’s forage value, i.e. 47% of indigenous species with known preferential grazing status are grouped here (Table 7). Decreasers are defined as grasses abundant in good veld and will decrease when over- or undergrazed, whereas Increasers will increase under any type of mismanagement. Andropogoneae and Eragrostideae contain respectively the highest number of Increaser I and Increaser II species. 123 Genet Resour Crop Evol Table 7 The number of species within Poaceae subfamilies and tribes indigenous to South Africa, Lesotho and Eswatini using a C3 and/or C4 photosynthetic pathway sourced mainly Subfamily Tribe C3 C4 Pioneer from Gibbs Russell et al. (1990) and successional and ecological grazing status (Van Oudtshoorn 2012) Subclimax Climax Decreaser Increaser I Increaser II Increaser III BOP clade Oryzoideae Ehrharteae Oryzeae Bambusoideae Pooideae 23 1 2 5 Arundinarieae 1 Bambuseae 1 Meliceae 3 Stipeae Brachypodieae 2 2 Bromeae 4 Triticeae 2 Poeae 2 1 1 1 45 1 1 4 3 2 2 5 2 2 PACMAD clade Aristidoideae Aristideae 2 Panicoideae Tristachyideae Centotheceae Paniceae 2 127 Paspaleae 3 Arundinelleae 1 13 3 Molinieae 4 Arundineae 2 3 25 21 1 1 27 7 7 16 1 87 5 20 5 2 2 2 1 Danthonioideae Danthonieae Chloridoideae Centropodieae 2 Triraphideae 4 114 1 3 3 1 1 1 1 Eragrostideae 84 7 11 5 2 1 25 2 Zoysieae 35 2 3 2 1 1 6 1 Cynodonteae 54 10 3 6 5 2 14 1 Endemism and conservation concern Figure 6 shows the collection intensity for the 257 grass species endemic to SA. The highest number of endemic species per QDGC is recorded in the Cape Floristic Region and in the Drakensberg Alpine Centre (Van Wyk and Smith 2001; Rutherford et al. 2006). In terms of the presence of endemic species in the study area, the Upper and Lower Karoo, Bushmanland and Lowveld are the main bioregions (refer to Fig. 2) containing the smallest number of endemic species. 123 9 1 1 23 Andropogoneae Arundinoideae 40 14 Danthonioideae (found mainly in the Cape Floristic Region and Succulent Karoo biome (see Online Resource 1) contributes to nearly half of the total number of endemic grass species in SA. Pentameris accounts for 28% of endemic species followed by Ehrharta (9%), Eragrostis (7%) and Tribolium (5%). Pentameris is mainly found in the Fynbos biome and Drakensberg Grassland bioregion, Ehrharta in the Fynbos and Succulent Karoo biome, Eragrostis in most parts of the study area and Tribolium in the Fynbos and Succulent Karoo biome (Fish et al. 2015). Genet Resour Crop Evol Fig. 5 The distribution of C3 and C4 photosynthetic pathways in Poaceae indigenous to South Africa, Lesotho and Eswatini. Sourced mainly from Gibbs Russell et al. (1990) and mapped on the bioregions of Rutherford et al. (2006) Pentameris and Tribolium are considered having a lower grazing value than Ehrharta (Van Oudtshoorn 2012). The conservational assessment for grass species published in the Red List of South African Plants (SANBI 2017) is listed in Table 8 together with regions or centres of endemism (sensu Van Wyk and Smith 2001). The genus Pentameris holds a true conservational concern with 24 species listed, present in all categories. There are two Panicum spp. and one Secale L. sp., both important genera in pasture production also listed, respectively as rare and critically endangered. Panicum sanctaluciense and Panicum dewinteri, both perennials, are found respectively in the tropical region (Online Resource 2) and the Mopane bioregion of the study area. Secale strictum (J.Presl) J.Presl subsp. africanum (Stapf) K.Hammer, is a perennial wild rye found in Renosterveld of the western mountain Karoo (Fish et al. 2015) and within the Hantam-Roggeveld Centre (Van Wyk and Smith 2001). Helictotrichon quinquesetum (Steud.) Schweick., only known from the slopes of Table Mountain, Cape Town (Fish et al. 2015), is the only grass species listed as possibly extinct. It is evident from Table 8 that the majority of grass species with a conservational concern are found in the Cape Floristic Region, followed by the Hatam–Roggeveld Centre. Habitat loss is identified as the major threat to South African plants, i.e. infrastructure development, urban expansion, cultivation of crops, commercial afforestation and mining (SANBI 2017). In terms of the conservation of genetic resources for possible future use, it is suggested that representative seed samples are collected from these species and stored in the SANFG. The screening and characterisation of these genetic resources will allow for re-introduction when certain populations go locally extinct. In addition, these efforts will allow these genetic resources to be included into breeding programs that can produce more adapted species or for rehabilitation of degraded rangelands. Conclusion The important role that members of the Poaceae plays in sustainable pasture production systems compels the SA-NFG to conserve SA’s indigenous grass genetic resources, not only those used in current pasture systems, but also genetic material that could be beneficial in future breeding programs and/or adapted to specific agro-ecological conditions. A modern classification framework was therefore used to document indigenous grass species, recorded in various southern African taxonomic reviews and botanical databases. Linking various attributes with phylogenetic classification and vegetation types assisted in distinguishing taxa with pasture potential and to prioritise future collection and conservation efforts. 123 Genet Resour Crop Evol Fig. 6 The collection intensity for the endemic grass species present in South Africa, Lesotho and Eswatini. Sourced from Fish et al. (2015) and mapped on the bioregions of Rutherford et al. (2006) Results showed that the inclusion of PHYTOBAS (South African National Vegetation Data Archive) added valuable data since 13 additional QDGCs were added to the dataset. Attention is drawn to the value of historical vegetation data and the proposed call for online open access is supported. The collection intensity map shows that the central Bushveld has high collections, but that in large parts of the central arid region no collections were done. There are eight subfamilies, 25 tribes, 151 genera and 685 species present in SA. Subfamilies Anomochlooideae, Pharoideae, Puelioideae (basal lineage) and Micrairoideae (PACMAD clade) are not represented in SA. Only 6% of world grass species (i.e. 685 spp.) are found in SA, with Panicoideae having the largest number of species (256 spp.). Aristidoideae, Panicoideae and Chloridoideae are represented in most of the bioregions whereas Bambusoideae is found only in one bioregion, namely the Drakensberg Grassland. The infrequent presence of Panicoideae in the western and central arid regions should be taken into account when focusing on the collection of potentially drought tolerant grass species. Instead, the presence of Chloridoideae in 123 these arid regions, containing important pasture species such as Chloris, Cynodon and Eragrostis, should be considered. The majority of grass species in SA are tufted, an attribute of the majority of key grass pasture species. Tribes with well-known pasture species contains only C4 species (Andropogoneae, Eragrostideae and Cynodonteae) whereas Paniceae have a mixture of C3 and C4 species (i.e. genus Panicum). Species with a C4 photosynthetic pathway is found in all the bioregions of SA whereas C3 species have low occurrences in the central arid region. Panicoideae contains more climax species compared to other subfamilies, especially within the tribes Paniceae and Andropogoneae. The 257 endemic grass species found largely in the Cape Floristic Region and in the Drakensberg Alpine Centre need an in depth assessment to determine the role the SA-NFG plays in their conservation as possible future pasture genetic resources. This is an urgent outcome for the 24 species of Pentameris that holds a true conservational concern. The collection of viable seed of two Panicum spp. and one Secale sp., listed respectively as rare and critically endangered and the possible extinct Secale Genet Resour Crop Evol Table 8 Grass species and infraspecific taxa on the ‘Red List of South African Plants’ in order of the least to highest risk of extinction (SANBI 2017). The superscript following some names indicates to which specific local region or centre of endemism (sensu Van Wyk and Smith 2001) the particular taxon is endemic, or if more widespread, confined to in the study area Rare Vulnerable (continued) Capeochloa setaceaCFR Elytrophorus globularis Dregeochloa calviniensis Helictotrichon barbatumSKR Panicum sancta-lucienseMC Helictotrichon namaquenseKBC Pentameris caulescens Pentameris clavata CFR Helictotrichon rogerellisiiCFR CFR Oryza longistaminata Pentameris glacialisCFR Pentameris hirtiglumis Pentameris calcicola var. hirsutaCFR CFR Pentameris longiglumis subsp. longiglumisCFR CFR Pentameris longipesAC Pentameris holciformis Pentameris horridaCFR Pentameris trifidaCFR Pentameris longiglumis subsp. gymnocolea Pentameris swartbergensis Pentameris unifloraCFR CFR CFR Near threatened Tribolium ciliareCFR Endangered Helictotrichon roggeveldenseHRC Eulalia aurea PC Pentameris bachmanniiCFR Oxyrhachis gracillima * SBC Pentameris dentataHRC CFR Pentameris eckloniiCFR Panicum dewinteri Pentameris aspera Sartidia jucundaSBC Pentameris calcicola var. calcicolaCFR Pentameris limaKBC Pentameris pholiuroidesCFR Pentameris scandensCFR Stipagrostis geminifoliaGC Vulnerable Critically endangered Capeochloa cincta subsp. sericeaAC Catabrosa drakensbergense DAC Ehrharta setacea subsp. unifloraCFR CFR Ellisochloa papposa Pentameris barbata subsp. orientalisCFR Pentameris elegansCFR Pentameris ellisiiCFR Secale strictum subsp. africanumHRC Tribolium pleuropogonCFR Helictotrichon quinquesetum (possibly extinct)CFR Rare, not exposed to any direct or potential threat; Near Threatened, likely to become at risk of extinction; Vulnerable, high risk of extinction; Endangered, very high risk of extinction; Critically Endangered, extremely high risk of extinction; AC, Albany Centre; CFR, Cape Floristic Region; DAC, Drakensberg Alpine Centre; GC, Gariep Centre; HRC, Hantam–Roggeveld Centre; KBC, Kamiesberg Centre; MC, Maputaland Centre; PC, Pondoland Centre; SBC, Soutpansberg Centre; SKR, Succulent Karoo Region *In the study area confined to the centre, but not endemic as it also occurs elsewhere in Africa/Madagascar strictum subsp. africanum is a critical undertaking for the SA-NFG. The SA-NFG conserve at present only a quarter of indigenous grass species with the current collection already facing serious risks. This will prompt decision makers to reinvest in the SA-NFG, focusing on conserving genetic resources for improved animal production, mitigate the possible effect of climate change on the valuable grass genetic resources, and consequently pasture production. It is proposed that a strategy be developed for the SA-NFG to collect and conserve seed of Paniceae since this tribe contains valuable pasture grass species, thus focusing on speciose subtropical genera such as Anthephora, Brachiaria, Digitaria, Panicum and Setaria. 123 Genet Resour Crop Evol Furthermore, temperate genera containing important pasture species i.e. Bromus, Hordeum and Festuca as well as endemic species with grazing value, such as Ehrharta, should also be the focus of seed collection efforts by the SA-NFG. The need for more emphasis on the description and formal recognition of infraspecific taxa, including ecotypes, are also highlighted. It is further proposed that stakeholders in biodiversity conservation strategize plant collections excursions to those areas, previously not sampled. The current systems of germplasm conservation at the SA-NFG is under revision. A long term storage facility is being prepared for purely conservation purposes, while active collections of important pasture species as well as indigenous species with agronomic potential will be made available for research purposes. The reason for the long term storage of all grass genetic resources was brought about by the changes in bioclimatic conditions predicted for southern Africa. The conservation of these resources at the SA-NFG could, under future bioclimatic conditions, result in potential rehabilitation of degraded areas, or reintroductions with better adapted ecotypes of the same species after local extinction of naturally occurring populations. The indigenous grass database developed for this study will be used to establish biogeographical patterns of the grass flora as well as for assessing their pasture potential. These results will be combined with the published results for indigenous legumes and used to develop a collective strategy in terms of prioritizing species, identifying key regions and planning characterization and evaluation studies. Acknowledgements We thank the South African National Biodiversity Institute (SANBI) for making available the distribution and descriptive data contained in the BODATSA database, the late Dr Bobby Westfall for administrating the data contained in the PHYTOBAS National Vegetation Data Archive and Elsa van Niekerk (ARC-PPR) for the graphics. Funding The Agricultural Research Council of South Africa funded this study. Data availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. 123 Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. References Barker NP (1993) A biosystematic study of Pentameris (Arundineae, Poaceae). Bothalia 23:25–47 Beck HE, Zimmermann NE, McVicar TR, Vergopolan N, Berg A, Wood EF (2018) Present and future Köppen–Geiger climate classification maps at 1-km resolution. Nature 5:180214 Bouchenak-Khelladi Y, Verboom GA, Savolainen V, Hodkinson TR (2010) Biogeography of the grasses (Poaceae): a phylogenetic approach to reveal evolutionary history in geographical space and geological time. Bot J 162:543–557 Buckler ES, Thornsberry JM, Kresovich S (2001) Molecular diversity, structure and domestication of grasses. Genet Res 77:213–218 Capstaff NM, Miller AJ (2018) Improving the yield and nutritional quality of forage crops. Front Plant Sci 9:535 Da Silva SC, Sbrissia AF, Pereira LET (2015) Ecophysiology of C4 forage grasses—understanding plant growth for optimising their use and management. Agriculture 5:598–625 Ellis RP (1974) The significance of the occurrence of both Kranz and non-Kranz leaf anatomy in the grass species Alloteropsis semialata. S Afr J Sci 70:169–173 Ellis RP (1988) Leaf anatomy and systematics of Panicum (Poaceae: Panicoideae) in southern Africa. Monogr Syst Bot Mo Bot Gard 25:129–156 Fish L, Mashau AC, Moeaha MJ, Nembudani MT (2015) Identification guide to the southern African grasses. An identification manual with keys, descriptions and distributions. Strelizia 36. South African National Biodiversity Institute, Pretoria Forani BD, Tainton NM, Booysen PV (1978) The development of a method for assessing veld condition in three grassveld types in Natal. Proc Grassld Soc S Afr 13:27–33 Germishuizen G, Meyer NL (2003) Plants of southern Africa: an annotated checklist. Strelitzia 14. South African National Biodiversity Institute, Pretoria Gibbs Russell GE (1983) The taxonomic position of C3 and C4 Alloteropsis semialata (Poaceae) in southern Africa. Bothalia 14:205–213 Gibbs Russell GE (1986) Significance of different centres of diversity in subfamilies of Poaceae in southern Africa. Palaeoecol of Africa 17:183–192 Gibbs Russell GE, Watson L, Koekemoer M, Smook L, Barker NP, Anderson HM, Dallwitz MJ (1990) Grasses of southern Africa. In: Leistner OA (ed) Memoirs of the botanical survey of South Africa No. 58. Botanical Research Institute, Pretoria Gibson DJ (2009) Grasses and grassland ecology. Oxford University Press, Oxford Genet Resour Crop Evol Hempson GP, Archibald S, Bond WJ, Ellis RP, Grant CC, Kruger FJ, Laurence M, Kruger LM, Moxley C, OwenSmith N, Peel MJS, Smit IPJ, Vickers KJ (2014) Ecology of grazing lawns in Africa. Biol Rev Cambridge Philos Soc 90:979–994 Hodkinson TR (2018) Evolution and taxonomy of the grasses (Poaceae): a model family for the study of species-rich groups. Annu Plant Rev 1:1–39 Ibrahim DG, Burke T, Ripley BS, Osborne CP (2009) A molecular phylogeny of the genus Alloteropsis (Panicoideae, Poaceae) suggests an evolutionary reversion from C4 to C3 photosynthesis. Ann Bot 103:127–136 Kruger AJ, Jooste M, Ellis RP, Erasmus M, Van Oudtshoorn F (1993) Collection and ex situ conservation of African grasses. In: Conference proceedings: the conservation and utilization of southern African botanical diversity, Cape Town. Unpublished. Poster presentation, Cape Town Le Roux MM, Wilkin P, Balkwill K, Boatwright JS, Bytebier B, Filer D, Klak C, Klopper RR, Koekemoer M, Livermore L, Lubke R, Magee AR, Manning JC, Paton A, Pearce T, Slingsby J, Van Wyk B-E, Victor JE, Von Staden L (2017) Producing a plant diversity portal for South Africa. Taxon 66:421–431 Linder HP, Lehmann CER, Archibald S, Osborne CP, Richardson DM (2018) Global grass (Poaceae) success underpinned by traits facilitating colonization, persistence and habitat transformation. Biol Rev Cambridge Philos Soc 93:1125–1144 Long SP (1999) Environmental responses. In: Sage RF, Monson RK (eds) C4 Plant biology. Academic Press, San Diego, pp 215–249 Maass BL, Pengelly BC (2019) Tropical and subtropical forage germplasm conservation and science on their deathbed! 1. A journey to crisis. Outlook Agr 48:198–209 Mashau AC, Coetzee AL (2019) Enneapogon limpopoensis (Poaceae: Chloridoideae: Eragrostideae): a new species from southern Africa. Kew Bull 74:4 Mashau AC, Fish L, Van Wyk AE (2010) Poaceae: two new species of Helictotrichon (Pooideae: Aveneae) from South Africa. Bothalia 40:179–183 Pau S, Edwards EJ, Still CJ (2012) Improving our understanding of environmental controls on the distribution of C3 and C4 grasses. Global Change Biol 19:184–196 Pengelly BC, Maass BL (2018) Implementation of the tropical and subtropical forages strategy 2017. Final Report to Global Crop Diversity Trust, Bonn, Germany Pengelly BC, Maass BL (2019) Tropical and subtropical forage germplasm conservation and science on their deathbed! 2. Genebanks, FAO and donors must take urgent steps to overcome the crisis. Outlook Agr 48:210–219 Rechenthin CA (1956) Elementary morphology of grass growth and how it affects utilization. J Range Manage 9:167–170 Reynolds SG (1995) Pasture-cattle-coconut systems. Food and Agriculture Organization of the United Nations, Rome http://www.fao.org/docrep/005/af298e/af298E00. htm#TOC Rutherford MC, Mucina L, Powrie LW (2006) Biomes and bioregions of southern Africa. In: Mucina L, Rutherford MC (eds) The vegetation of South Africa, Lesotho and Swaziland, Strelitzia 19. South African National Biodiversity Institute, Pretoria, pp 30–51 SANBI (2017) Red list of South African Plants. http://redlist. sanbi.org/ Shackleton SE (1991) Autecological studies on grass species in southern Africa—a literature survey. J Grassl Soc South Afr 8:114–120 Soreng RJ, Peterson PM, Romaschenko K, Davidse G, Teisher JK, Clark LG, Barberá P, Gillespie LJ, Zuloaga FO (2017) A worldwide phylogenetic classification of the Poaceae (Gramineae) II: an update and a comparison of two 2015 classifications. J Syst Evol 55:259–290 Specht A, Bolton MP, Kingsford B, Specht RL, Belbin L (2018) A story of data won, data lost and data re-found: the realities of ecological data preservation. Biodivers Data J 6:e28073 Sungkaew S, Stapleton CMA, Salamin N, Hodkinson TR (2009) Non-monophyly of the woody bamboos (Bambuseae; Poaceae): a multi-gene region phylogenetic analysis of Bambusoideae s.s. J Plant Res 122:95–108 Truter WF, Botha PR, Dannhauser CS, Maasdorp BV, Miles N, Smith A, Snyman HA, Tainton NM (2015) Southern African pasture and forage science entering the 21st century: past to present. Afr J Range For Sci 32:73–89 Trytsman M, Van Wyk AE, Masemola EL (2011) Systematics, diversity and forage value of indigenous legumes of southern Africa. Afr J Biotechnol 10:13773–13779 Trytsman M, Westfall RH, Breytenbach PJJ, Calitz FJ, Van Wyk AE (2016) Diversity and biogeographical patterns of legumes (Leguminosae) indigenous to southern Africa. PhytoKeys 70:53–96 Trytsman M, Masemola EL, Müller FL, Calitz FJ, Van Wyk AE (2019) Assessing legumes indigenous to South Africa, Lesotho and Swaziland for their pasture potential. Afr J Range For Sci 36:27–40 Tzvelev NN (1989) The system of grasses (Poaceae) and their evolution. Bot Rev 55:142–168 Van Oudtshoorn F (2012) Guide to grasses of Southern Africa. Briza publications, Pretoria Van Wyk AE (1995) The role of the plant taxonomist in the conservation of biodiversity in Africa: a practical approach. In: Bennun LA, Aman RA, Crafter SA (eds) Conservation of biodiversity in Africa: local initiatives and institutional roles. Proceedings of an international meeting, National Museums of Kenya, 30 August–3 September 1992, Nairobi, Kenya. Centre for Biodiversity, Kenya, pp. 329–334 Van Wyk AE, Smith G (2001) Regions of floristic endemism in southern Africa. A review with emphasis on succulents. Umdaus Press, Hatfield Victor JE, Mashau AC, Ngobeni VJ (2012) The taxonomic and conservation status of Agrostis eriantha var. planifolia. Bothalia 42:202–204 Victor JE, Smith GF, Van Wyk AE (2015) A method for establishing taxonomic research priorities in a megadiverse country. Phytotaxa 203:55–62 Vogel JC, Fuls A, Ellis RP (1978) The geographical distribution of Kranz grasses in South Africa. S Afr J Sci 74:209–215 Zhou Y, Lambrides CJ, Fukai S (2014) Drought resistance and soil water extraction of a perennial C4 grass: contributions of root and rhizome traits. Funct Plant Biol 41:505–519 Zwerts JA, Prins HHT, Bomhoff D, Verhagen I, Swart JM, de Boer WF (2015) Competition between a lawn-forming 123 Genet Resour Crop Evol Cynodon dactylon and a tufted grass species Hyparrhenia hirta on a South-African dystrophic savanna. PLoS ONE 10:e0140789. https://doi.org/10.1371/journal.pone.01407 89 123 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.