Cornelia Löhne

Documented living plant collections distinguish botanic gardens from other green spaces and horticultural landscapes. With more than 3,500 collections worldwide, these institutions steward at least 105,634 species—around 30% of all land plant diversity—while fulfilling amenity, educational, scientific and conservation roles. However, twenty-first-century challenges demand a re-evaluation of how these collections are documented and managed. We argue that meeting these emerging needs requires higher standards of coordinated information management and innovation in data infrastructures across the global network. This Perspective critically examines data management practices of living collections supporting scientific research and conservation, from institutional to global levels. We identify the renewed demands on living collections, highlight exemplar global data infrastructures, define data challenges inherent to living collections and explore how current systems fall short in enabling a connected global system. Finally, we outline a vision for high-performance collections, fully integrated into a robust global data ecosystem.

Ex situ living plant collections play a crucial role in providing nature-based solutions to twenty-first century global challenges. However, the complex dynamics of these artificial ecosystems are poorly quantified and understood, affecting biodiversity storage, conservation and utilization. To evaluate the management of ex situ plant diversity, we analysed a century of data comprising 2.2 million records, from a meta-collection currently holding ~500,000 accessions and 41% of global ex situ species diversity. Our study provides critical insights into the historical evolution, current state and future trajectory of global living collections. We reveal sigmoidal growth of a meta-collection that has reached capacity in both total accessions and total diversity, and identify intrinsic constraints on biodiversity management, including a median survival probability of 15 years. We explore the impact of external constraints and quantify the influence of the Convention on Biological Diversity, which we link to reduced acquisition of wild-origin and internationally sourced material by 44% and 38%, respectively. We further define the impact of these constraints on ex situ conservation but highlight targeted initiatives that successfully mitigate these challenges. Ultimately, our study underscores the urgent need for strategic prioritization and the re-evaluation of ex situ biodiversity management to achieve both scientific and conservation goals.

Field observations of morphologically intermediate water lilies in central Canada suggested a hybrid origin involving the parents Nymphaea odorata Aiton and Nymphaea leibergii Morong despite the fertile nature of these plants. Sequencing of the nrITS and the plastid rps4–trnT–trnF regions further including all members of the north temperate Nymphaea subg. Nymphaea clade, and samples from other hybrids occurring in North America in the wild, allowed us to determine that individuals of N. leibergii and of N. odorata were the maternal and paternal parents, respectively. Hybrids of New England have all proven to be sterile, are genetically variable, and probably are F1. By comparison, the plants of east-central Saskatchewan and west-central Manitoba are fully fertile and genetically uniform based on ISSR and sequence data. On the basis of this evidence, the latter are here described as a new species, Nymphaea loriana sp. nov., which may have originated during the Holocene climatic optimum about 6000 years ago in a past contact zone of the parents. Further hybrids detected between N. odorata and Nymphaea tetragona Georgi, as well as between N. leibergii and N. tetragona, were always sterile. Gene trees of the temperate clade of Nymphaea converge on a clade of small-flowered water lilies (sect. Chamaenymphaea), including N. leibergii, N. tetragona, and Nymphaea pygmaea. Nuclear ITS further resolves an American clade (Nymphaea mexicana Zucc. – N. odorata) sister to all remaining species. This split into two major subclades also appears in the otherwise less resolved rps4–trnT–trnF tree. Thus the origin of N. loriana is a reticulation between long-separated parental lineages.

The Nymphaeales (water lilies) clade has diverged as the second branch in the tree of angiosperms and is composed of the three families Cabombaceae, Nymphaeaceae and Hydatellaceae. Extant species diversity is constituted by 82 species, about half of which belong to the nearly globally distributed genus Nymphaea . DNA sequence datasets of multiple non‐coding and rapidly evolving regions from all three genomic compartments (ca. 8 kb of sequence per taxon) for a dense sampling of Nymphaeales, Austrobaileyales and Amborella were examined. In an attempt to review the literature on water lilies of the past decades a matrix comprising 62 morphological characters was generated. The crown group of extant Nymphaeales is supported by a series of synapomorphies, several of which have evolved in line with the acquisition of herbaceous habits and adaptations to an aquatic lifestyle such as the loss of cambium and sclerenchyma. Further innovations evolved subsequently within the diversification of the water lily crown group such as hydropotes, or an aril as floating device for the seeds in core Nymphaeaceae that have evolved fruits ripening under water. Both Hydatellaceae and Cabombaceae exhibit many derived features that in part may be explained as adaptations to anemophily. The Nymphaeaceae are supported as monophyletic by most character partitions, including morphology, as are Nuphar and Barclaya as successive sisters to the core Nymphaeacae ( Nymphaea, Ondinea, Euryale, Victoria ). Parsimony analysis of the morphological dataset alone yielded a well resolved and statistically supported tree. Ondinea appears as a close relative of the Australian subg. Anecphya clade within Nymphaea by all genomic compartments and morphology. Earlier hypotheses of Nymphaea being paraphyletic to the Euryale‐Victoria clade are inferred in nuclear trees, albeit with low support. Different morphological characters equivocally support a position of the Euryale‐Victoria clade as sister to the subg. Hydrocallis‐Lotus clade within Nymphaea or as sister to all species of Nymphaea . The diversification of the water lily clade is further characterized by a trend towards increased complexity in floral architecture.

With the present study we attempt to elucidate the history of the order Nymphaeales—water‐lilies and relatives—in time and space. On the basis of a dense taxon sampling that covers all genera of Cabombaceae and Nymphaeaceae and all subgenera of the genus Nymphaea , and on the basis of well‐supported phylogenetic hypotheses, we estimate divergence times in Nymphaeales. Distribution data for all species are used to reconstruct ancestral ranges and to identify possible dispersal events in the biogeographic history of Nymphaeales. Considering the fossil record for the clade as well as geologic history, a plausible historical scenario can be drawn. The assessment of extant biogeography shows that there are several centres of species diversity for Nymphaeales, i.e., northern South America, Central America, the Zambezian region of Africa and northern Australia. However, the diversification of the Nymphaeales started in the Northern Hemisphere when the northern landmasses were covered by tropical vegetation, except for the very high latitudes. The estimation of divergence times depicted two distinct radiation events, a rapid first differentiation into three major lineages during the Paleocene (Cabombaceae, Nuphar , remaining Nymphaeaceae) and the radiation of core Nymphaeaceae ( Victoria , Euryale , Nymphaea incl. Ondinea ) from the Late Oligocene to Middle Miocene. The second radiation probably started in the Northern Hemisphere. We hypothesize that subsequent spatial separation and southward relocation of the descendants led to the disjunct distribution of extant sister lineages in core Nymphaeaceae, e.g., Euryale‐Victoria or Nymphaea subgg. Hydrocallis and Lotos .

This study represents the first comprehensive analysis of phylogenetic relationships within the Australian water-lilies, Nymphaea subg. Anecphya. Our 51-accession dataset covers all 10 species of the subgenus, except the newly described N. alexii, and includes information from the nuclear ITS as well as from the chloroplast trnT–trnF region. The results show that molecular data are consistent with morphology, because the subdivision of subg. Anecphya into two major clades, a large-seeded and a small-seeded group, could be confirmed. Within the large-seeded group, Nymphaea atrans and N. immutabilis seem to form one clade, whereas samples of N. gigantea, N. georginae, N. macrosperma and N. carpentariae form another. Relationships within the small-seeded group, containing all samples of N. violacea, N. elleniae and N. hastifolia, are less clear, since the trees obtained from the chloroplast and the nuclear marker are incongruent. The samples of N. violacea do not form a monophyletic group in each of the trees, but—at least in the ITS tree—group with either N. elleniae or N. hastifolia/Ondinea, respectively. Polymorphisms among ITS paralogues, i.e. substitutions at single nucleotide positions and length polymorphisms, have been observed in some samples of N. violacea. This fact as well as the incongruent phylogenetic signal obtained from the chloroplast and the nuclear genomes point to recent hybridisation or introgression in this group. Remarkably, Ondinea purpurea is resolved within the small-seeded group by both markers and seems to have a close relationship to N. hastifolia. Although incomplete lineage sorting cannot be fully excluded to explain high variability in N. violacea, molecular data potentially hint to a case of still imperfect taxonomy.