Abstract
We present a densely-sampled phylogenomic study of the mulberry tribe (Moreae, Moraceae), an economically important clade with a global distribution, revealing multiple losses of inflexed stamens, a character traditionally used to circumscribe Moreae. Inflexed stamens facilitate ballistic pollen release and are associated with wind pollination, and the results presented here suggest that losses of this character state may have evolved repeatedly in Moraceae. Neither Moreae nor several of its major genera (Morus, Streblus, Trophis) were found to be monophyletic. A revised system for a monophyletic Moreae is presented, including the reinstatement of the genera Ampalis, Maillardia, Pachytrophe, Taxotrophis, and Paratrophis, and the recognition of the new genus Afromorus, based on Morus subgenus Afromorus. Pseudostreblus is reinstated and transferred to the Parartocarpeae, and Sloetiopsis is reinstated and transferred to the Dorstenieae. The tribe Olmediae is reinstated, replacing the Castilleae, owing to the reinstatement of the type genus Olmedia, and its exclusion from Moreae. Streblus s.s. is excluded from Moreae and transferred to the Olmediae, which is characterized primarily by involucrate inflorescences without regard to stamen position. Eight new combinations are made.
Introduction
The preservation of plesiomorphic (ancestral) characters can result in species that are similar in appearance but distantly related, connected only by a remote common ancestor. The mulberry family (Moraceae Gaudich., seven tribes, ca. 39 genera and 1,200 species) illustrates this principle well. Inflexed stamens in bud—an adaptation to wind pollination that allows explosive pollen dispersal when flowers open—were traditionally used to define a tribe of the family, the Moreae (mulberries and their allies). Yet phylogenetic analyses have revealed that inflexed stamens, an ancestral feature of both Moraceae and their sister family Urticaceae (nettles), have been lost repeatedly (Clement & Weiblen, 2009). Thus, for example, Cecropiaceae C.C. Berg, traditionally distinguished from the nettle family by the absence of inflexed stamens, is in fact embedded within the Urticaceae Juss. (Berg, 1978; Clement & Weiblen, 2009). Likewise, while mulberries (Morus L.), paper mulberries (Broussonetia L’Hér. ex Vent.), and osage oranges (Maclura Nutt.) were once treated as tribe Moreae Gaudich. on account of their inflexed stamens, phylogenetic analyses reveal them to belong to three distinct clades, each containing an assemblage of genera with and without inflexed stamens (Clement & Weiblen, 2009; Zerega & Gardner, 2019).
This study focuses on the tribe Moreae, a clade of six genera and an estimated 66 species (Clement & Weiblen, 2009). This widespread group of plants contains species of ecological and cultural importance, as well as the economically important white mulberry (Morus alba L.), whose leaves sustain Bombyx mori L. (Bombycidae), the invertebrate proletariat of the silk industry.
Morphological basis for higher classification in Moraceae
By the end of the nineteenth century, Engler (1889) had circumscribed a Moraceae that is quite similar to the modern concept of the family, with two subfamilies: the Moroideae, comprising tribes Dorstenieae Gaudich., Broussonetieae Gaudich., Fatouae Engler, Moreae, and Strebleae Bureau, characterized by stamens inflexed in bud (broadly construed, including Dorstenia where they straighten gradually rather than spring outward suddenly); and the Artocarpoideae, composed of tribes Brosimae Trécul, Euartocarpeae Trécul, Ficeae Dumort., and Olmedieae Trécul and characterized by stamens straight in bud. The two most influential scholars of Moraceae classification in the 20th century were E.J.H. Corner and C.C. Berg. Corner considered inflorescence architecture to be the most important character for higher-rank taxonomy within the family and inflexed stamens consequently of secondary importance(Corner, 1962). Berg by contrast questioned the utility of inflorescence architecture and took into account a variety of characters, especially the presence of inflexed stamens (Berg, 1977b, 2001).
The taxonomic history of the family has leaned heavily on this stamen character. Engler’s Moreae (1899), unchanged in substance from Bureau’s (1873), contained six genera, Ampalis Bojer, Pachytrophe Bureau, Paratrophis Blume, Pseudomorus Bureau, and Morus, all with inflexed stamens that spring out suddenly at anthesis (Table 1). Corner’s expanded Moreae consisted of seven genera with either straight or inflexed stamens, but with pistillate inflorescences never condensed into a head: Fatoua Gaudich, Morus, Sorocea A. St.-Hil. (apparently including Paraclarisia Ducke), Clarisia Ruiz & Pav., Ampalis, Pachytrophe, and Streblus Lour. (including Bleekrodea Blume, Paratrophis, Pseudomorus, Taxotrophis Blume, Sloetiopsis Engl. and Neosloetiopsis Engl.). Corner himself, however, found the diversity of his Moreae unsatisfactory, noting that “[t]oo many genera on insufficient and invalid grounds trouble this small tribe” (Corner, 1962). Perhaps in response, Berg’s Moreae comprised all of the genera with inflexed stamens and none without, including Broussonetia and Maclura but excluding Sorocea and Clarisia, providing a simple character with which to delimit the tribe(Berg, 2001; Berg & al., 2006).
Recent phylogenetic work has supported a third approach, with the Moreae comprising six genera, including genera with both straight (Sorocea, Bagassa Aubl.) and inflexed (Broussonetia, Maclura) stamens (Clement & Weiblen, 2009). These studies also suggest that the character state of inflexed stamens is plesiomorphic and is the ancestral state for both Moraceae and Urticaceae (Datwyler & Weiblen, 2004; Zerega & al., 2005; Clement & Weiblen, 2009). Inflexed stamens, which are precursors to ballistic pollen release, are associated with wind pollination (Bawa & Crisp, 1980; Berg, 2001), while the loss of inflexed stamens is associated with animal pollination. Although dominant in the Moreae, inflexed stamens also occur in two other tribes: Maclureae W.L. Clement & G. Weiblen and Dorstenieae. Genera lacking inflexed stamens occur in all seven tribes of Moraceae (Clement & Weiblen, 2009; Zerega & Gardner, 2019).
Taxonomic summary of Moreae
Following Clement & Weiblen (2009), the Moreae comprise six genera and an estimated 66 species: Bagassa (1), Milicia Sim (2), Morus (16), Sorocea (16), Streblus (23), and Trophis P. Browne. (8) (Berg, 1977a, 2001; Berg & al., 2006; Clement & Weiblen, 2009; Filho & al., 2009; Machado & al., 2013; Santos & Neto, 2015). Here, we present an overview of these genera.
Morus L., the true mulberries, is characterized by leaves with crenate margins and trinerved bases, stamens inflexed in bud, and many-flowered pistillate spikes whose four-parted perianths become fleshy in fruit, the aggregations superficially resembling a blackberry. Morus comprises approximately 16 species whose delimitation requires further research. There are three subgenera: Morus (ca. 14 species) which is found in temperate to tropical Asia and from North America to Mexico; Gomphomorus Leroy, a single species restricted to tropical South America; and Afromorus (Bureau ex Leroy), a single species restricted to tropical Africa. Previous phylogenetic work has suggested that these three subgenera may not form a monophyletic clade (Nepal, 2012). Milicia (2 spp., Africa) has inflorescences which somewhat resemble those of Morus, but the leaves of Milicia, with entire margins and pinnate venation, prevent any confusion of the two genera.
Streblus Lour., with 23 species from India to Southeast Asia and Oceania, is morphologically heterogeneous, but its species are all characterized by stamens inflexed in bud and pistillate flowers with more or less free tepals that enclose the fruit loosely or not at all. Initially described by Loureiro based on the widespread S. asper Lour.—notable for its discoid-capitate staminate inflorescences with the rudiments of an involucre—Streblus was broadened by Corner (1962), bringing in as sections Taxotrophis, Phyllochlamys Bureau, Paratrophis (including Pseudomorus), Pseudostreblus Bureau, Bleekrodea, and Sloetia Teijsm. & Binn. (apparently including Sloetiopsis but without making any combinations). None of these have discoid-capitate inflorescences; they mostly have spicate staminate inflorescences, and the latter two can have bisexual inflorescences.
Corner viewed these sections as fragments of an ancient lineage preserving ancestral characters (Corner, 1975). Following additional work by Corner (1970, 1975), Berg included the genera Ampalis and Pachytrophe as section Ampalis (Bojer) C.C. Berg, subsumed Sloetiopsis and section Taxotrophis into section Streblus, and excluded section Bleekrodea, reinstating it as a genus (Berg, 1988). Berg later reinstated section Taxotrophis, whose species are unique in bearing axillary spines (Berg, 2005; Berg & al., 2006). In 2009, Clement and Weiblen reinstated Sloetia at generic rank and transferred it and Bleekrodea to Dorstenieae based on phylogenetic evidence (Clement & Weiblen, 2009).
Trophis P. Browne is characterized by stamens inflexed in bud, spicate staminate inflorescences, and tubular pistillate perianths, becoming fleshy and enclosing the fruit, except for the monotypic section Olmedia (Ruiz & Pav.) C.C. Berg (T. caucana), which has discoid-capitate staminate inflorescences with an involucre. Trophis (regrettably conserved over Linnaeus’s Bucephalon L.) as recognized by Berg (1988, 2001) had five sections: Trophis, restricted to Latin America; Calpidochlamys (Diels) Corner (previously included in Paratrophis and Uromorus), restricted to Southeast Asia (Corner, 1962); Maillardia (Frapp. ex Duch.) C.C. Berg, restricted to Africa; Malaisia (Blanco) C.C. Berg, restricted to Southeast Asia; and Olmedia, restricted to Latin America. The sinking of Olmedia into Trophis by Berg (1988), requiring the re-typification of the tribe Olmedieae Trécul, which became Castilleae C.C. Berg. Recently, Malaisia Blanco was reinstated as a genus and transferred to Dorstenieae based on phylogenetic evidence (Clement & Weiblen, 2009), reducing the current number of sections to four.
Two Neotropical genera included in Moreae by Clement and Weiblen (2009) but not by Berg (2001) have straight stamens. While the monotypic Bagassa Aubl., with its long staminate catkins, is wind pollinated (Bawa & Crisp, 1980), evidence suggests that Sorocea A. St.-Hil. (19 spp.), which produces racemose staminate inflorescences that do not have as many flowers as those of Bagassa, likely contains both wind- and insect-pollinated species (Zapata & Arroyo, 1978; Bawa & al., 1985; Lewis, 1986). The pistillate perianths are subtended by pluricellular hairs, believed to serve as a substrate for a fungus which, in turn attracts pollinators (Berg, 2001). The staminate flowers of Sorocea affinis Helmsl. are apparently fragrant (fide S. Zona 778, 21 Nov. 1997, FTBG♂♀), suggesting insect pollination for that species. If insect pollination is confirmed in Sorocea, it would likely represent another transition from ancestral wind to derived insect pollination in Moraceae.
Disagreement between the two principal recent monographers of the Moraceae, Corner (1962) and Berg (2001, 2005, 2006), over the delimitation of the Moreae has been compounded by a series of phylogenies based on the analysis of molecular (DNA sequence) data that have redefined genera and their relationships. This has resulted in a poorly delimited tribe, both with respect to diagnostic morphological characters, the genera it comprises, and the rank of several taxa (see above). Ensuring that the tribe and its genera are monophyletic will result in a classification that better reflects evolutionary history, providing a framework for answering broader scientific questions. Our aim, therefore, was to generate a comprehensive phylogeny of the Moreae (Fig. 1) and its near allies by sampling all of the potential genera and species in the tribe sensu Berg, and sensu Clement & Weiblen (2009) and use this to redelimit the tribe and its genera. We aimed to do so through increased taxon and genome sampling compared to previous studies, including a nearly comprehensive sample of the taxa in Moreae (56/66 species) and the allied tribes Artocarpeae (76/83 taxa) and Maclureae (12/12). Our data set combines phylogenomic data generated using two largely non-overlapping sets of enrichment baits, one developed specifically for the Moraceae (Gardner et al., 2016), the other for the whole of the Angiosperms (Johnson & al., 2019), allowing us to explore the possibilities and challenges of combining samples based on largely non-overlapping loci. The resulting generic revision lays the groundwork for species-level revisionary work and provides clarity to this economically important clade.
We also set out to reconstruct the evolutionary history of inflexed stamens within Moraceae using ancestral state reconstruction. The loss of inflexed stamens in Castilleae+Ficeae and Artocarpeae have been associated with transitions from wind to animal pollination (Momose & al., 1998; Sakai & al., 2000; Datwyler & Weiblen, 2004; Gardner & al., 2018). A more complete picture of evolutionary transitions between inflexed and straight stamens may help focus further research on transitions in pollination biology with Moraceae.
Materials and Methods
We used target enrichment sequencing (HybSeq) (Weitemier & al., 2014) to capture 333 genes previously developed for phylogenetic work in Moraceae (the “Moraceae333”) (Gardner & al., 2016; Johnson & al., 2016). This method allows for efficient capture of hundreds of loci and is suitable for both fresh material and degraded DNA from herbarium material (Villaverde & al., 2018; Brewer & al., 2019), which comprises much of the material employed in this study.
Taxon sampling, library preparation, and sequencing
We sampled 56 out of 66 species (and all genera) in Moreae as well as select outgroup taxa from Artocarpeae using DNA from leaf tissue preserved on silica gel or—in most cases—samples from herbarium specimens (Table S1). In either case, DNA was extracted using a modified CTAB method, usually with increased incubation times to maximize yield from herbarium tissue(Doyle & Doyle, 1987; Hale & al., 2020). Samples were quantified using a Qubit fluorometer (Invitrogen, Life Technologies, California, USA), and herbarium samples were run on a gel to test for degradation. For most samples 200 ng of DNA was used for library preparation; for some low-yield samples, as little as 50 ng was used, and for very degraded samples, as much input as possible was used, up to 400 ng. Undegraded DNA was fragmented either using NEB DNA Fragmentase (New England Biolabs, Ipswich, Massachusetts, USA) or on a Covaris M220 (Covaris, Wobum, Massachusetts, USA). DNA samples with an average fragment size less than 500 bp were not fragmented at all, and partially-degraded samples with an average fragment size of over 500 bp were fragmented on the Covaris M220. TruSeq-style library preparation was carried out using either the KAPA Hyper Prep kit (Kapa Biosystems, Wilmington, MA) or the NEB DNA Ultra 2 kit following the manufacturer’s protocols, except that end repair, A-tailing, and adapter ligation were carried out in reduced-volume reactions (0.25x for KAPA and 0.5x for NEB) to reduce costs. Final products were quantified on the Qubit and combined in equal molecular weights into pools of 16–20 samples. The pools totaled 1,200 µg each if enough library preparation was available. Pools were hybridized for 16–24 hours to custom Moraceae probes (Gardner et al, 2016) using a MYBaits kit (Arbor Biosciences, Ann Arbor, Michigan, USA) following the manufacturer’s protocol, except that the probes were diluted 1:1 with nuclease-free water. Hybridization products were reamplified using KAPA Hot Start PCR reagents following the MYbaits protocol, quantified on the Qubit, and quality-checked on an Agilent BioAnalyzer (Agilent Technologies, Palo Alto, California, USA). Samples with adapter dimer peaks were cleaned using 0.7x SPRI beads and re-run on the Qubit and BioAnalyzer. An initial sequencing run took place on a MiSeq 2 x 300bp run (v3) (Illumina, San Diego, California, USA) at the Field Museum of Natural History, and then additional samples were sequenced on a HiSeq 4000 2 x 150bp run at the Northwestern University Genomics Core.
We also included 47 Moraceae and Urticaceae samples enriched for the Angiosperm353 probes and sequenced as part of the Plant and Fungal Trees of Life project (PAFTOL, RBG Kew; https://www.kew.org/science/our-science/projects/plant-and-fungal-trees-of-life;Table S1). Sample preparation and sequencing followed Johnson et al. (2019). The PAFTOL samples were enriched with a universal probe set developed for angiosperms (the “Angiosperms353”).
Finally, we used samples sequenced for previous phylogenetics projects in Artocarpus J.R. Forst. & G. Forst. and Parartocarpeae Zerega & E.M. Gardner to complete our sampling (Johnson & al., 2016; Gardner, 2017; Kates & al., 2018; Zerega & Gardner, 2019). The final dataset contained 247 samples.
Assembly of reads
We trimmed reads using Trimmomatic (ILLUMINACLIP: TruSeq3-PE.fa:2:30:10 HEADCROP:3 LEADING:30 TRAILING:25 SLIDINGWINDOW:4:25 MINLEN:20) (Bolger & al., 2014) and assembled them with HybPiper, which produces gene-by-gene, reference-guided, de novo assemblies (Johnson & al., 2016). For the samples enriched with the Angiosperms353 baits, we used the reference described by Johnson et al. (2019), and for the samples enriched with the Moraceae333 baits, we used the reference described in Zerega & Gardner (2019). To increase overlap between the two data sets beyond the 5 genes they inherently have in common, we used HybPiper to assemble the Angiosperms353-enriched reads using the Moraceae333 targets and vice-versa, in order to capture any additional genes found in off-target reads. The exception was most of the Artocarpus data set, which was previously assembled for another study using only the Moraceae333 targets. We used the HybPiper script “intronerate.py” to build “supercontig” sequences for each gene, consisting of exons as well as any assembled flanking non-coding sequences (intronic or intergenic). To these new assemblies, we added 76 Moraceae333 assemblies from Gardner (2017) to complete sampling for the Artocarpeae.
Main phylogenetic analyses
The full data set was analyzed using the “exon” sequences to ensure good alignment across the entire family. To maximize phylogenetic resolution within Moreae, a smaller data set consisting only of Moreae taxa and a single outgroup taxon (Artocarpus heterophyllus) was analyzed using the “supercontig” sequences but following the same methodology. For each of the 686 genes assembled, we discarded sequences shorter than 25% of the average length for the gene and discarded genes containing sequences for less than 30 samples. We aligned each gene with MAFFT (Katoh & Standley, 2013)and used Trimal (Capella-Gutiérrez & al., 2009) to discard sequences with an average pairwise identity of less than 0.5 to all other sequences in the alignment (indicative of poor a poor quality sequence that could not be properly aligned) as well as columns containing more than 75% gaps. We used RAxML 8.2.4 (Stamatakis, 2014) to estimate a maximum-likelihood tree for a partitioned supermatrix of all genes as well as for each gene individually (GTRCAT model, 200 bootstrap replicates). We then used ASTRAL-III (Zhang & al., 2017), a summary-coalescent method, to estimate a species tree from all gene trees, estimating node support using both bootstrap (160 replicates, resampling across and within genes) and local posterior probability (normalized quartet scores, representing gene tree concordance). Finally, we used ASTRAL-III to test whether, for each node, the null hypothesis of a polytomy could be rejected (Sayyari & Mirarab, 2018). Alignment, trimming, and estimation of gene trees were parallelized using GNU Parallel (Tange, 2018).
Whole chloroplast phylogenetic tree
We also built a whole-chloroplast phylogenetic tree as follows. Rather than assembling full-length genomes, which can be extremely slow to align, we assembled and aligned the genomes in sections, dramatically speeding up the process. We created HybPiper targets using the chloroplast genome of Morus indica L. (NCBI RefSeq accession no. NC_008359.1) and the associated gene annotations. Each target consisted of a gene feature concatenated with any internal or subsequent non-coding sequence, terminating one base before the next gene feature began. These intervals were generated by manually editing the NCBI gff3 file in Excel (Microsoft Corp., Redmond, Washington, USA) and then extracted using BedTools (Quinlan & Hall, 2010). Assemblies for all Moraceae samples and Boehmeria nivea (L.) Gaudich. were carried out in HybPiper using a coverage cutoff of 2 and otherwise with default parameters. Within each target, sequencing less than 25% of the average length were discarded, and after alignment with MAFFT, sequences with an average pairwise identity of less than 0.7 were discarded (the higher cutoff reflecting the conserved nature of chloroplast DNA). Alignments were then concatenated into a supermatrix, and samples with more than 50% undetermined characters were discarded. Alignments were then rebuilt, filtered, and concatenated, and columns in the final supermatrix with more than 75% missing characters were discarded. A maximum-likelihood tree was generated using RAxML 8.2.4 under the GTRCAT model, with 200 rapid bootstrap replicates.
Phylogenetic network analysis
To further investigate relationships within the Paratrophis clade, including the proper placement of Morus insignis and Trophis philippinensis (Bureau) Corner, we constructed phylogenetic networks based on two reduced 8-taxon datasets (“exon” and “supercontig”) consisting of those taxa plus Streblus anthropophagorum (Seem.) Corner, S. glaber (Merr.) Corner, S. glaber subsp. australianus (C.T. White) C.C. Berg, and S. heterophyllus (Blume) Corner, with S. mauritianus (Jacq.) Blume as the outgroup. Alignment preparation followed the workflow outline above except that only genes with sequences for all eight taxa were retained. Rooted gene trees were used to infer the network in PhyloNet using the “InferNetwork_MPL” command, allowing a maximum of four hybridization events and collapsing gene tree nodes with less than 30% bootstrap support (Yu & Nakhleh, 2015; Wen & al., 2018) (Wen et al., 2018; Yu and Nakhleh, 2015); AIC was scored for the best five networks from 10 runs, using the number of branch lengths and hybridization events calculated as the number of parameters (Kamneva & al., 2017).
ITS and rbcL phylogenetic trees
While we did not have our own sequences for Streblus tonkinensis (Eberh. & Dubard) Corner, S. ascendens Corner, S. banksii (Cheeseman) C.J. Webb, and S. smithii (Cheeseman) Corner, either ITS or rbcL sequences existed in NCBI GenBank for these taxa (Fig. S2). We were not able to examine the underlying specimens ourselves, but we considered the chances of misidentification low because those species are morphologically and/or geographically distinctive. We constructed ITS and rbcL data sets from our own samples using HybPiper, using Streblus sequences for these loci (obtained from GenBank) as targets and reducing the coverage cutoff at 2 to recover the loci from off-target reads. To these sequences, we added GenBank sequences of Streblus taxa as well as Trophis caucana (Pittier) C.C. Berg. For each locus, we aligned sequences with MAFFT, discarded short sequences (<490 bp), trimmed alignments to remove columns with over 75% gaps, and built maximum-likelihood trees using RAxML (GTRGAMMA, 1000 rapid bootstraps).
Divergence time estimation
The supermatrix maximum-likelihood tree was time calibrated using ape v 5.3 (Paradis & Schliep, 2019) in R v 3.5.1 (Team, 2019). First, the tree was pruned of duplicate taxa and non-Moraceae outgroup taxa, and edge lengths (in substitutions per site) were multiplied by the number of sites in the alignment (to convert them to substitutions). The following stem nodes were constrained with minimum ages (in million years, Ma) based on fossil data, following(Zhang & al., 2019): Ficus, 56 Ma; Broussonetia, 33.9 Ma; Morus (subg. Morus, based on U.S.S.R. locality of the fossil), 33.9 Ma; and Artocarpus, 64 Ma.
The crown node of Moraceae was constrained to a minimum age of 73.2 Ma and a maximum age of 84.7 Ma.(Zhang & al., 2019), which had a more extensive outgroup sampling that we have here. The tree was then time-calibrated using the chronos function under two different models (“relaxed” and “correlated”) (Kim & Sanderson, 2008; Paradis, 2013). The smoothing parameter (λ) was chosen using the cross-validation method in the chronopl function (testing λ = 0 and 0-5 through 1015), selecting the value of λ that minimized the cross-validation statistic (Sanderson, 2002). The resulting trees were visualized using the densiTree function in phangorn 2.4.0 (Schliep, 2011), and the time calibrated tree representing the central tendency of these analyses was selected for use in all further analyses, also taking into account the results of past family-wide studies. As the penalized likelihood approach used does not integrate over model uncertainty or uncertainty in calibration placement and timing, confidence intervals on node ages are not provided in this study. A geologic timescale based on the strat2012 dataset added to tree figures using PHYLOCH v 1.5-3 (Heibl, 2008).
Ancestral state reconstruction for inflexed stamens
All taxa were coded for presence (1) or absence (0) of inflexed stamens in bud. Taxa such as Dorsteniae with stamens that are inflexed in bud but gradually straighten were coded as 0. We reconstructed ancestral character states on the entire phylogeny by stochastic mapping using the make.simmap function in the phytools v 0.6-99 (Revell, 2012). We also tested for trait-associated shifts in diversification rates using BAMM v 2.5.0 (Rabosky & al., 2013, 2014a; Rabosky, 2014), specifying the amount of missing taxa per clade to account for the high proportion of missing taxa in Ficeae, Castilleae, and Dorstenieae. Parameters were optimized using BAMMtools v 2.1.7 (Rabosky & al., 2014b). We also tested for trait-dependent diversification using diversitree v 0.9-13 (FitzJohn, 2012). We evaluated BiSSE and trait-independent models on the entire tree and on a pruned tree consisting only of the densely-sampled Maclureae+Moreae+Artocarpeae clade.
Reads from have been deposited in Genbank (Non-PAFTOL samples: BioProject PRJNA322184; PAFTOL samples: BioProject PRJEB35285, subprojects PRJEB37667 for Moraceae and PRJEB37665 for Urticaceae). Tree files and scripts outlining the phylogenetic analyses have been deposited in the Dryad Data Repository (###TBA###).
Results
The final data set contained 247 samples, 47 enriched with the Angiosperms353 baits, 196 enriched with the Moraceae333 baits, and 4 extracted from whole genomes) and 619 genes (286 Angiosperms353 and all of the Moraceae333) (Table S1). The “exon” supermatrix (all taxa) contained 613,126 characters, and the “supercontig” supermatrix (Moreae and select outgroup taxa only) contained 799,926 characters. The chloroplast data set contained 113 loci.
On average, we assembled 39 Moraceae333 genes from the Angiosperms353-enriched samples and 20 Angiosperms353 genes from the Moraceae333-enriched samples, and the average pairwise overlap in assembled loci was 210 (Tables S1, S2). Nineteen taxa were represented in both sets of samples, and 15 of these were always monophyletic (Figs. 2, 3). Twelve resolved as sister pairs: Antiaropsis decipiens (31 loci overlapping), Bagassa guianensis Aubl. (38), Batocarpus amazonicus (Ducke) Fosberg (22), Batocarpus orinocensis H. Karst. (34), Brosimum alicastrum Sw. (42), Clarisia racemosa Ruiz & Pav. (59), Maclura africana (Bureau) Corner (213), Milicia excelsa (Welw.) C.C. Berg (45), Streblus asper (Retz.) Lour. (75), Streblus mauritianus (Jacq.) Blume (40), Streblus usambarensis (Engl.) C.C. Berg (43), and Trophis caucana (32). Three more represented by more than two samples always resolved as a clade: Sorocea bonplandii (Baillon) W.C. Burger, Lanj. & de Boer (3 samples; 19–35 loci overlapping), Maclura tinctoria (L.) D. Don ex Steud. (4 samples; 37–43 loci overlapping), and Malaisia scandens (Lour.) Planch. (3 samples; 27–43 loci overlapping). Three additional taxa resolved as sister pairs in the ASTRAL analysis but as a grade in the supermatrix analysis: Maclura cochinchinensis (Lour.) Corner (28 loci overlapping), Streblus heterophyllus (Blume) Corner (50), and Trophis montana (Leandri) C.C. Berg (56). Only one species, Utsetela gabonensis Pellegr. (18 loci overlapping) was never monophyletic, always forming a grade with U. neglecta Jongkind.
The supermatrix and species trees were broadly concordant except within Artocarpus, where the species tree was much more consistent with previous analyses within that genus (Gardner et al., in review) (Figs. 2, 3). Otherwise, most differences were at shallow phylogenetic depths, such as relationships within Streblus section Paratrophis. For both data sets, the polytomy hypothesis was rejected (P < 0.05) for Moraceae and all tribe and genus-level clades (following the revised classification presented below) (Figs. 2, 3).
Of the genera in Moreae sensu Clement & Weiblen, only Milicia and Sorocea were monophyletic; Morus, Streblus, and Trophis were not; Bagassa is monotypic and resolved as sister to Sorocea, with which it shares straight stamens. Four Moreae species resolved with other tribes: Trophis caucana was nested within Castilleae, Streblus asper (Retz.) Lour. was sister to Castilleae, Streblus indicus (Bureau) Corner was sister to Parartocarpeae, and Streblus usambarensis was nested within Dorstenieae. Moreae was otherwise monophyletic, comprising the following subclades: (1) Streblus section Taxotrophis, sister to all other Moreae; (2) (a) Trophis section Maillardia, (b) Milicia + Morus subgenus Afromorus, (c) Streblus section Ampalis, (d) Streblus section Paratrophis in part + Morus subgenus Gomphomorus, (e) Streblus section Paratrophis in part + Trophis section Calpidochlamys; (3) Bagassa + Sorocea; (4) Trophis sections Trophis and Echinocarpa C.C. Berg + Morus. These are roughly geographic clades: (1) Southeast Asia; (2) (a) Madagascar, (b) Southeast Africa, (c) Madagascar, (d) Pacific + South America, (e) Southeast Asia + Pacific; (3) South America; (4) South America.
While Trophis philippinensis was inside the Paratrophis clade in all analyses, the position of Morus insignis was not stable. In the “exon” analyses, it was always in Paratrophis (Fig. 2), although the polotomy hypothesis could not be rejected for its position as sister to S. anthropophagorum+S. heterophyllus. In the “supercontig” and chloroplast analyses, it was sister to Paratrophis, and the polotomy test was rejected for that node (Figs. 3, S1). In the five phylogenetic networks reported (Fig. 4, Table S3), M. insignis was inside Paratrophis in three of them, had a hybrid origin involving Paratrophis in one, and was only unambiguously sister to Paratrophis in one network, although even in that one it was involved in a hybridization involving Paratrophis.
The chloroplast phylogenetic tree (Fig. S1) was generally in agreement with the nuclear phylogenetic trees, with three notable differences. Streblus indicus and the Parartocarpeae formed a grade paraphyletic to Dorstenieae, Castilleae, and Ficeae, rather than a clade. Streblus zeylanicus (Thwaites) Kurz was sister to S. taxoides (B. Heyne ex Roth) Kurz (instead of sister to all of section Taxotrophis), although with low support (62%), and S. glaber subsp. australianus was not sister to subsp. glaber, although its placement was not strongly supported (80%). Finally, Morus insignis was sister to Paratrophis, agreeing with the “supercontig” analysis but not the “exon” analysis.
The ITS and rbcL phylogenetic trees based on few characters, were not very well resolved, with only few nodes attaining 100% bootstrap support (Fig. S2). Nevertheless, with the exception of a few stray taxa (one Dorstenia sample in ITS and one Streblus indicus samples in rbcL), both phylogenetic trees reconstructed monophyletic tribes with at least moderate support. In the ITS phylogenetic tree, Streblus smithii and S. banksii were both in a well-supported clade comprising section Paratrophis, and S. tonkinensis was part of the Castilleae clade, which also contained S. asper and Trophis caucana. In the rbcL phylogenetic tree, the two Streblus ascendens samples, the four samples of the undetermined Streblus from Papua New Guinea, and S. smithii were part of a well-supported clade comprising section Paratrophis.
Time-calibration and ancestral state reconstruction
The cross-validation criterion was minimized by a smoothing parameter (λ) value of 106. The tree calibrated under the correlated model (logLik = −17; p-logLik = −17; ΦIC = 1108) was less sensitive to changes in λ and generally had younger ages than the tree calibrated under the relaxed model (logLik = −56.7; p-logLik = −53129706, ΦIC = 4209013), the latter of which had perhaps implausibly long terminal branches (Table 2, Fig. S3). The crown age of Moreae was Paleocene (59.1 Ma) under the correlated model and late Cretaceous (75.4 Ma) under the relaxed model. Because the correlated tree was more consistent with past family-wide studies (Zerega & al., 2005; Zhang & al., 2019), we used that tree for all further analyses (Fig. 5). The BAMM analysis found a single credible rate shift, not surprisingly at the crown node of Ficus (Fig. S4); that rate shift was—also not surprisingly—associated with a loss of inflexed stamens, which are never found in Ficus (P = 0.035); no rate shifts were found within Moreae. Ancestral reconstruction on the entire tree found that ancestral Moraceae had stamens inflexed in bud; these were lost nine times (in Ficeae, Olmedieae [=Castilleae], Parartocarpeae, twice in Dorstenieae, Maclura section Cudrania (Trécul) Corner, Artocarpeae, Bagassa, and Sorocea) and regained once (in Trophis caucaua) (Fig. 6). Model testing on both the whole tree and on the Moreae+Artocarpeae+Maclureae subtree indicated that a BiSSE model was not substantially better than a trait-independent model (Moreae only: AIC for BiSSE = 1110.4; trait-independent = 1112.8; ΔAIC = 2.4; whole family: AIC for BiSSE = 1488.3; trait-independent = 1490.4; ΔAIC = 2.1). In any event, the reconstructions were identical (Figs. 6).
Discussion
Sequencing and combination of data sets
This study was materially improved by our ability to combine samples enriched with two largely non-overlapping bait sets. Despite minimal by-design overlap, we were able to assemble many overlapping loci for samples with moderate to deep coverage (Table S2), adding taxa that would not otherwise have been included in the study and replicating taxa to confirm unexpected phylogenetic placement (e.g, Streblus indicus). Taxa replicated across the two data sets performed well in phylogenetic analyses, with 15 out of 19 always resolving as monophyletic in supermatrix analyses and 18/19 so resolving in ASTRAL analyses, with only Utsetela gabonensis always forming a grade instead (with a difficult-to-distinguish congener). Our results should embolden others to combine and repurpose data sets in similar ways.
Higher taxonomy in Moraceae and the delimitation of the tribe Moreae
Our results support Corner’s overall approach to the classification of Moraceae, if not all of its details, including the primacy of inflorescence architecture and the unreliability of inflexed stamens for higher taxonomy (Corner, 1962). Inflexed stamens are a plesiomorphy that was lost nine times in Moraceae (Fig. 6), a preserved ancestral character whose past taxonomic importance accounts for the rather extreme non-monophyly of the Moreae. Streblus s.l. provides the best illustration of this principle, appearing in four out of seven tribe-level clades within Moraceae. We may justly call its disparate sections, as Corner did, “fragments of an ancestral Streblus.” (Corner, 1975)—or in modern parlance, a paraphyletic remnant preserving plesiomorphic staminate flowers similar to those likely to have occurred in the ancestor of all Moraceae, with four free tepals and four inflexed stamens (Clement & Weiblen, 2009). If one of the goals of modern systematics is to establish taxonomic frameworks that reflect as far as possible real evolutionary relationships, our results may serve as a warning to carefully investigate whether characters used for taxonomy are derived (synapomorphic) or ancestral (symplesiomorphic).
Broadly speaking, a tribal delimitation based on inflorescence architecture (supplemented with other characters in some cases) agrees best with the phylogenetic trees presented here. Moreae (as revised below) have unisexual spicate or racemose inflorescences (the pistillate ones sometimes uniflorous), with the globose-capitate pistillate inflorescences of the monotypic Bagassa as the sole exception. Stamens may be either inflexed or straight. Elsewhere in the family, racemes and spikes are rare, with the former found in Maclura (in part) and the latter found in the related genera Broussonetia, Allaeanthus Thwaites, and Malaisia and arguably in Batocarpus H. Karst. and Clarisia; all of these have capitate pistillate inflorescences. The taxa of Streblus s.l. and Trophis s.l. that must be excluded from Moreae all have inflorescences that do not fit our general rule: discoid-capitate (Streblus asper and Trophis caucana), cymose (Streblus indicus), or bisexual (Streblus usambarensis). In these cases, perhaps Corner did not take his emphasis on inflorescence architecture quite far enough, including too wide of a variety in this one tribe. In critiquing the utility of inflorescence architecture for classification, Berg (1977b) noted the similarity in the inflorescence structure of Bleekrodea (then part of Moreae, and included in Streblus by Corner) to that of Utsetela and Helianthostylis (Dorstenieae), an observation that proved prescient when Bleekrodea was found to belong to Dorstenieae (Clement & Weiblen, 2009).
Olmedieae (as revised below, including Castilleae) can be defined entirely based upon the presence of a discoid inflorescence subtended by an involucre of imbricate bracts. Two species with such involucres previously classified as Moreae, Streblus asper and Trophis caucana (=Olmedia aspera) always appeared in the Olmedieae (=Castilleae) clade in our analyses (Figs. 2–3). Olmedia aspera (=Trophis caucana)—the nomenclatural type of the tribe Olmedieae—was transferred to Moreae because of its inflexed stamens and lack of self-pruning branches (Berg, 1977b). In inflorescence morphology, however, T. caucana closely resembles other Castilleae, always subtended by an involucre of imbricate bracts (Berg, 1977b, 2001). The staminate inflorescences of Streblus asper are strikingly similar to those of T. caucana, and it is remarkable that the affinity between S. asper and the Olmedieae has not been seriously considered until now. Previous barcoding or phylogenetic studies have placed Trophis caucana (Kress & al., 2009) and Streblus tonkinensis (Chen & al., 2016) (closely allied to S. asper) in the Castilleae clade, but those results went unremarked upon, perhaps because of the broad scale of the studies (respectively, forest community phylogenetics and the vascular plants of China).
The remaining five tribes can all be broadly defined based on inflorescence architecture as Corner argued, sometimes supplemented by other characters as Berg preferred, allowing of course for the exceptions made inevitable by the vicissitudes of evolution. Ficeae of course is defined by the syconium (essentially an urceolate disc that has been closed at the top). Maclureae have densely-packed globose infructescences and can be distinguished from Artocarpeae by their four stamens (inflexed in most sections) and armature. Artocarpeae also have densely-packed globose infructescences but only one straight stamen and no armature. Parartocarpeae have a few connate involucral bracts and (in large part) flowers embedded in fleshy receptacles. Dorstenieae are perhaps the most heterogenous group, but in large part they have bisexual inflorescences, often capitate or discoid, and often with ballistically-ejected endocarps.
The role of inflexed stamens in the evolution of Moraceae
The repeated losses of inflexed stamens (Fig. 6), which are associated with wind pollination (Bawa & Crisp, 1980; Berg, 2001), raise the possibility that transitions from wind to animal pollination, which have already been documented in Moraceae (Momose & al., 1998; Sakai & al., 2000; Datwyler & Weiblen, 2004; Gardner & al., 2018), are even more common within the family. Generally considered rare (Culley & al., 2002), the shift from wind to animal pollination may be a repeated feature of Moraceae deserving of further investigation. Further investigation of Sorocea, with its straight stamens and sometimes-scented inflorescences, may reveal that, like Artocarpus, it contains both wind and animal pollination. And while little is known about pollination in the Dorstenieae, the presence of unisexual inflorescences and inflexed stamens (e.g., Broussonetia) as well as bisexual inflorescences with straight stamens (e.g., Brosimum, Dorstenia) raises the possibility of transitions within that clade as well; taxa with inflexed stamens but bisexual inflorescences such as Bleekrodea and Sloetia (which is visited by bees, EMG pers. obs.) may represent remnants of an intermediate state. Finally, the conclusion that a single transition to insect pollination preceded the split between the Ficeae and Olmedieae (Castilleae) should be reevaluated in light of the position of Streblus asper as sister to the latter (Figs. 2, 3). The only shift in diversification rates our analyses recovered was on the branch leading to the very diverse Ficus, suggesting that the shift away from wind pollination may not by itself lead to increased diversification.
Explanation of taxonomic revisions
We present a generic revision of Moreae (Fig. 7) based on the present phylogenetic study as well as morphological characters. Arranging monophyletic and morphologically coherent genera requires one change of rank and 8 new combinations, but no entirely new names. These revisions provide a framework for within-genus revisionary work, which will require more intensive sampling and review of specimens within the genera circumscribed here. We provide an explanation of the taxonomic changes, followed by a formal presentation of the affected tribes and genera, with complete species lists and synonymy.
Trophis and the reinstatement of the Olmedieae — To make Trophis monophyletic, we propose that it be applied strictly to the Neotropical clade that includes the type species Trophis americana L. (=Trophis racemosa (L.) Urb.). To accomplish this, we reinstate the genus Maillardia and transfer Trophis philippinensis to Paratrophis (discussed below), reinstating its former name Paratrophis philippinensis (Bureau) Fern.-Vill. Because Trophis caucana does not belong in Trophis or Moreae but rather with the members of Castilleae, we reinstate its former name Olmedia aspera Ruiz & Pav. and transfer it to that tribe, whose former name Olmedieae must now be reinstated on account of its priority.
Streblus — We restrict the genus Streblus to three species comprising most of section Streblus—Streblus asper, Streblus tonkinensis, and Streblus celebensis C.C. Berg— and transfer the genus to Olmedieae. Although S. celebensis was not included in our phylogeny, the sub-involucrate inflorescences are similar to those of S. asper, with which S. celebensis differs primarily in vegetative characters, the latter having broadly-toothed margins in the distal half of the leaf. Both occur in Sulawesi, where at least one specimen with intermediate leaf morphology has been collected (Sulawesi, Kendari: Kjellberg 452, 24 Feb. 1929, L, det. S. asper by E.J.H. Corner). We therefore retain its taxonomic position in Streblus. The remaining member of section Streblus, S. usambarensis, belongs with Dorstenieae, and we therefore transfer it to that tribe, reinstating its former name Sloetiopsis usambarensis. Streblus indicus is sister to Parartocarpeae, and we therefore reinstate its former name Pseudostreblus indicus and transfer it to Parartocarpeae.
The remaining species of Streblus s.l. are properly placed in Moreae but are still paraphyletic. We therefore reinstate the genera Ampalis, Pachytrophe, Paratrophis, and Taxotrophis, largely corresponding to the former sections but requiring some new combinations. Paratrophis as presented here is united by spicate male inflorescences (usually) with peltate or reniform bracts (except for Paratrophis philippinensis (=Trophis philippinensis), which does not have fleshy tepals in fruit). Within Paratrophis we include Paratrophis ascendens (Corner) E.M. Gardner (=Streblus ascendens Corner) based on its inflorescence morphology and phylogenetic position in the rbcL tree (Fig. S2b). Its previous position within the monotypic section Protostreblus was due to the type specimen’s spiral phyllotaxy, but the latter may be atypical, as a more recent collection Womersley NGF 24791 (K, L, BO), has distichous leaves. Berg (1988), recognizing the affinities between Pachytrophe dimepate Bureau and Ampalis mauritiana, included both in Streblus section Ampalis, noting that in light of the diversity found in Streblus s.l., separation at the sectional level was no longer warranted; although our phylogenetic results support this grouping, we provisionally retain them as separate genera because the differences between them in phyllotaxy, stipule amplexicaulity, and embryo characters equal or exceed the distinctions between other genera in Moreae following our atomization of Streblus s.l. Further intensive study of that species complex may of course warrant a different approach, including potentially reducing both to a section of Paratrophis. Our species lists within the former components of Streblus follow Berg’s approach (1988, 2006), with two exceptions. We provisionally recognize Taxotrophis zeylanica (Thwaites) Thwaites as distinct from Taxotrophis taxoides (B. Heyne ex Roth) W.L. Chew ex E.M. Gardner (=Streblus taxoides (B. Heyne ex Roth) Kurz) based on our phylogenetic work and following the Flora of China, which recognizes Streblus zeylanicus as distinct from S. taxoides based on its clustered pistillate inflorescences. In addition, we provisionally recognize Paratrophis australiana (=S. glaber subsp. australianus) as distinct from Paratrophis glabra based on its geographic and consistent morphological distinctiveness. Taxotrophis and Paratrophis warrant further investigation to refine species limits.
Morus — Morus has never been revised, and the species concepts are often based on minor morphological differences (Berg et al., 2006; Berg 2001), and the paraphyly of several species in our analyses suggests that a broad M. alba similar to Bureau’s (1873) may be worth a second look. The monotypic subgenus Afromorus is sister to Milicia, but the leaf morphology is markedly different, instead resembling other Morus species in its trinerved based and crenate margins. We therefore raise Afromorus to genus level, requiring one new combination. Morus insignis, from western Central and South America, bears a remarkable resemblance to Paratrophis, in particular P. pendulina, especially in leaf morphology, which in M. insignis is not consistently trinerved as in other mulberries. The infructescence appears superficially like a mulberry because of its basally fleshy tepals, although with more loosely-packed flowers, but closer inspection places it firmly within Paratrophis, with drupes protruding from the persistent tepals, peltate bracts, and a sterile groove. Because analyses differ as to whether M. insignis is sister to Paratrophis or part of it, it seems best to treat them as congeners.
These changes result in eleven monophyletic genera of Moreae, providing a framework for revisionary work within the genera. Below, we present a complete genus and species list for Moreae with brief descriptions for all genera and synonymies for new combinations and reinstated or recircumscribed taxa. All cited protologues were reviewed, and dates for works published piecemeal were confirmed by reference to Taxonomic Literature 2 online (https://www.sil.si.edu/DigitalCollections/tl-2/index.cfm).
Taxonomic Treatment
Key to the tribes of Moraceae
This key follows the tribal circumscription of Clement & Weiblen (2009), as subsequently modified by Zerega & al. (2010), Chung & al. (2017), Zerega & Gardner (2019), and this study.
1. Inflorescences urceolate with the opening entirely closed by ostiolar bracts (i.e. syconium) such that the flowers are enclosed at all stages of development; lamina with waxy glandular spots at the base of the midrib or in the axils of the basal lateral veins beneath – Ficeae (Ficus)
1. Inflorescences capitate, spicate, discoid, or urceolate, but flowers are not entirely enclosed at all developmental stages (i.e. mature anthers or stigmas are open to the surrounding environment); lamina without waxy glandular spots. – 2
2. Inflorescences (at least staminate) with an involucre of imbricate bracts; often with self-pruning horizontal branches (except Olmedia, Poulsenia, Streblus) – Olmedieae (Antiaris, Antiaropsis, Castilla, Helicostylis, Maquira, Mesogyne, Naucleopsis, Olmedia, Perebea, Poulsenia, Pseudolmedia, Sparattosyce, Streblus)
2. Inflorescences not involucrate; plants without self-pruning branches – 3
3. Plants woody; dioecious; pistillate inflorescences globose-capitate; spines axillary or terminating short shoots; dioecious – Maclureae (Maclura)
3. Plants woody, herbaceous, or succulent; dioecious or monoecious; pistillate inflorescences various; spines absent or if present, then pistillate inflorescences not globose capitate – 4
4. Trees or shrubs; monoecious; inflorescences unisexual; staminate flowers with one stamen (rarely two);– Artocarpeae (Artocarpus, Batocarpus, Clarisia)
4. Trees, shrubs, lianas, herbaceous, or succulent; monoecious or dioecious; inflorescences unisexual or bisexual; staminate flowers with more than one stamen (or if one stamen then dioecious) – 5
5. Trees or shrubs; monoecious; inflorescences unisexual; stamens straight in bud or staminate flowers 5-parted and inflexed in bud;– Parartocarpeae (Hullettia, Parartocarpus, Pseudostreblus)
5. Trees, shrubs, lianas, herbaceous, or succulent; not as above but otherwise monoecious or dioecious; bisexual or unisexual inflorescences, stamens straight or inflexed in bud but staminate flowers never 5-parted – 6
6. Trees, shrubs, lianas, herbaceous, or succulent; bisexual inflorescences (or if unisexual then a climber or herbaceous) – Dorstenieae (Allaeanthus in part, Bleekrodea, Bosqueiopsis, Brosimum, Broussonetia in part, Dorstenia, Fatoua, Helianthostylis, Malaisia, Scyphosyce, Sloetia, Sloetiopsis, Treculia, Trilepsium, Trymatococcus, Utsetela)
6. Trees or shrubs; unisexual inflorescences – 7
7. Trees; stamens inflexed in bud; pistillate inflorescences globose-capitate – Dorstenieae (Allaeanthus in part, Broussonetia in part)
7. Trees or shrubs; stamens inflexed or straight in bud; if stamens inflexed in bud, pistillate inflorescence not globose-capitate – Moreae (Afromorus, Ampalis, Bagassa, Maillardia, Milicia, Morus, Pachytrophe, Paratrophis, Taxotrophis, Sorocea, Trophis)
Revisions
Tribe MOREAE
Moreae Gaudich. in Freyc., Voy. Uranie, Bot. (1830) 509.
Moreae subtribe Soroceae Miq in Martius, Fl. Bras. 4, 1, fasc. 12: 111 (1853), p.t. –
Soroceae (Miq.) C.C. Berg, Blumea 50: 537 (2005), p.p.
Strebleae Bureau in DC., Prodr. 17: 215 (1873), p.p.
Tree or shrubs, monoecious or dioecious. Leaves alternate or opposite, distichous or spirally arranged; stipules lateral to amplexicaul. Inflorescences unisexual, uniflorous or racemose, spicate, capitate, or globose; bracteate; tepals 4, free to connate; staminate flowers with 4 stamens, filaments straight or inflexed in bud, pistillode usually present; pistillate flowers with (mostly) free ovary, 2 stigmas. Fruits drupaceous or achene-like with a fleshy persistent perianth, dehiscent or not. Seeds with or without endosperm, testa usually with a thick vascularized part below the hilum, cotyledons equal or unequal, straight or folded.
Genera and distribution: eleven genera and 63 species with a worldwide distribution.
Afromorus
Afromorus (A. Chev. ex J.-F. Leroy) E.M. Gardner, stat. nov. — based on Morus L. subg. Afromorus [A. Chev., Rev. Bot. Appl. Agr. Trop. 29 (315–316): 70 (1949), invalidly published]; J.-F. Leroy, Rev. Bot. Appl. Agr. Trop. 29 (323–324): 482 (1949) & Bull. Mus. Hist. Nat. Paris, ser. 2, 21: 732 (1949). TYPE: Afromorus mesozygia (Stapf ex A. Chev.) E.M. Gardner.
Dioecious trees, shoot apices deciduous. Leaves distichous, triplinerved or at least trinerved at the base. Stipules free, more or less lateral. Inflorescences solitary or paired, bracts of varying shapes. Staminate inflorescences spicate, to 2.5mm long, flower 4-parted, tepals imbricate, ciliolate, pistillode small, apiculate. Pistillate inflorescences subglobose, ca. 5 mm across, flowers 4-parted, tepals ciliolate, stigma bifid, equal or unequal, arms filiform to 5 mm long. Infructescences subglobose or less often slightly elongate, ca. 1 mm across, tepals fleshy, yellowish to greenish, drupes ca. 5 x 3–5 mm. Seeds ca. 4.5 x 2.5–4.5 mm.
Species and distribution: one species, in tropical Africa.
Note: The leaves of Afromorus—with crenate margins and trinerved bases—bear a striking resemblance to those of Morus, and it is thus not surprising that the former was heretofore included within the latter. However, leaves of Afromorus are distinct in being usually completely triplinerved, without an upper pinnately-veined portion as in most species of Morus. As the genus is monotypic, the type must be its only species, Afromorus mesozygia (Stapf ex A. Chev.) E.M. Gardner.
Leroy’s protologue of Morus subg. Afromorus was published only in French and was therefore not a valid diagnosis or description in 1949. However, because the subgenus was monotypic, the protologue referred to a single name based on a validly-published description, Morus lactea (Sim.) Mildbr. (=Celtis lactea Sim.), and was therefore validly published by reference to the latter.
1. Afromorus mesozygia (Stapf ex A. Chev.) E.M. Gardner, comb. nov. — based on Morus mesozygia Stapf ex A. Chev. [Végétaux utiles de l’Afrique tropicale française 5: 263 (1909), nomen solum], J. Bot. (Morot) ser. 2, t. 2: 99 (1909).
Celtis lactea Sim, For. Fl. Port. E. Afr.: 97, t. 96 (1909), probably post-dating Morus mesozygia (fide Berg. 1977) – Morus lactea (Sim) Mildbr., Notizbl. Bot. Gart. Berlin 8: 243 (1922) – Morus mesozygia var. lactea (Sim) A. Chev., Rev. Bot. Appl. Agr. Trop. 29: 72 (1949).
Morus mesozygia var. sanda A. Chev., Rev. Bot. Appl. Agr. Trop. 29: 71 (1949), invalidly published (under Art. 39.1, Latin description or diagnosis lacking).
Morus mesozygia var. colossea A. Chev., Rev. Bot. Appl. Agr. Trop. 29: 71 (1949), invalidly published (under Art. 39.1, Latin description or diagnosis lacking).
Ampalis
Ampalis Bojer, Hort. Maurit. 291 (1837)
Streblus Lour. subgen. Parastreblus Blume, Mus. Bot. Ludg.-Bat. 2: 89 (1856) Streblus Lous. sect. Ampalis (Bojer) C.C. Berg, Proc. Kon. Ned. Akad. Wetensch. C 91(4): 358 (1988), p.p.
Dioecious trees or shrubs. Leaves spirally arranged to nearly distichous, pinnately veined. Stipules free, nearly lateral. Inflorescences solitary or paired in the leaf axils, spicate, with flowers in longitudinal rows, bracts present but sparse, usually subpeltate. Staminate inflorescences to 9 cm long, flowers 4-parted, decussate-imbricate, stamens 4, inflexed in bud, pistillode present. Pistillate inflorescences to 12 cm long, flowers tepals 4, separate, decussate-imbricate, ovary free, stigmas 2, equal. Infructescences with enlarged fleshy perianths, surrounding drupaceous fruits, the latter ca. 5–6 mm long. Seeds ca. 4 x 4 mm, testa thickened and not distinctly vascularized, cotyledons equal.
Species and distribution: One species, native to Madagascar and Comoros.
1. Ampalis mauritiana (Jacq.) Urb., Symb. Antill. 8: 165 (1920) – Morus mauritiana Jacq., Collect. 3: 206 (“1789”, 1791) – Streblus mauritianus (Jacq.) Blume, Mus. Bot. Lugd.-Bat. 2: 80 (1856).
Streblus maritimus Palaky, Catal. Pl. Madag. 2: 31 (1907)
Morus nitida Willem. In Useri, Ann. Bot. 18: 56 (1796)
M. ampalis Poir. In Lam., Encycl. Bot. 4: 380 (1797)
Trophis cylindrica Roxb., Fl. Indica 3: 599 (1832), pro syn. M. mauritianae. Ampalis madagascariensis, Bojer, Hort. Maurit. 291 (1837) nom. illeg., pro syn. Ampalis madagascariensis var. occidentalis Léandri, Mém. Inst. Sci. Madag., ser. B, 1: 12, pl. (1948).
Morus rigida Hassk., Cat. Hort. Bog. 74 (1844).
Note: May be considered congeneric with Pachytrophe, with which it shares inflorescence and infructescence morphology. They are distinguished on the basis of phyllotaxy, which is (usually) spiral in Ampalis and distichous in Pachytrophe.
Bagassa
Bagassa Aubl., Pl. Guiane. Suppl. 15 (1775).
Dioecious trees. Leaves opposite and decussate, lamina triplinerved, 3-lobed to entire. Stipules free, lateral. Inflorescences solitary or paired in the leaf axils, bracteate. Staminate inflorescences spicate with an abaxial sterile groove, to 12 cm long, flowers in longitudinal rows with, tepals 4, stamens 2, straight in bud, pistillode present. Pistillate inflorescences globose-capitate, ca. 1–1.5 cm across, flowers 4-lobed to 4-parted, stigmas 2, filiform.
Infructescences globose, ca. 2.5–3.5 cm across, green, tepals fleshy, yellowish to greenish, drupes ca. 7–8 mm long. Seeds ca. 3 x 2 mm, testa thin and not vascularized, cotyledons equal.
Species and distribution: One, in tropical South America.
1. Bagassa guianensis Aubl., Pl. Guiane. Supple. 15 (1775)
Maillardia
Maillardia Frapp. ex Duch. in Maillard, Notes sur l’île de la Réunion, Annex P. 3 (1862)
Trophis P. Br. sect. Maillardia (Duch.) Corner, Gard. Bull. Singapore 19: 230 (1962).
Dioecious trees or shrubs. Leaves distichous, pinnately veined. Stipules free, semi-amplexicaul. Inflorescences solitary or paired in the leaf axils, bracts subpeltate. Staminate inflorescences spicate with an abaxial sterile groove, flowers 4-parted, decussate-imbricate, stamens 4, inflexed in bud, pistillode present. Pistillate inflorescences up to three together, uni- or bi-florous, perianth tubular 4-lobed, ovary adnate to the perianth, stigmas 2, equal. Infructescences with enlarged fleshy perianths surrounding drupaceous fruits, the latter to 18 mm long. Seeds to 13 mm long, testa thin, with a thickened vascularized part below the hilum, cotyledons unequal.
Species and distribution: Two species, in Madagascar, Comoros, and the Seychelles.
1. Maillardia borbonica Duch., Ann. Notes Réunion, Bot. 1: 148 (1862) – Trophis borbonica (Duch.) C.C.Berg, Proc. Kon. Ned. Akad. Wetensch. C 91(4): 355 (1988) Maillardia lancifolia Frapp. ex Duch., Ann. Notes Réunion, Bot. 1: 148 (1862). Trophis borbonica (Duch.) C.C. Berg, Proc. Kon. Ned. Akad. Wetensch. C 91: 355 (1988).
2. Maillardia montana Leandri, Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 1: 25 (1948) – Trophis montana (Leandri) C.C.Berg, Proc. Kon. Ned. Akad. Wetensch. C 91(4): 355 (1988).
Maillardia occidentalis Leandri, Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 1: 26 (1948)
Maillardia orientalis Léandri, Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 1: 27 (1948)
Maillardia mandrarensis Léandri, Mém. Inst. Sci. Madagascar, Sér. B, Biol. Vég. 1: 28 (1948).
Maillardia pendula Fosberg, Kew Bull. 29: 266, t. 2 (1974).
Milicia
Milicia Sim, For. Fl. Port. E. Afr. 97 (1909).
Dioecious trees. Leaves distichous, lamina pinnately veined. Stipules free, not fully amplexicaul. Inflorescences spicate, usually solitary in the leaf axils or on leafless nodes, spicate, flowers in longitudinal rows alternating with rows of bracts, bracts mostly basally attached, abaxial sterile groove present. Staminate inflorescences to 20 cm long, flowers 4-parted, imbricate, stamens 4, inflexed in bud, pistillode present. Pistillate inflorescences to 4.5 cm long, flowers 4-parted, decussate-imbricate, ovary free, stigmas 2, unequal. Infructescences with enlarged fleshy perianths, surrounding slightly flattened drupaceous fruits, the latter to ca. 3 mm long. Seeds ca. 2 mm long, testa thin with a slightly thickened vascularized part below the hilum, cotyledons equal.
Species and distribution: Two species in tropical Africa.
1. Milicia excelsa (Welw.) C.C. Berg, Bull. Jard. Bot. Natl. Belg. 52(1-2): 227 (1982) – Morus excelsa Welw., Trans. Linn. Soc. Lond. (Bot.) 27: 69, t. 23 (1869) – Maclura excelsa (Welw.) Bur. In DC., Prodr. 17: 231 (1873) – Chlorophora excelsa (Welw.) Benth. & Hook., Gen. Pl. 3(1): 363 (1880).
Chlorophora tenuifolia Engl., Bot. Jahrb. 20: 139 (1894).
Milicia africana Sim, For. Fl. Port. E. Afr. 97, t. 122 (1909).
Chlorophora alba A. Chev., Bull. Soc. Bot. Fr. 58 (Mém. 8d): 209 (1912).
2. Milicia regia (A. Chev.) C.C. Berg, Bull. Jard. Bot. Natl. Belg. 52(1-2): 227 (1982) – Chlorophora regia A. Chev., Bull. Soc. Bot. Fr. 58 (Mém. 8d): 209 (1912) – Maclura regia (A. Chev.) Corner, Gard. Bull. Singapore 19: 237 (1962).
Morus
Morus L., Sp. Pl. 986 (1753).
Dioecious trees or shrubs. Leaves trinerved (to five-nerved) at the base. Stipules free, nearly lateral. Inflorescences solitary or paired in the leaf axils, without an obvious sterile groove, interfloral bracts absent. Staminate inflorescences spicate or racemose, to 8 cm long, flowers 4-parted, imbricate, stamens 4, inflexed in bud, pistillode present. Pistillate inflorescences subcapitate to spicate, up to 16 cm long, flowers tepals 4, separate, decussate-imbricate, ovary free, stigmas 2, equal. Infructescences with enlarged fleshy perianths enclosing achene-like fruits, the latter ca. 1 mm long. Seeds less than 1 mm long, cotyledons equal.
Species and distribution: Sixteen species, Asia and North to Central America; introduced worldwide.
1. Morus alba L., Sp. Pl. 2: 986 (1753).
var. alba
var. multicaulis (Perrottet) Loudon, Arbor. Frutic. Brit. 3: 1348 (1838).
1. 2. Morus australis Poir. in Desrousseaux et al., Encycl. 4: 380 (1797).
2. Morus boninensis Koidz., Bot. Mag. (Tokyo) 31: 38 (1917)
3. Morus cathayana Hemsl., J. Linn. Soc., Bot. 26: 456 (1894)
var. cathayana
var. gongshanensis (Z. Y. Cao) Z. Y. Cao, Acta Bot. Yunnan. 17: 154 (1995).
1. 5. Morus celtidifolia Kunth, Nov. Gen. Sp. [H.B.K.] 2: 33 (1817).
2. 6. Morus liboensis S. S. Chang, Acta Phytotax. Sin. 22: 66 (1984).
3. 7. Morus macroura Miq., Pl. Jungh. 1: 42 (1851).
var. macroura
var. laxiflora G.K.Upadhyay & A.A.Ansari, Rheedea 20: 44 (2010).
1. 8. Morus koordersiana J.-F.Leroy, Bull. Mus. Natl. Hist. Nat. sér. 2, 21: 729 (1949) (endemic to Sumatra and possibly synonymous with M. macroura. However, it was not cited by Berg et al. (2006) either as a good species or as a synonym of the latter).
2. 9. Morus microphylla Buckley, Acad. Nat. Sci. Philadelphia 1862: 8 (1863) (recognized in the Flora of North America (Flora of North America Editorial Committee, 1993) but likely conspecific with M. celtidifolia and considered so by Berg (2001)).
3. 10. Morus mongolica (Bureau) C. K. Schneid. in Sargent, Pl. Wilson. 3: 296 (1916).
4. 11. Morus nigra L. Sp. Pl. 2: 986 (1753).
5. 12. Morus notabilis C. K. Schneid. in Sargent, Pl. Wilson. 3: 293 (1916).
6. 13. Morus rubra L., Sp. Pl. 986 (1753).
var. rubra
var. murrayana (Saar & Galla) Saar, Phytologia 91: 106 (2009).
1. 14. Morus serrata Roxb., Fl. Ind., ed. 1832, 3: 596 (1832).
2. 15. Morus trilobata (S. S. Chang) Z. Y. Cao, Acta Phytotax. Sin. 29: 265 (1991).
3. 16. Morus wittiorum Hand.-Mazz., Anz. Akad. Wiss. Wien, Math.-Naturwiss. Kl. 58: 88 (1921)
Pachytrophe
Pachytrophe Bureau, Prodr. [A.P. de Candolle] 17: 234 (1873). – Ampalis Boj. sect.
Pachytrophe (Bur.) Baillon, Hist. Pl. 6: 191 (1875–76)
Streblus Lour. sect. Ampalis (Boj.) C.C. Berg, Proc. Kon. Ned. Akad. Wetensch. C 91(4): 358 (1988), p.p. Dioecious trees. Leaves distichous, pinnately veined. Stipules connate, semi-amplexicaul. Inflorescences usually paired in the leaf axils, spicate, with an abaxial sterile groove, flowers sessile in longitudinal rows, tepals 4, decussate-imbricate, ciliolate. Staminate inflorescences up to 5 cm. long, bracts basally attached, flowers with filaments inflexed in bud, anthers introrse, pistillode quadrangular, ca. 0.5 mm long. Pistillate inflorescences up to 2.5 cm long, bracts (sub)peltate, flowers 2–14, tepals free, stigma bifid, arms equal.
Infructescences reddish, with enlarged fleshy perianths ca. 6–8 mm long, drupes thinly fleshy. Seeds ca. 4.5 mm long.
Species and distribution: One, in tropical Africa (Madagascar).
Note: Pachytrophe could be considered congeneric with Ampalis, within which Baillon treated it as a section (although without making any new specific combinations).
1. Pachytrophe dimepate Bureau, Prodr. [A.P. de Candolle] 17: 234 (1873).
Pachytrophe obovata Bureau, Prodr. [A.P. de Candolle] 17: 235 (1873).
Plecospermum bureaui A.G. Richt., Term. Füzetek 18: 296 (1895), nomen nudum et superfl., pro. syn. nom. ined. “Plecospermum obovatum Bureau” (Boivin 1717, P).
Plecospermum (?) laurifolium Baill. in Grandidier, Hist. Madag. Vol. 35, Hist. Nat.
Pl., Tome 5, Atlas 3, 2e partie: pl. 294a (1895) – Pachytrophe obovata var.
laurifolia (Baill.) Léandri, Mém. Inst. Sci. Madag., ser. B, 1: 16 (1948).
Pachytrophe obovata var. montana Léandri, Mém. Inst. Sci. Madag., ser. B, 1: 16 (1948).
Paratrophis
Paratrophis Blume, Ann. Mus. Bot. Lugduno-Batavi 2 (1856) 81.
Pseudomorus Bureau, Ann. Sci. Nat., Bot. sér. 5, 11: 371 (1869).
Uromorus Bureau in A.DC., Prodr. 17: 236 (1873).
Calpidochlamys Diels, Bot. Jahrb. Syst. 67: 172 (1935); — Trophis sect.
Calpidochlamys Corner, Gard. Bull. Singapore 19: 230 (1962).
Chevalierodendron J.-F. Leroy, Compt. Rend. Hebd. Séances Acad. Sci. 227: 146 (1948).
Streblus Lour. sect. Paratrophis (Blume) Corner, Gard. Bull. Singapore 19: 216 (1962).
Streblus Lour. sect. Protostreblus Corner, Blumea 18: 393 (1970).
Morus L. subg. Gomphomorus J.-F. Leroy, Bull. Mus. Hist. Nat. (Paris), Sér. 2, 21: 732 (1949)
Dioecious trees or shrubs. Leaves distichous (or spiral in P. ascendens), pinnately veined, sometimes trinerved at the base. Stipules free, lateral. Inflorescences axillary, solitary or up to 5 together, spicate, with an abaxial sterile groove, interfloral bracts mostly peltate, flowers sessile in longitudinal rows, tepals 4, valvate, ciliolate. Staminate inflorescences up to at least 20 cm. long, flowers with filaments inflexed in bud, pistillode present. Pistillate inflorescences up to at least 10 cm long, flowers usually at least 2 (to many), tepals free (except in P. philippinensis), stigma bifid, arms equal. Fruits drupaceous, red to black, with tepals persistent but usually not enlarged or fleshy (except in P. philippinensis and P. insignis), up to ca. 1 cm long. Seeds up to ca. 8 x 6 mm, cotyledons equal.
Species and distribution: Twelve species from the Malesian region to Australia, New Zealand, and Oceania, Central and western South America.
1. Paratrophis anthropophagorum (Seem.) Benth. & Hook. f. ex Drake, Ill. Fl. Ins. Pacif. 296 1892
Trophis anthropophagorum Seem. [Bonplandia 9 (17–18): 259 (1861), nomen nudum] Fl. Vit. 258, t. 68 (1868); — Uromorus anthropophagorum (Seem.) Bureau in de Candolle, Prodr. 17 (1873) 236; — Streblus anthropophagorum (Seem.) Corner, Gard. Bull. Sing.19 (1962) 221.
Caturus oblongatus Seem., Fl. Vit. 254 (1868).
Paratrophis ostermeyri Rech., Fedd. Rep. 5: 130 (1908).
Paratrophis viridissima Rech., Fedd. Rep. 5: 130 (1908).
Paratrophis zahlbruckneri Rech., Fedd. Rep. 5: 130 (1908).
Pseudomorus brunoniana (Endl.) Bureau var. tahitensis J. Nadeaud, Énum. Pl. Tahiti 43 (1873) – Uromorus tahitensis (J. Nadeaud) Bureau in de Candolle, Prodr. 17: 237 (1873) – Paratrophis tahitensis (Bureau) Benth. & Hook.f. ex Drake, Ill. Ins. Mar.
Pacif. Fasc. 7: 296 (1892); et Fl. Polyn. Franc. 193 (1892). – Streblus tahitensis (J. Nadeaud) E.J.H. Corner, Gard. Bull. Singapore 19: 225 (1962).
1. 2. Paratrophis ascendens (Corner) E.M. Gardner comb. nov. — based on Streblus ascendens Corner, Blumea 18: 395, t.1 (1970).
2. Paratrophis australiana C.T. White, Contr. Arnold Arbor. 4: 15 (1933) — Streblus glaber (Merr.) Corner var. australianus (C.T. White) Corner, Gard. Bull. Singapore 19: 221 (1962) — Streblus glaber (Merr.) Corner subsp. australianus (C.T. White) C.C. Berg, Blumea 50: 548 (2005).
3. Paratrophis banksii Cheeseman, Man. N.Z. Fl. 633 (1906) – Streblus banksii
(Cheeseman) C.J. Webb, in Connor & Edgar, New Zealand J. of Bot. 25 136 (1987).
Paratrophis heterophylla var. elliptica Kirk T.N.Z.L 29: 500, t. 46 (1897); — Streblus heterophyllus var. ellipticus (Kirk) Corner, Gard. Bull. Singapore 19: 222 (1962).
Note:— Morphologically very close to and not always distinguishable from P. microphylla.
1. 5. Paratrophis glabra (Merr.) Steenis, J. Bot. 72: 8 (1934) – Gironniera glabra Merr., Philipp. J. Sci., 1, Suppl. 42 (1906); Enum. Philipp. Flow. Pl. 2: 35 (1923) — Chevalierodendron glabrum (Merr.) J.-F. Leroy, Compt. Rend. Hebd. Séances Acad. Sci. 227: 146 (1948) — Streblus glaber (Merr.) Corner, Gard. Bull. Singapore 19: 221 (1962). Aphananthe negrosensis Elmer, Leafl. Philipp. Bot. 2: 575 (1909).
Pseudostreblus caudatus Ridl., J. Fed. Malay States Mus. 6: 54 (1915).
Streblus laevifolius Diels, Bot. Jahrb. Syst. 67: 171 (1935).
Streblus urophyllus Diels, Bot. Jahrb. Syst. 67: 172 (1935) — Streblus glaber (Merr.) Corner subsp. urophyllus (Diels) C.C. Berg, Blumea 50: 548 (2005);
Streblus urophyllus Diels var. salicifolius Corner, Gard. Bull. Singapore. 19: 225 (1962).
Note: Berg et al. (2006) treated P. australiana and Streblus urophyllus as subspecies of Streblus glaber. Paratrophis australiana, endemic to Australia, has crenate leaf margins and somewhat larger staminate inflorescences with more interfloral bracts than P. glabra; these morphological differences are consistent and geographically confined, and we therefore reinstate P. australiana. This stands in contrast to Streblus urophyllus, which we provisionally maintain in synonomy under P. glabra. Although collections from Mt.
Wilhelm in New Guinea are remarkable for their thick coriaceous leaves and spinose margins, similar toothed margins can be found at higher elevations in Borneo and Sulawesi, suggesting that the striking leaf morphology of S. urophyllus is an alpine effect not indicative of speciation. We reserve judgment on the status of Streblus urophyllus var. salicifolius Corner, applied by Corner (1962) to specimens with linear leaves, provisionally maintaining it in synonomy following Berg et al. (2006).
1. 6. Paratrophis insignis (Bureau) E.M. Gardner comb. nov. — based on Morus insignis
Bureau, in De Candolle, Prodr. 17: 247 (1873).
Morus peruviana Planch. ex Koidz., Fl. Symb. Orient.-Asiat. 88 (1930).
Morus trianae J.-F.Leroy, Bull. Mus. Hist. Nat. (Paris) Sér. 2, 21: 731 (1949).
Morus marmolii Legname, Lilloa 33: 334 (1973)
Note: The infructescences differ from most other Paratrophis in their fleshy free tepals and denser aggregation of flowers, but the staminate inflorescences are typical of the genus.
1. 7. Paratrophis microphylla (Raoul) Cockayne, Bot. Notes Kennedy’s Bush & Scenic Res. Port Hills, Lyttelton (Rep. Scenery Preserv.) 3 (1915) – Epicarpurus microphyllus Raoul, Choix Pl. Nouv.-Zel. 14. (1846) – Taxotrophis microphylla (Raoul) F. Muell., Fragm. (Mueller) 6(47): 193 (1868).
Paratrophis heterophylla Blume, Mus. Bot. 2(1-8): 81. (1856).
1. 8. Paratrophis pendulina (Endl.) E.M. Gardner, comb. nov. — based on Morus pendulina Endl., Prodromus Florae Norfolkicae 40 (1833) – Pseudomorus pendulina (Endl.) Stearn, J. Arnold Arbor. 28: 427 (1947) – Pseudomorus brunoniana var. pendulina (Endl.) Bureau, Ann. Sci. Nat. Bot. sér. 5, 11: 372 (1869) – Streblus pendulinus (Endl.) F. Muell., Fragmenta Phytographiae Australiae (1868).
Morus brunoniana Endl., Atakta Bot. t. 32 (1835) – Streblus brunonianus (Endl.) F. Muell., Fragm. Phyt. Australiae 6: 192 (1868) – Pseudomorus brunoniana (Endl.) Bureau, Ann. Sci. Nat., Bot. sér. 5, 11: 372 (1869)
Pseudomorus brunoniana var. australiana Bureau, Ann. Sci. Nat., Bot. sér. 5, 11: 373 (1869) – Pseudomorus pendulina var. australiana (Bureau) Stearn, J. Arnold Arbor. 28: 427 (1947).
Pseudomorus brunoniana var. australiana subvar. castaneaefolia Bureau, Ann. Sci. Nat., Bot. sér. 5, 11: 372 (1869).
Pseudomorus brunoniana var. obtusa Bureau, Ann. Sci. Nat., Bot. sér. 5, 11: 373 (1869) – Pseudomorus pendulina var. obtusa (Bureau) Stearn, J. Arnold Arbor. 28: 428 (1947), as var. obtusata.
Pseudomorus sandwicensis O. Deg., Fl. Hawaiiensis fam. 38 (1938) – Pseudomorus brunoniana var. sandwicensis (O. Deg.) Skottsb., Acta Horti Gothob. 15: 347 (1944) – Pseuomorus pendulina var. sandwicensis (O. Degener) Stearn, J. Arnold Arbor. 28: 428 (1947) – Streblus sandwicensis (O. Deg.) H. St. John, Pacific Trop. Bot. Gard. Mem. 1: 374. 1973.
Note: This complex requires further investigation. Although the types of Morus pendulina, Morus brunoniana, Pseudomorus sandwicensis, and Pseudomorus brunoniana var. obtusa differ from one another, it is difficult to discern clear geographic patterns that warrant maintaining separate species, and we therefore provisionally treat them all as one widespread and variable species.
1. 9. Paratrophis philippinensis (Bureau) Fern.-Vill., Nov. App. 98 (1880) — Uromorus philippinensis Bureau in A.DC., Prodr. 17: 237 (1873) — Trophis philippinensis (Bureau) Corner, Gard. Bull. Singapore 19: 231 (1962).
Sloetia minahassae Koord., Versl. Minahasa 645 (1898).
Paratrophis grandifolia Elmer, Leafl. Philipp. Bot. 5: 1814 (1913). Calpidochlamys branderhorstii Diels, Bot. Jahrb. Syst. 67: 173 (1935) — Trophis branderhorstii (Diels) Corner, Gard. Bull. Singapore 19: 231 (1962).
Calpidochlamys drupacea Diels, Bot. Jahrb. Syst. 67: 173 (1935) — Trophis drupacea (Diels) Corner, Gard. Bull. Singapore 19: 231 (1962).
Note: This species is the only Paratrophis with fused tepals and one of only two whose tepals develop into a fleshy accessory fruit. The staminate inflorescences, however, are typical of the genus.
1. 10. Paratrophis sclerophylla (Corner) E.M. Gardner, comb. nov. — based on Streblus sclerophyllus Corner, Blumea 18: 399 (1970).
2. 11. Paratrophis smithii Cheeseman T.N.Z.L 20:148 (1888) – Streblus smithii (Cheeseman) Corner Gard. Bull. Singapore 19: 224 (1962).
Note: Morphologically very close to P. anthropophagorum.
1. 12. Paratrophis solomonensis (Corner) E.M. Gardner, comb. nov. — based on Streblus solomonensis Corner, Gard. Bull. Singapore 19: 224 (1962).
Note:— Morphologically very close to P. anthropophagorum.
Sorocea
Sorocea A. St-Hil., Mém. Mus. Hist. Nat. 7: 473 (1821).
Balanostreblus Kurz, J. Asiat. Soc. Bengal., Pt. 2, Nat. Hist., 42: 247 (1873), p.p.
Pseudosorocea Baill., Hist. Pl. 6: 210 (1875).
Trophisomia Rojas Acosta, Bull. Acad. Inst. Geogr. Bot. 24: 211 (1914).
Paraclarisia Ducke, Arq. Serv. Florest. 1(1): 2 (1939).
Dioecious trees. Leaves alternate, distichous, lamina pinnately veined, 3-lobed to entire. Stipules free, lateral. Inflorescences solitary or paired in the leaf axils or below the leaves, racemose to spicate to subcapitate or uniflorous, bracts basally attached to peltate.
Staminate flowers 4-lobed to 4-parted, tepals decussate-imbricate, stamens (3–)4, straight in bud, pistillode usually absent. Pistillate flowers tubular, 4-lobed to sub-entire, ovary basally adnate to the perianth, stigmas 2, short, usually tongue-shaped. Infructescences with drupaceous fruits enclosed by fleshy enlarged perianths, the latter red to orange, turning black at maturity. Seeds large, testa thin, embryo green, cotyledons very unequal, the smaller one minute and enclosed by the larger.
Species and distribution: 19 species in Central and South America from Mexico to Argentina.
Note: This species list follows the Flora Neotropica monograph (Berg 2001) with the addition of four new species and one status change published later. Names synonymized by Berg (2001) but recognized by Burger et al. (1962) are noted in parentheses. Complete synonymies can be found in those treatments.
Sorocea subg. Sorocea
Sorocea affinis Hemsl., Biol. Cent.-Amer., Bot. 3: 150 (1883).
Sorocea angustifolia Al.Santos & Romaniuc, Novon 24: 199 (2015).
Sorocea bonplandii (Baill.) W.C. Burger, Lanj. & de Boer, Acta Bot. Neerl. 11: 465 (1962).
Sorocea briquetii J.F.Macbr., Publ. Field Columb. Mus., Bot. Ser. 11: 16 (1931) (incl. S. pileate W.C. Burger).
Sorocea carautana M.D.M.Vianna, Carrijo & Romaniuc, Novon 19: 549 (2009).
Sorocea ganevii R.M.Castro, Neodiversity 1: 18 (2006).
Sorocea guilleminiana Gaudich., Voy. Bonite, Bot. 3: t. 74 (1843) (incl. S. klotzschiana Baill. and S. macrogyna Lanj. & Wess. Boer).
Sorocea hilarii Gaudich., Voy. Bonite, Bot. 3: t. 71 (1843) (incl. S. racemosa Gaudich.).
Sorocea jaramilloi C.C.Berg, Novon 6: 241 (1996).
Sorocea longipedicellata A.F.P. Machado, M.D.M. Vianna & Romaniuc, Syst. Bot. 38: 687 (2013).
Sorocea muriculata Miq., C.F.P. von Martius & auct. suc. (eds.), Fl. Bras. 4(1): 113 (1853).
subsp. muriculata (incl. S. amazonica Miq.)
subsp. uaupensis (Baill.) C.C. Berg, Proc. Kon. Ned. Akad. Wetensch. C 88: 387 (1985) (incl. S. guayanensis W.C. Burger).
Sorocea pubivena Hemsl., Biol. Cent.-Amer., Bot. 3: 150 (1883).
subsp. pubivena (incl. S. cufodontii W.C. Burger)
subsp. hirtella (Mildbr.) C.C. Berg, Novon 6: 243 (1996) (incl. S. opima J.F. Macbr.)
subsp. oligotricha (Akkermans & C.C. Berg) C.C. Berg, Novon 6: 243 (1996).
Sorocea ruminata C.C.Berg, Novon 6: 244 (1996).
Sorocea sarcocarpa Lanj. & Wess. Boer, Acta Bot. Neerl. 11: 452 (1962).
Sorocea steinbachii C.C. Berg, Proc. Kon. Ned. Akad. Wetensch. C 88: 385 (1985).
Sorocea trophoides W.C. Burger, Acta Bot. Neerl. 11: 450 (1962).
Sorocea subg. Paraclarisia (Ducke) W.C. Burger, Lanj. & Wess. Boer, Acta Bot. Neerl. 11: 468 (1962).
Sorocea duckei W.C.Burger, Acta Bot. Neerl. 11: 473 (1962).
Sorocea sprucei (Baill.) J.F.Macbr., Publ. Field Mus. Nat. Hist., Bot. Ser. 11: 16 (1931).
subsp. sprucei (incl. S. arnoldoi Lanj. & Wess. Boer).
subsp. saxicola (Hassl.) C.C.Berg, Proc. Kon. Ned. Akad. Wetensch., Ser. C., Biol. Med. Sci. 88: 391 (1985).
Sorocea subumbellata (C.C. Berg) Cornejo, Novon 19: 297 (2009).
Taxotrophis
Taxotrophis Blume, Ann. Mus. Bot. Lugduno-Batavi 2 77 (1856); Hutch., Bull. Misc. Inform. Kew 147 (1918).
Streblus Lour. sect. Taxotrophis (Blume) Corner, Gard. Bull. Singapore 19: 218 (1962); Berg et al., Fl. Males. Ser. 1, Vol. 17, Pt. 2 (2006).
Monoecious or dioecious trees or shrubs, usually with lateral or terminal thorns. Leaves distichous, pinnately veined, petioles adaxially pubescent. Stipules free, lateral. Inflorescences axillary, solitary or paired, with an abaxial sterile groove, interfloral bracts basally attached, flowers with 4 free tepals, valvate. Staminate inflorescences spicate to sub-capitate, flowers with filaments inflexed in bud, pistillode present. Pistillate inflorescences uniflorous or sub-spicate, flowers usually pedicellate, stigma bifid, arms equal. Fruits drupaceous, up to ca. 1 cm long, usually loosely enclosed by the enlarged persistent tepals. Seeds up to ca. 8 x 6 mm, cotyledons subequal to unequal.
Species and distribution: Six species, ranging from Sri Lanka to New Guinea.
1. Taxotrophis ilicifolia S. Vidal, Revis. Pl. Vasc. Filip. 249 (1886).
Balanostreblus ilicifolia Kurz, J. Asiat. Soc. Bengal, Pt. 2, Nat. Hist. 42(4): 248 (1874), p.p.
Pseudotrophis laxiflora Warb., Bot. Jahrb. Syst. 13 294 (1891).
Taxotrophis obtusa Elmer, Leafl. Philipp. Bot. 5 1813 (1913).
Taxotrophis laxiflora Hutch., Bull. Misc. Inform. Kew 151 (1918). – Streblus laxiflorus (Hutch.) Corner, Gard. Bull. Singapore 19 229 (1962).
Taxotrophis triapiculata Gamble, Bull. Misc. Inform. Kew 188 (1913).
Taxotrophis eberhardtii Gagnep., Fl. Indo-Chine 5 700 (1928).
Taxotrophis macrophylla auct. non Boerl.: Burkill, Dict. Econ. Prod. Malay Penins. 2126 (1935).
Streblus ilicifolius (S. Vidal) Corner, Gard. Bull. Singapore 19: 227 (1962).
1. 2. Taxotrophis macrophylla (Blume) Boerl., Handl. Fl. Ned. Ind. 3: 359 (1900).
Streblus macrophyllus Blume, Ann. Mus. Bot. Lugduno-Batavi 2: 80 (1856) —
Diplocosmacrophyllus (Blume) Bureau in A.DC., Prodr. 17: 216 (1873).
Pseudotrophis mindanaensis Warb. in Perkins, Fragm. Fl. Philipp. 1: 165 (1905); Elmer, Leafl. Philipp. Bot. 5: 1815 (1913), ‘Taxatrophis mindanaensis’ in nota.
Paratrophis caudata Merr., Philipp. J. Sci., 1, Suppl. 183 (1906).
Taxotrophis balansae Hutch., Bull. Misc. Inform. Kew 151 (1918).
Dimerocarpus brenieri Gagnep., Bull. Mus. Hist. Nat. (Paris) 27: 441 (1921).
1. 3. Taxotrophis perakensis (Corner) E.M. Gardner, comb. nov., based on Streblus perakensis Corner, Gard. Bull. Singapore 19: 223 (1962).
Note: Although Corner (1962) considered S. perakensis species part of section Paratrophis, albeit with some hesitation, Berg et al. (2006), whom we follow, placed it in Taxotrophis based on the spines that appear on some specimens. This is a rather variable species that requires further investigation to properly elucidate its limits and affinities.
1. 4. Taxotrophis spinosa Steenis in Backer & Bakh.f., Fl. Java 2 (1965) 16.
Urtica spinosa Blume, Bijdr. (1825) 507. — Streblus spinosus (Blume) Corner, Gard.
Bull. Singapore 19: 229 (1962).
Taxotrophis javanica Blume, Ann. Mus. Bot. Lugduno-Batavi 2: 77, t. 26 (1856).
1. 5. Taxotrophis taxoides (B. Heyne ex Roth) W.L. Chew ex E.M. Gardner, comb. nov.
Trophis taxoides B. Heyne ex Roth, Nov. Pl. Sp. 368 (1821). — Trophis taxiformis Spreng., Syst. Veg. 3 902 (1826), nom. nov. illeg. — Streblus taxoides (B. Heyne ex Roth) Kurz, Forest Fl. Burma 2: 465 (1877) — Phyllochlamys taxoides (B. Heyne ex Roth) Koord., Exkurs.-Fl. Java 2: 89 (1912).
Trophis spinosa Roxb., Fl. Ind., ed. Carey 3: 762 (1832), non Willd. 1806, nec Blume 1826. — Epicarpurus spinosus (Roxb.) Wight, Icon. Pl. Ind. Orient. 6: 7, t. 1962 (1853), p.p. — Phyllochlamys spinosa (Roxb.) Bureau in A.DC., Prodr. 17: 218 (1873);
Epicarpurus timorensis Decne., Nouv. Ann. Mus. Hist. Nat. 3: 499, t. 21 (1834).
Taxotrophis roxburghii Blume, Ann. Mus. Bot. Lugduno-Batavi 2: 78 (1856)
Streblus microphyllus Kurz, Prelim. Rep. Forest Pegu App. A, cxviii; App. B, 84 (1875); — Streblus taxoides (B. Heyne ex Roth) Kurz var. microphylla (Kurz) Kurz, Forest Fl. Burma 2: 465 (1877).
Phyllochlamys wallichii King ex Hook.f., Fl. Brit. India 5: 489 (1888);
Phyllochlamys taxoides (B. Heyne ex Roth) Koord. var. parvifolia Merr., Enum. Philipp. Flow. Pl. 2: 38 (1923).
Taxotrophis poilanei Gagnep., Fl. Indo-Chine 5: 701 (1928).
Taxotrophis crenata Gagnep., Fl. Indo-chine 5: 702, t. 82 (1928). — Streblus crenatus
(Gagnep.) Corner, Gard. Bull. Singapore 19: 226 (1962).
Phyllochlamys tridentata Gagnep., Fl. Indo-Chine 5: 714 (1928).
Note: In the 1950s, Dr. Chew Wee Lek annotated quite a lot of specimens at K and SING with the new combination Taxotrophis taxoides. However, the combination was never published, probably because the need for the combination was obviated in 1962 when Corner, his doctoral supervisor, reduced Taxotrophis to a section of Streblus.
1. 6. Taxotrophis zeylanica (Thwaites) Thwaites, Enum. Pl. Zeyl. [Thwaites] 264 (1861).
Epicarpurus zeylanicus Thwaites, Hooker’s J. Bot. Kew Gard. Misc. 4: 1 (1852) —
Diplocos zeylanica (Thwaites) Bureau, Prodr. [A. P. de Candolle] 17: 215 (1873)
— Streblus zeylanicus (Thwaites) Kurz, Forest Fl. Burma 2: 464 (1877).
Taxotrophis caudata Hutchinson, Bull. Misc. Inform. Kew 1918(4): 149 (1918).
Trophis
Trophis Lour., P. Browne, Civ. Nat. Hist. Jamaica 357 (1756), nom. cons.
Bucephalon L. Sp. Pl. 1190 (1753), nom. rejic.
Skutchia Pax & K. Hoffm. ex C.V. Morton, J. Wash. Acad. Sci. 27: 306 (1937).
Dioecious trees or shrubs. Leaves distichous, pinnately veined. Stipules free, lateral. Inflorescences axillary or just below the leaves, solitary or paired, interfloral bracts basally attached. Staminate inflorescences spicate to racemose with an abaxial sterile groove, tepals 4, basally connate, stamens 4, filaments inflexed in bud, pistillode present. Pistillate inflorescences spicate to racemose or subcapitate, tepals 4, connate, forming a tubular perianth, ovary adnate to the perianth or not, stigma bifid, arms equal. Fruits drupaceous, up to ca. 1.5 cm long, adnate to the perianth or not, the perianth enlarged and fleshy or not. Seeds up to ca. 1 cm long, cotyledons equal.
Species and distribution: Five species in the neotropics.
Trophis P. Browne sect. Trophis
1. Trophis cuspidata Lundell, Amer. Midl. Naturalist 19: 427 (1938).
2. Trophis mexicana (Liebm.) Bureau, Prodr. [A. P. de Candolle] 17: 253 (1873).
3. Trophis noraminervae Cuevas & Carvajal, Acta Bot. Mex. 47: 2 (1-7, fig.) (1999).
4. Trophis racemosa Urb., Symb. Antill. (Urban). 4(2): 195 (1905).
Trophis P. Browne sect. Echinocarpa C.C. Berg, Proc. Kon. Ned. Akad. Wetensch., Ser. C, Biol. Med. Sci. 91: 353 (1988).
5. Trophis involucrata W.C. Burger, Phytologia 26: 432 (1973).
Tribe OLMEDIEAE
Olmedieae Trécul, Ann. Sci. Nat., Bot. sér. 3, 8 (1847) 126
Castilleae C.C. Berg, Acta Bot. Neerl. 26 (1977) 78
Strebleae Bureau in DC., Prodr. 17: 215 (1873), p.p.
Tree or shrubs, monoecious or dioecious (or androdioecious), mostly with self-pruning branches. Leaves alternate or opposite, distichous or spirally arranged; stipules lateral to amplexicaul. Inflorescences mostly unisexual, capitate, mostly discoid to urceolate, involucrate, tepals mostly 4, connate or not. Staminate inflorescences usually many-flowered; stamens 4 or fewer, with filaments straight or less often inflexed in bud, pistillode mostly absent. Pistillate inflorescences one to many-flowered, ovary free or not, stigmas 2, filiform. Fruits mostly drupaceous, mostly enclosed by a fleshy perianth or embedded in a fleshy receptacle. Seeds with or without endosperm, testa thin, vascularized, cotyledons mostly equal.
Genera and distribution: 13 genera with 63 species. Eight neotropical genera: Castilla (3 spp.), Helicostylis (7), Maquira (4), Naucleopsis (22), Olmedia (1), Perebea (9), Poulsenia (1), and Pseudolmedia (9); and five Paleotropical genera: the widespread Antiaris (1); Antiaropsis (2) in New Guinea; Mesogyne (1) in Africa; Sparattosyce (1) in New Caledonia; and Streblus (3) in South to Southeast Asia.
Olmedia
Olmedia Ruiz & Pav., Syst. Veg. Fl. Peruv. Chil.1:257.1798. Type — Olmedia aspera Ruiz & Pav.
Trophis section Olmedia (Ruiz & Pav.) Berg, Proc. Kon. Ned. Acad. Wetensch. C. 91: 354. 1988.
Dioecious trees or shrubs. Leaves distichous, lamina pinnately veined. Stipules free, not fully amplexicaul. Inflorescences unisexual, with a well-developed involucre. Staminate inflorescences discoid, multiflorous; tepals 4, valvate, stamens 4, inflexed in bud, pistillode absent. Pistillate inflorescences usually uniflorous, perianth tubular, 4-dentate, ovary free, stigmas 2, equal. Fruits drupaceous, surrounded by fleshy persistent perianth and subtended by spreading, fleshy involucral bracts. Seeds ca. 5 mm long, cotyledons equal.
Species and distribution: One species, in the Neotropics.
1. Olmedia aspera Ruiz & Pav., Syst. Veg. Fl. Peruv. Chil.1:257.1798
Olmedia caucana Pittier, Contr. U.S. Natl. Herb. 13:434. 1912. — Trophis caucana
(Pitttier) C.C. Berg, Proc. Kon. Ned. Acad. Wetensch. C. 91: 354. 1988
Olmedia poeppigiana Klotzsch, Linnaea 20: 525. 1847, as a synonym of O. aspera Poeppig & Endlicher, Nov. Gen. 2: 31. 1838, based on Poeppig s.n. or 1267, non O.poeppigiana Martius Flora (or Bot. Zeit) 24 (Beibl. 2): 93. 1841 (= Helicostylis tomentosa (Poeppig & Endlicher) Rusby).
Olmedia falcifolia Pittier, Contr. U.S. Natl. Herb. 13: 435. 1912.
Trophis aurantiaca Herzog, Repert. Spec. Nov. 7: 51. 1909
Streblus
Streblus Lour., Fl. Cochinch. (1790) 615.
Achymus Juss., Dict. Sci. Nat. 1, Suppl. 31 (1816).
Epicarpurus Blume, Bijdr. 488 (1825).
Albrandia Gaudich. in Freyc., Voy. Uranie, Bot. 509 (1830).
Calius Blanco, Fl. Filip. 698 (1837).
Teonongia Stapf, Hooker’s Icon. Pl. 30: t. 2947 (1911).
Diplothorax Gagnep., Bull. Soc. Bot. France 75 98 (1928).
Trees or shrubs, dioecious or monoecious. Leaves distichous, lamina pinnately veined. Stipules free, lateral. Inflorescences bisexual or unisexual, capitate, with a rudimentary involucre, bracts basally attached. Staminate inflorescences discoid capitate, multiflorous; tepals 4, imbricate, stamens 4, inflexed in bud, pistillode present but small. Pistillate inflorescences usually uniflorous, tepals 4, ovary free, stigmas 2, equal. Fruits drupaceous, up to ca. 8 mm long, initially enclosed by enlarged but not fleshy tepals, which may open later. Seeds ca. 5 mm long, cotyledons equal or very unequal.
Species and distribution: three species from India to South China and from mainland Southeast Asia to the Philippines and the Moluccas.
1. Streblus asper (Retz.) Lour., Fl. Cochinch. 2: 615 (1790).
Trophis aspera Retz., Observ. Bot. 5 (1788) – Epicarpurus asper (Retz.) Steud. Nomencl. Bot. ed. 2, 1: 556 (1840).
Trophis cochinchinensis Poir., Encycl. 8 (1808) 123.
Epicarpurus orientalis Blume, Bijdr. (1825) 488
Calius lactescens Blanco, Fl. Filip. (1837) 698; ed. 3, 3 (1879) 1103, t. 171. —
Streblus lactescens (Blanco) Blume, Ann. Mus. Bot. Lugduno-Batavi 2 (1856) 80.
Achymus pallens Sol. ex Blume, Ann. Mus. Bot. Lugduno-Batavi 2 (1856) 79.
Cudrania crenata C.H. Wright, J. Linn. Soc., Bot. 26 (1899) 469. — Vanieria crenata
(C.H. Wright) Chun, J. Arnold Arbor. 8 (1927) 21.
Diplothorax tonkinensis Gagnep., Bull. Soc. Bot. France 75 (1928) 98.
Note: Retzius’s Trophis aspera predated Loureiro’s Streblus asper by two years, making the latter an implied new combination under Article 41.1 of the Code (cf. Ex. 10).
Synonymies of the S. asper have sometimes erroneously included Trophis aculeata Roth, which is actually Maclura spinosa (Roxb. ex Willd.) C.C. Berg.
1. 2. Streblus celebensis C.C. Berg, Blumea 50 547 (2005).
2. Streblus tonkinensis (Eberh. & Dubard) Corner, Gard. Bull. Singapore 19 (1962): 228. Bleeokrodea tonkinensis Eberh. & Dubard, Compt. Rend. Hebd. SÈances Acad. Sci. 145 (1907): 632. — Teonongia tonkinensis (Eberh. & Dubard) Stapf.
Tribe DORSTENIEAE
Dorstenieae Gaudich. in Freyc., Voy. Uranie, Bot. (1830).
Broussonetieae Gaudich. in Freyc., Voy. Uranie, Bot. 508 (1830).
Brosimeae Trécul, Ann. Sci. Nat., Bot. sér. 3, 8: 146 (1847).
Fatoueae Engl., Nat. Pflanzenfam. 3, 1: 71 (1888).
Tree, shrubs, lianas, and herbs, monoecious or less often dioecious. Leaves alternate or less commonly (sub)opposite, distichous or spirally arranged; stipules lateral to fully amplexicaul. Inflorescences unisexual or bisexual, cymose, spicate, globose, or discoid to turbinate or cup-shaped, multiflorous or uniflorous (the latter in pistillate inflorescences only), bracteate or not, interfloral bracts mostly peltate. Staminate flowers with tepals (1–)2–4 or absent, stamens 1–4 with filaments straight or inflexed in bud, pistillode present or (more often) absent. Pistillate flowers free or connate or embedded in the receptacle, tepals 2–4, ovary free or not, stigmas 1 or 1, equal or unequal. Fruits drupaceous or drupe-like due to a persistent fleshy perianth and/or receptacle, the whole endocarp unit often ballistically ejected from the infructescence. Seeds with or without endosperm, large or small; cotyledons equal or unequal.
Genera and distribution: eleven genera and 63 species with a worldwide distribution: Bleekrodea (Africa and Southeast Asia), Bosqueiopsis (Africa), Brosimum (Neotropics), Broussonetia (Asia to Oceania; introduced worldwide), Dorstenia (Africa and South America, with one species in India), Fatoua (Madagascar and Japan to New Caledonia; introduced worldwide), Helianthostylis (Neotropics), Malaisia (Southeast Asia to New Caledonia), Scyphosyce (Africa), Sloetia (Southeast Asia), Sloetiopsis (Africa), Trilepisium (Africa), Trymatococcus (Neotropics), Utsetela (Africa).
Sloetiopsis
Sloetiopsis Engl., Bot. Jahrb. 39: 573 (1907).
Neosloetiopsis Engl., Bot. Jahrb. 51: 426 (1914)
Streblus auct. non Lour., C.C. Berg, Proc. Kon. Ned. Acad. Wetensch. C. 91: 357. 1988.
Tree or shrubs, dioecious (or monoecious). Leaves distichous, pinnately veined, cystoliths present; stipules free, nearly amplexicaul. Inflorescences unisexual (or bisexual); staminate inflorescences spicate with an abaxial sterile groove, bracts mostly peltate, tepals 4, free or basally connate filaments inflexed in bud, pistillode small; pistillate inflorescences uniflorous, bracts mostly basally attached, tepals 4, free, imbricate, 2 stigmas. Fruits drupaceous, fleshy endocarp dehiscent, tepals enlarged and persistent but not fleshy. Seeds globose, ca. 4 mm, endocarp coriaceous with a hard disc against the hilum, testa vascularized with a thick apical cap, cotyledons equal.
1. Sloetiopsis usambarensis Engl., Bot. Jahrb. 39: 573 (1907) — Streblus usambarensis (Engl.) C.C. Berg, Proc. Kon. Ned. Acad. Wetensch. C. 91: 357. 1988. Neosloetiopsis kamerunensis Engl., Bot. Jahrb. 51: 426 (1914).
Note. The nearly amplexicaul stipules and occasional bisexual inflorescences reflect an affinity with the Southeast Asian Sloetia, also a member of the Dorstenieae.
Tribe PARARTOCARPEAE
Parartocarpeae N.J.C. Zerega & E.M. Gardner Zerega, Phytotaxa 388, 253–265 (2019). Shrubs to large trees, monoecious or dioecious; abundant white exudate. Leaves distichous or spirally arranged; simple; entire; pinnately veined; thin to thick coriaceous; glabrous, pubescent, scabrid, or hispid pubescent. Stipules axillary, simple or paired, lateral.
Inflorescences solitary or paired in leaf axils, unisexual (or bisexual with a single apical pistillate flower), uniflorous, cymose, or capitate with stamens or ovaries sunken into the receptacle; pedunculate; involucre of 3–8 triangular bracts, basally connate. Staminate inflorescences with numerous flowers, tepals 4–5, stamens 5 and normally positioned or 1– 3 in cavities in the receptacle with the anthers exerted through perforations in the upper surface of the receptacle, filaments free or united. Pistillate inflorescences uniflorous or (sub)globose with ovaries solitary in each cavity, unilocular, the style apical with a short exerted stigma. Fruits drupaceous, enclosed by persistant tepals, or aggregated into syncarps formed by the enlargement of the entire inflorescence.
Genera and distribution: Three genera and five species, from Southern China to the Solomon Islands: Hullettia, Parartocarpus, and Pseudostreblus.
Pseudostreblus
Pseudostreblus Bureau, in A.P. de Candolle, Prodr. 17: 220 (1873).
Streblus Lour. sect. Pseudostreblus (Bureau) Corner, Gard. Bull. Singapore 19: 217 (1962).
Monoecious trees. Leaves distichous, pinnately veined. Stipules free, lateral. Inflorescences axillary. Staminate inflorescences cymose, tepals 5, imbricate, filaments inflexed in bud but apparently straightening gradually upon anthesis, pistillode minute, conical, pubescent, interfloral bracts few, basally attached. Pistillate inflorescences uniflorous, pedunculate, involucral bracts 3, ±connate, tepals 4, imbricate, stigma bifid, arms equal. Fruits drupaceous, ca. 1 cm long, with tepals enlarged and loosely enclosing the fruit. Seeds ellipsoid, ca. 6 x 8 mm, cotyledons unequal.
Species and distribution: One, in South to Southeast Asia and Southern China. Note: Pseudostreblus is remarkable for its five-parted staminate flowers.
1. Pseudostreblus indicus Bureau, in A.P. de Candolle, Prodr. 17: 220 (1873).
Streblus indicus (Bureau) Corner, Gard. Bull. Singapore 19: 226 (1962).
Conclusion
The revisions presented here, based on the best phylogenetic evidence available to date, provide for seven monophyletic tribes of Moraceae and eleven monophyletic genera within Moreae. We hope that the resulting taxonomic and phylogenetic framework will provide a solid foundation for further research into the systematics and evolution of Moraceae.
Author contributions
EMG led the writing of the manuscript, did field, lab, and herbarium work, conducted all phylogenetic analyses, and was responsible for the taxonomic treatment. MG performed lab work for the Moraceae333 samples. RC, SD, NE, and OM performed lab work for the PAFTOL samples, and OM curated the PAFTOL data. DA collaborated on all herbarium work at BO. Sahromi facilitated access and participated in field collections at the Bogor Botanical Gardens. NJCZ provided overall guidance on Moraceae and facilitated sequencing at Northwestern University. WJB and FF acquired funding for and supervised the PAFTOL project. AM performed herbarium sampling at K and supervised the Moraceae portion of the PAFTOL project. EMG designed the overall study with input from ALH and NJCZ. ALH supervised the overall study. All authors commented on and contributed to the manuscript.
Conflicts of interest
The authors are unaware of any conflicts of interest affecting this research.
Acknowledgements
This work was supported by the United States National Science Foundation (DBI award number 1711391, and DEB award numbers 0919119 and 1501373); the SYNTHESYS Project http://www.synthesys.info/ which is financed by European Community Research Infrastructure Action under the FP7 “Capacities” Program” the National Parks Board of Singapore through a Flora of Singapore Fellowship and a Research Fellowship; grants from the Calleva Foundation and the Sackler Trust to the Plant & Fungal Trees of Life Project at the Royal Botanic Gardens, Kew. We thank the Pritzker Laboratory for Molecular Systematics at the Field Museum of Natural History (K. Feldheim) for access to laboratory and sequencing facilities; S. Knapp and J. Wajer (BM) for helpful consultation on nomenclatural issues; W. Clement for helpful comments on the circumscription of the Olmedieae; the Bogor Botanical Gardens, Singapore Botanic Gardens, and The Morton Arboretum for access to living collections; and the following herbaria for access to collections for examination and/or sampling: BM, BKF, BO, F, IBSC, K, FTBG, L, MO, NY, SING, SAN, SAR, SNP, and U. Finally, we gratefully acknowledge the careful work of past scholars of Moraceae, in particular the late C.C. Berg and E.J.H. Corner.