Skip to main content

Phylogenetic and comparative analyses of Hydnora abyssinica plastomes provide evidence for hidden diversity within Hydnoraceae



To date, plastid genomes have been published for all but two holoparasitic angiosperm families. However, only a single or a few plastomes represent most of these families. Of the approximately 40 genera of holoparasitic angiosperms, a complete plastid genome sequence is available for only about half. In addition, less than 15 species are currently represented with more than one published plastid genome, most of which belong to the Orobanchaceae. Therefore, a significant portion of the holoparasitic plant plastome diversity remains unexplored. This limited information could hinder potential evolutionary pattern recognition as well as the exploration of inter- and intra-species plastid genome diversity in the most extreme holoparasitic angiosperms.


Here, we report the first plastomes of Kenyan Hydnora abyssinica accessions. The plastomes have a typical quadripartite structure and encode 24 unique genes. Phylogenetic tree reconstruction recovers the Kenyan accessions as monophyletic and together in a clade with the Namibian H. abyssinica accession and the recently published H. arabica from Oman. Hydnora abyssinica as a whole however is recovered as non-monophyletic, with H. arabica nested within. This result is supported by distinct structural plastome synapomorphies as well as pairwise distance estimates that reveal hidden diversity within the Hydnora species in Africa.


We propose to increase efforts to sample widespread holoparasitic species for their plastid genomes, as is the case with H. abyssinica, which is widely distributed in Africa. Morphological reinvestigation and further molecular data are needed to fully investigate the diversity of H. abyssinica along the entire range of distribution, as well as the diversity of currently synonymized taxa.

Peer Review reports


Holoparasitism has evolved in ten out of twelve parasitic angiosperm lineages, resulting in a great diversity of genera and species [1]. The trophic change from auto- to full heterotrophy is accompanied by drastic changes to the plastid genome [2], culminating in highly condensed plastomes with respect to gene content as well as genome size [3,4,5]. Alongside the gene function losses, changes to the otherwise extremely conserved plastome structure are often observed. Many so called minimal plastomes do not maintain the typical quadripartite structure due to the loss of one of the inverted repeat copies [3,4,5,6,7]. Additionally, many smaller inversions and translocations can be observed [2, 8], which may be correlated to the increased AT content of plastomes of heterotrophic plants [9, 10]. Based on published data, general models were proposed to explain lineage-independent patterns of plastome condensation during adaptation to a heterotrophic lifestyle [7, 9,10,11,12]. However to date, no such models exist for the “final stages of plastome condensation”, the minimal plastomes, as a variety of different structures and gene contents have been observed [3,4,5,6,7]. This could either be due to lineage-specific factors being at play or be at least partially, due to lack of data.

At the time of this study, all holoparasitic angiosperm families are represented with at least a single published plastome, with the exception of Mystropetalaceae (Santalales) and Rafflesiaceae (Malpighiales). For the latter, independent studies have hypothesized the complete plastid genome loss [13, 14]. The NCBI database reviewing ( reveals that, out of the ~ 40 genera of holoparasitic angiosperms, only about half of them have full plastid genome sequences available. In addition, less than 15 species are currently represented by more than one published plastid genome (mostly limited to the Orobanchaceae). Therefore, a significant fraction of the diversity remains currently unexplored. This limited information might hinder, among others, potential evolutionary pattern recognition as well as explorations on inter-species and intra-species diversity of plastid genomes.

Plastomes of Hydnoraceae are highly condensed with respect to genome size and gene content and among the smallest known to date [7, 8]. The typical quadripartite plastome structure is shown missing in genus Prosopanche de Bary as well as in Hydnora esculenta Jum. & H. Perrier [6, 8]. The remaining published species of Hydnora Thunb., clustered in two clades, show differences with respect to their IR positioning and content, allowing for multiple evolutionary scenarios of IR and direct repeat evolution [8]. Apart from structural differences, the plastomes show an extreme degree of divergence on nucleotide sequence level [6]. The nucleotide composition is highly biased towards A and T, with an average of 23.1% GC in genus Hydnora and average of 21.7% GC in Prosopanche [8], compared to an average of 34–39% in photosynthetic angiosperms [15].

The family Hydnoraceae consists of the two genera Hydnora [16] and Prosopanche [17]. The former being distributed from Southern Africa to the Arabian Peninsula and Madagascar, the latter from South to Central America [18, 19]. This holoparasitic family of root parasites is among the oldest parasitic angiosperm lineages [20], and places within the magnoliid order Piperales. Together with Aristolochiaceae, Asaraceae, and Lactoridaceae they form the perianth-bearing clade within the order [21]. Prosopanche consists of seven species [22] whereas eight species are currently recognized in Hydnora [23, 24]. Hydnoraceae are unique among angiosperms due to their extreme loss of morphological features, e.g. the loss of leaves or scales and the ability to release a rotting smell that attracts beetles for pollination, hence described as the “strangest plants on earth” [25]. The flowers and fruits are the only aboveground parts of the mostly subterranean Hydnoraceae plants. The only exception to this is H. triceps Drege & E. Mey., which is known to complete its entire life cycle below ground [26, 27]. According to the Flora of tropical East Africa (FTEA), three Hydnora species have been recorded in Kenya, namely H. sinandevu Beentje & Q. Luke, H. abyssinica A. Braun, and H. africana Thunb [28]. These species are commonly being used as medicine, as source of food, and their roots are essential for tanning leather [24]. Hydnora abyssinca is described to have the largest distribution range among the members of the family, ranging from Southern Africa to northern East Africa [29]. Based on a comprehensive analysis of herbarium specimens, Musselman [18] recognized H. abyssinica, and treated Hydnora johannis Becc. and Hydnora solmsiana Dinter as synonyms of the first [18]. Additionally, Hydnora species are rarely found and collected due to their cryptic nature and seasonal appearance. Herbarium material preservation is often poor, which has made systematic revisions challenging. Apart from what is known about the natural history of Hydnora [18, 23, 26, 30]; there is only limited knowledge available on the morphogenesis and growth of this strange group of parasitic plants.

Here, we explore inter-species plastome diversity in Hydnoraceae for the first time. We report seven new plastomes of Hydnora abyssinica from Kenya and compare them with published plastomes of H. abyssinica from Namibia as well as to H. arabica Bolin & Musselman from Oman. Based on our genome reconstruction and a resolved phylogenetic hypothesis including a wider sampling in Hydnoraceae, we analyze their structure, nucleotide composition, as well as codon usage biasness.


The plastomes of Kenyan Hydnora abyssinica

We reconstructed seven circular plastid genomes for the newly sequenced H. abyssinica accessions (H0 - H6). The plastomes are identical with respect to gene content and quadripartite plastome structure, with little variation on sequence level. The average plastome size is 24,549 bp, with accession H2 (24,526 bp) having the smallest and accession H4 (24,565 bp) having the largest plastome size/length(Additional file 1: Table S1). The majority of the length difference for H2 stems from shorter IR copies, whereas the size increase for H4 stems from a slightly increased large single copy region. The plastomes show little variation with respect to overall nucleotide composition bias, with a 23.2% of GC in all accessions, except H3 where we recorded 23.5% of GC instead (Additional file 1: Table S1). We included ambiguous characters at plastome positions with low readmapping support and within the ycf2 gene of accessions H3 and H5 to allow for circularization. PCR and Sanger sequencing were used to verify the regions, for example, verify the IR – SSC ssc region.

The plastid genomes of all seven Kenyan H. abyssinica accessions contain an identical gene set, consisting of 24 unique genes (Fig. 1). The gene set is comprised of 17 protein-coding genes, four rRNA genes as well as three tRNAs. The single remaining intron is the trans-spliced intron of the rps12 gene (Fig. 1). Two complete (rrn4.5, rrn5) and two partial (rrn23, rpl16) genes are duplicated in the inverted repeats of the Kenyan H. abyssinica accessions (Fig. 1).

Fig. 1
figure 1

The plastid genome of Kenyan Hydnora abyssinica and its flower. The accession H0 is depicted as representative of accessions H0 – H6. Genes inside the outer circle are transcribed clockwise, genes on the outside are transcribed counter clockwise. The inner grey circle visualizes the GC content across the plastome. The center of the figure contains a picture of a longitudinally dissected Hydnora abyssinica flower from Kenya shortly before anthesis

Phylogenetic placement of Kenyan Hydnora abyssinica

Phylogenetic tree reconstruction of an 83-plastid loci data set recovers the Kenyan H. abyssinica accessions as monophyletic with 100 bootstrap support (BS). An equally high BS is recovered for the sister group relationship of H. abyssinica accessions H0 and H6. However, other nodes among the Kenyan accessions receive low to no BS (Fig. 2A, Fig. S1). The newly sequenced accessions place sister to H. arabica with 100 BS and successively sister to the Namibian H. abyssinica (BS 100). Hydnora formally referred to as Hydnora clade II was expanded by Jost et al. [8],using the newly sequenced plastomes.

Branch lengths within Hydnora clade II are relatively short, with the longest branches separating the Namibian H. abyssinica from H. arabica and the Kenyan sampling (Fig. 2B). Within the latter, we observed short branch lengths stemming from low sequence variation compared to what is seen in Hydnoraceae in general (Additional file 2: Table S2).

Fig. 2
figure 2

Phylogenetic placement of Kenyan Hydnora abyssinica accessions based on an 83 plastid-loci data set as clado- (A) and phylogram (B). Note that H.abyssinica, with H. arabica nested within, is not recovered as monophyletic, with Hydnora arabica. Outgroups are represented in a condensed form in (A) and completely removed in (B) to allow for better focus on the study group. The complete trees can be found in Figure S1. Inferences were calculated using RAxML. Bootstrap support values are shown along the respective nodes. Members of Hydnora clades I and II (based on Jost et al. [8] are highlighted in green and blue, respectively

Structural plastome comparisons among H. abyssinica and relatives

Although branch lengths among the focal taxa are relatively short, a structural plastome comparison reveals striking differences (Fig. 3). The orientation of the rps12 exon 2 is the only exception to an otherwise congruent gene order and orientation. The exon is inverted in Namibian H. abyssinica compared to the Kenyan accessions and H. arabica (Fig. 3). Additional differences concern the gene content of the Hydnora clade II IRs, in the plastome of Namibian H. abyssinica, the repeat consists of partial copies of rpl16 and rps3, whereas the IR of Kenyan H. abyssinica accessions contains full-length copies of rrn4.5 and rrn5, along with partial copies of rrn23 and rpl16 (highlighted green in Fig. 3). The IRs of H. arabica are identical to the ones of Kenyan H. abyssinica with respect to gene content, while additionally including a partial copy of rps3 (Fig. 3). Lastly, the rps19 and ycf1 genes do not partially overlap in the plastome of Namibian H. abyssinica and instead have a four bp spacer in between, whereas the genes show a four bp overlap in H. arabica and Kenyan H. abyssinica.

Fig. 3
figure 3

Structural plastome comparision among Namibian and Kenyan H. abyssinica and H. arabica (Hydnora clade II). The linear plastomes highlight incongruences and similarities among Namibian and Kenyan H. abyssinica and H. arabica. The green area highlights the positions and size of the respective IRs. Plastome regions (LSC, SSC, and IR copies) are indicated below

Relative synonymous codon usage

RSCU analysis of the Hydnoraceae protein-coding genes reveals an extreme bias towards A and T nucleotides (Additional file 3: Table S3), complementary to the plastomes’ overall bias (Additional file 1: Table S1). For all amino acids, the most used codons correspond to the ones with the highest AT content (Additional file 3: Table S3). The preferred codons are mostly consistent among the sampled Hydnora and Prosopanche species, but also between the genera and the outgroup (Aristolochia, Lactoris, and Thottea). For four amino acids however, namely alanine, glycine, isoleucine, and serine, differences within the parasitic family have been uncovered (Additional file 4: Table S3). For alanine, the most used codon in Prosopanche and Hydnora clade I (H. africana, H. longicollis Welw, H. triceps, and H. visseri Bolin, E. Mass & Musselman) was GCT. In Hydnora clade II and H. esculenta however, GCA was the preferred codon, with GCT being only the second most used synonymous codon. The same pattern has been identified for the preferred codons for isoleucine. Prosopanche and Hydnora clade I favor ATA, whereas Hydnora clade II and H. esculenta favor ATT. Here, P. bonacinae Speg. is considered an outlier, also with ATT estimated to be the most used codon (Additional file 3: Table S3). The predominant use of synonymous codons for serine highlights a case of different codon usage between Prosopanche and Hydnora. The most used synonymous serine codon for the latter is TCA, whereas Prosopanche favors TCT, with equal proportional use of TCT and AGT in P. bonacinae and P. americana. The most diverse use of preferred synonymous codons was observed for glycine, where Hydnoraceae species shifted between GGA and GGT (Additional file 3: Table S3). Hydnora clade II, H. visseri and P. americana prefer GGA, whereas the remainder of the sampling favor GGT. However, H. longicollis is estimated to have equal usage of GGA and GGT.

Comparison of the RSCU among Hydnoraceae and heterotrophic outgroup taxa reveals a much more even usage of synonymous codons for the latter, with only a low fraction of codons showing ≤ 10% usage (Additional file 3: Table S3).


Hidden diversity within Hydnora abyssinica and relatives

The non-monophyly of H. abyssinica accessions from Kenya and Namibia, supported by structural plastome evidence, serves as first phylogenomic evidence for hidden diversity within H. abyssinica. The observed plastome reconfiguration, along with the inverted repeat content, compliment a supported tree reconstruction (Figs. 2 and 3). Based on these lines of evidence and to aid in clarification for future studies, we propose to recognize the Kenyan and Namibian accessions of H. abyssinica as distinctly different species. This proposal receives support by a recent morphological systematic revision, reconsidering H. arabica from Oman as H. hanningtonii and H. abyssinica from Namibia as H. solmsiana [31]. Following this taxonomic revision, our study provides the first plastid genomes for Hydnora abyssinica. However, it needs to be noted that the current molecular evidence for splitting species stems from plastid data only. Further nuclear data will be needed to confirm our findings as well as the new systematic treatments in general.

Morphological characterization of many holoparasites, including Hydnoraceae, proves difficult due to a lack of features combined with the uncertain assignment of some remaining ones [31]. Furthermore, herbarium specimens are often poorly preserved and mostly floral structures are documented, whereas conserved underground organs are rare. Hydnora abyssinica shows the widest distribution range among all members of Hydnoraceae. It is occurring in arid and semi-arid biomes from South Africa to northern East Africa [23,24,25, 27, 28, 31,32,33,34]. An approximately equally broad distribution range was proposed for closely related Prosopanche americana in the New World. Meanwhile, several additional Prosopanche species have been published from southern America to Central America in recent years [35,36,37,38,39]. Our study might provide evidence, based on molecular data, that additional species could potentially be recognized given detailed morphological and field observations as started by Hatt et al. [30]. Indeed, morphologically similar species to H. abyssinica have been published in the past, namely H. sinandevu (not sampled here) as well as H. solmsiana Dinter, and H. johannis Becc. The latter two have been treated as synonym of H. abysssinica [18, 31]. The first, H. sinandevu, is treated as an independent species and parasitizes a different host plant in Kenya (Commiphora africana (A. Rich.) Engl., Burseraceae)). In comparison, all other species mentioned above exclusively parasitize Fabaceae [31]. However, further synonyms of H. abyssinica exist. According to taxonomic treatments of the last decades, at least eight further synonymous taxa have been described: Hydnora aethiopica Decne., Hydnora angolensis Decne., Hydnora abyssinica A. Braun var. quinquefida Engl, Hydnora bogoscensis Becc, Hydnora cornii Vaccaneo, Hydnora gigantea Chiov., Hydnora hanningtinii Rendle, and Hydnora michaelis Peter [18, 23, 28, 31, 40]. Therefore, it is not unlikely that a greater diversity is currently hidden due to the lack of morphological characteristics and the pooling of taxa based on inaccessible characters.

For the future, an expanded taxon sampling of accessions currently considered H. abyssinica and its synonyms, covering the entire known range, is highly desirable to reveal diversity within Hydnora clade II and support further taxonomic revision. Our study showed the viability of molecular data to aide in species deciphering and the present dataset might serve as a starting point to design additional molecular markers to help with species identification in the light of limited morphological data.

Plastome structural features support phylogenetic findings and systematic revision

Within Hydnoraceae, especially genus Hydnora, a fair number of structural plastome differences have been observed, which seem to correlate to specific clades [8]. The proposed Hydnora clade II, comprised of H. arabica and Namibian H. abyssinica (named H. hanningtonii and H. solmsiana from now on), is characterized by inverted repeats distinct from the majority of Hydnora species, along with two independent translocation and inversion events on the branch leading to it [8]. Phylogenetic inference of the newly sequenced accessions of this study reconstructs their placement within this Hydnora clade II (Fig. 2, Additional file 2: Fig. S1). This placement is fully supported based on bootstrap values as well as supported by structural plastome elements. The sampled H. abyssinica, H. hanningtonii, and H. solmsiana share two independent events of sequence translocation and inversion (rps3 – rpl16 region and rpl14) compared to the remainder of the Hydnoraceae published to date, along with an inversion of rps7 [8]. This adds support to the hypothesis made by Jost et al. [8], proposing that these events happened on the branch leading to this clade and prior to the diversification within. Two major structural differences within the clade can be observed however: (i) the inversion of the rps12 exon 2 of H. solmsiana (Namibia) compared to H. hanningtonii (Oman) and H. abyssinica (Kenya), and (ii) the distinct IR gene content for each of the three (Fig. 3).

Our findings indirectly support the new systematic treatment of Hydnora [31] as otherwise H. abyssinica would render non-monophyletic, as H. arabica is nested within. In line with the topological placement and structural plastome evidence is the result of the pairwise distance analysis, which highlights a larger distance of H. solmsiana to the remainder of Hydnora clade II than for H. hanningtonii to the Kenyan accessions (Additional file 3: Table S2). Based on the data at hand and little variation within the newly sequenced accessions, even on full plastome level (Additional file 3: Table S2), we are however not able to confidently resolve the relationships among the Kenyan accessions further; nuclear markers will potentially be required.

Hydnoracae plastome evolution

The seven reconstructed plastomes of Kenyan Hydnora abyssinica accessions reveal a striking degree of congruence, both on structural and sequence level (Additional file 1: Table S1 and Additional file 3: Table S2). All share the identical quadripartite structure with four genes duplicated within the IRs (rrn4.5, rrn5, rrn23, rpl16), and an identical gene order and orientation. With an average length of 24,549 bp, the plastomes of Kenyan H. abyssinica place on the lower size range within Hydnoraceae, together with H. esculenta (24,479 bp), H. hanningtonii (24,790 bp), and H. solmsiana (24,672, Jost et al. [8]). The studied accessions contain a set of 24 unique genes (Fig. 1), identical with other published Hydnora [7, 8]. Together, Hydnoraceae possess some of the most condensed parasitic plastid genomes known to date [6,7,8], for example “Pilostyles (Apodanthaceae) [3, 4], and Balanophora (Balanophoraceae) [5]”.

Adaptation to the holoparasitic lifestyle leads to an extreme bias for Adenosine and Thyrosine nucleotides, compared to the generally already AT rich plastomes of autotrophic plants. This compositional bias affects both coding and non-coding regions [9] as well as all three codon positions, resulting in notable changes in codon usage. Hydnoraceae plastomes consist to ~ 76–80% of AT (Additional file 1: Table S1, [6,7,8], complemented by an extreme nucleotide bias for chosen synonymous codons (Additional file 4: Table S3). This compositional bias could be the result of the general trend towards A and T nucleotides during a plants adaptation to a heterotrophic lifestyle [9], or be to compensate for limited tRNA import after the plastome-encoded tRNA genes have been lost. In eukaryotes, the codons of highly expressed genes are best recognized by the most abundant tRNAs, resulting from a translation efficiency selection [41, 42]. Codon usage between Hydnora and Prosopanche is congruent with mostly identical prevalence for each of the synonymous codons per amino acid (Additional file 4: Table S3). Even though the Prosopanche plastomes maintain two additional tRNAs (trnY-GUA and trnW-CCA) compared to Hydnora [6, 8], the codons corresponding to trnY and trnW do not show significantly higher usage in Prosopanche compared to the sister genus (Additional file 4: Table S3). However, since the RSCU analysis reveals that these codons are clearly being used in Hydnora, but lost from their plastome, an import mechanism for them has to exist. To service a specific codon, the corresponding tRNA as binding partner is required or an unspecific one that is able to read multiple codons, following the wobble [43] or superwobble rules [44]. An import for these two tRNAs either could be Hydnora-specific or have already existed in a condensed Hydnoraceae MRCA plastome, resulting in non-essential trnY and trnW copies in Prosopanche. Such a cytosolic import is likely to already exist for most of the other tRNAs lost, yet still used by the Hydnoraceae plastomes. This theory gains additional support by point mutation altering the anticodon of trnW in P. americana, rendering it putatively non-functional [6].


We newly sequenced and assembled for the first time plastid genomes of seven H. abyssinica accessions from Kenya and compared them with publicly available Hydnoraceae plastomes. Based on multiple lines of evidence and following a taxonomic revision by Hatt et al. [31], we support the recognition of H. solmsiana in Namibian (previously recognized as H. abyssinica) and H. hanningtonii (the correct name of H. arabica) in Oman. The plastomes of Kenyan H. abyssinica are identical to one another with respect to gene content and quadripartite plastome structure, with little variation on sequence level. Plastome comparison reveals that the orientation of the rps12 exon 2 is the only exception to an otherwise congruent gene order and orientation within Hydnora clade II. The exon is inverted in H. solmsiana compared to the H. abyssinica and H. hanningtonii, serving as a structural feature supporting phylogenetic findings. The phylogenetic tree reconstruction based on an 83-plastid loci alignment displayed H. solmsiana sister to a clade comprising H. hanningtonii and H. abyssinica, support the segregation of H. solmsiana from the synonymy of H. abyssinica, as proposed in the taxonomic revision of the genus. An expanded taxon sampling considering H. abyssinica and its relatives, covering the entire distribution range is highly desirable to uncover the diversity within Hydnora clade II and further aid a taxonomic revision. The present dataset might serve as a starting point to design additional molecular markers to help with species identification in the light of limited morphological data. However, a complementary and complete morphological reinvestigation is required as well.

Materials and methods

Sample collection and DNA extraction

Seven populations of H. abyssinica (H0 – H6) were collected by the authors with field assistance by scientists from the National Museum of Kenya at Mt. Kasigau (Taita Hills), a mountain range located in the Taita-Taveta County in south-eastern Kenya. The formal identification of the plant was done by Professor Guang Guang-Wan Hu from Wuhan Botanical Garden, Chinese Academy of Sciences with the help of National Museums of Kenya senior Scientist Dr. Geofrey Mwachala. These locations can be characterized as dry savanna grassland marked by Fabaceae trees (e.g. Albizia Durazz., Delonix Raf., Kigelia DC., Piliostigima Hochst.), and Euphorbiaceae species (e.g. Euphorbia tirukali L.) as well as in the Tsavo East and West National Park. The corresponding collection information and vouchers are provided in Additional file 5: Table S4. Samples were deposited at the Wuhan Botanical Garden herbarium. Within the Taita Hills mountain range, Hydnora could not be observed at Ngangao, Iyale, Msidunyi, or Ndiwenyi, but exclusively at Mt. Kasigau, Voi, Mbolo/Ronge juu, Kishushe, and Ndii near the border of Kenya and Tanzania. Collection permits were obtained under the Sino Africa Joint Research Center in 2019. Genomic DNA was isolated using a cetyltrimethylammonium bromide method [45]. All sampled Kenyan Hydnora abyssinica were growing on Vachelia, Albizia, and Kigelia.

Illumina sequencing and plastome assembly

The H. abyssinica libraries were constructed with an insert size of 250 bp, following the manufacturer’s standard method of TruSeq DNA sample preparation kits and was sequenced on an Illumina Hiseq-2000 platform at Novogene Co (Beijing, China). Low-quality data and adaptors were filtered and trimmed (quality scores < 20 were trimmed). We obtained between ~ 45 mio. raw reads for accession H. abyssinica H0 and ~ 22 mio reads for accessions H1-H6.

The sequence raw data were assembled using both a de novo assembly and a seed-based assembly approach. The de novo assemblies were done using CLC Genomics workbench (v. 11.1, Biomatters, Ltd., New Zealand), allowing for automatic estimation of optimal bubble and word sizes for each sample. Subsequently, to verify the correctness of the assemblies, readmappings were conducted using the same software. The assemblies were filtered for contigs with putative plastid origin (BLASTn, E-value 1e-10). Seed-based reconstructions of plastid scaffolds were carried out in GetOrganelle (v., [46] using standard settings. The filtered contigs as well as the plastome scaffolds created by GetOrganelle were imported into Geneious (v. 11.1.5, Biomatters, Ltd., New Zealand), where they were further assembled into complete plastid genomes.

Primer design and plastome verification

To verify plastome regions with low read coverage, to bridge assembly gaps, and to verify IR boundaries we designed primers (Additional file 5: Table S5) using PhyDE (Phylogenetic Data Editor, [47]). The Polymerase chain reaction (PCR) amplification was carried out on a A300 thermal cycler (Hangzhou Long Gene Scientific Instruments Co., Ltd.) with the following settings: denaturation at 98ºC for 2 min, followed by 35 cycles of 98ºC for 10 s, 10 s annealing at 57ºC, 45 s extension at 72ºC, and a final extension step at 72ºC for 5 min. The amplifications were performed in 40 µL reactions, containing 20 µL 2xT8 High-Fidelity Master Mix (Tsingke Biotechnology Co., Ltd.), 2 µL of each primer, 1 µL DNA template, and 15 µL double distilled water. A DNA agarose gel extraction kit was used to clean the PCR products after separation by electrophoresis on a 0.5% agarose gel. Products were then Sanger sequenced as both the forward and reverse strand on an ABI 3730 XL capillary electrophoresis sequencer (Tsingke Biotechnology Co., Ltd., Wuhan, China). The sequenced products were then aligned to the respective plastome regions after manually checking the respective chromatogram.

Genome annotation and visualization

Newly sequenced and assembled plastomes were annotated in Geneious (v. 11.1.5, Biomatters, Ltd., New Zealand), using the published plastome of H. abyssinica (NC_065049) [8]. Annotations were manually inspected and adjusted where necessary. For protein-coding genes, annotations contain complete open reading frames, including a start and stop codon. The plastomes were visualized as both circular and linear structures using OGDRAW (Organellar Genome DRAW)[48].

Phylogenetic analyses

To reconstruct the phylogenetic placement of the seven newly sequenced H. abyssinica accessions, we created alignments of 83 protein-coding and rRNA genes based on the data set of [8], including a representative sampling among ANA-grade taxa, magnoliids, eudicots, and monocots. Hydnoraceae are represented with 21 genes. Alignments were constructed using the MAFFT algorithm (v. 7.450, [49, 50] and manually adjusted using AliView (v. 1.20) [51]. All genes were subsequently concatenated using Geneious (v. 11.1.5, Biomatters, Ltd., New Zealand) and inferences were calculated using RAxML (v. 8.2.12) [52] implemented in Cipres Science Gate [53, 54], applying a gene partitioning approach. The best-fitting nucleotide substitution model was estimated prior using PartitionFinder2 [55, 56]. Tree files were visualized using TreeGraph 2 [57].

Relative synonymous codon usage

The relative synonymous codon usage (RSCU, [58]) was estimated for the set of Hydnoraceae protein coding genes using the Sequence Manipulation Suite (SMS,[59]). The comparison was done among the Hydnoraceae sampling of Jost et al. [8] and the newly sequenced H. abyssinica H0 of this study as well as closely related heterotrophic taxa (Aristolchia fimbriata Cham. CM034085, Lactoris fernandeziana Phil. NC_065383, and Thottea sumatrana (Merr.) Ding Hou NC_065017). Accession H. abyssinica H0 serves as a representative for the six other, newly sequenced accessions. The RSCU is estimated as the observed frequency of a specific codon divided by its expected frequency, assuming equal codon usage for synonymous codons.

Data availability

All the accession numbers and the newly sequenced (OP495727,OP495728,OP495729,OP495730,OP495731,OP495732,OP495733) and assembled plastomes for seven accessions of Hydnora abyssinica have been uploaded to the National Center for Biotechnology Information database as attached to the additional file Table S5.



Bayesian Inference


Bootstrap support


Cetyltrimethylammonium bromide

IR (IRa/IRb):

Inverted repeats


Large single copy


Maximum Likelihood


Polymerase chain reaction


Relative synonymous codon usage


Ribosomal RNA genes


Sino Africa Joint Research Center


Short small single copy


Transfer RNAs genes


Wuhan Botanical Garden


  1. Nickrent DL. Parasitic angiosperms: how often and how many. Taxon. 2020;69(1):5–27.

    Article  Google Scholar 

  2. Wicke S, Naumann J. Molecular evolution of plastid genomes in parasitic flowering plants. In: advances in botanical research. Elsevier. 2018; 315–47.

  3. Arias-Agudelo LM, González F, Isaza JP, Alzate JF, Pabón-Mora N. Plastome reduction and gene content in New World Pilostyles (Apodanthaceae) unveils high similarities to african and australian congeners. Mol phylogenet Evol. 2019;135:193–202.

    Article  PubMed  Google Scholar 

  4. Bellot S, Renner SS. The plastomes of two species in the endoparasite genus Pilostyles (Apodanthaceae) each retain just five or six possibly functional genes. GBE. 2016;8:189–201.

    CAS  Google Scholar 

  5. Su H-J, Barkman TJ, Hao W, Jones SS, Naumann J, Skippington E, Wafula EK, Hu J-M, Palmer JD, DePamphilis CW. Novel genetic code and record-setting AT-richness in the highly reduced plastid genome of the holoparasitic plant Balanophora. Proc Natl Acad Sci. 2019;116:934–43.

    Article  CAS  PubMed  Google Scholar 

  6. Jost M, Naumann J, Rocamundi N, Cocucci AA, Wanke S. The first plastid genome of the holoparasitic genus Prosopanche (Hydnoraceae). Plants. 2020;9:306.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Naumann J, Der JP, Wafula EK, Jones SS, Wagner ST, Honaas LA, Ralph PE, Bolin JF, Maass E, Neinhuis C. Detecting and characterizing the highly divergent plastid genome of the nonphotosynthetic parasitic plant Hydnora visseri (Hydnoraceae). GBE. 2016;8:345–63.

    CAS  PubMed  PubMed Central  Google Scholar 

  8. Jost M, Naumann J, Bolin JF, Martel C, Rocamundi N, Cocucci AA, Lupton D, Neinhuis C, Wanke S. Structural plastome evolution in holoparasitic Hydnoraceae with special focus on inverted and direct repeats. GBE. 2022;14:77.

    Google Scholar 

  9. Wicke S, Müller KF, de Pamphilis CW, Quandt D, Wickett NJ, Zhang Y, Renner SS, Schneeweiss GM. Mechanisms of functional and physical genome reduction in photosynthetic and nonphotosynthetic parasitic plants of the broomrape family. PLANT CELL. 2013;25:3711–25.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Wicke S, Müller KF, DePamphilis CW, Quandt D, Bellot S, Schneeweiss GM. Mechanistic model of evolutionary rate variation en route to a nonphotosynthetic lifestyle in plants. Proc Natl Acad Sci. 2016;113:9045–50.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Barrett CF, Davis JI. The plastid genome of the mycoheterotrophic Corallorhiza striata (Orchidaceae) is in the relatively early stages of degradation. AJB. 2012;99:1513–23.

    Article  CAS  PubMed  Google Scholar 

  12. Graham SW, Lam VK, Merckx VS. Plastomes on the edge: the evolutionary breakdown of mycoheterotroph plastid genomes. New Phytol. 2017;214:48–55.

    Article  CAS  PubMed  Google Scholar 

  13. Cai L, Arnold BJ, Xi Z, Khost DE, Patel N, Hartmann CB, Manickam S, Sasirat S, Nikolov LA, Mathews S. Deeply altered genome architecture in the endoparasitic flowering plant Sapria himalayana Griff. (Rafflesiaceae). Curr Biol. 2021;31:1002–11.

    Article  CAS  PubMed  Google Scholar 

  14. Molina J, Hazzouri KM, Nickrent D, Geisler M, Meyer RS, Pentony MM, Flowers JM, Pelser P, Barcelona J, Inovejas SA. Possible loss of the chloroplast genome in the parasitic flowering plant Rafflesia lagascae (Rafflesiaceae). Mol Biol Evol. 2014;31:793–803.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Cai Z, Penaflor C, Kuehl JV, Leebens-Mack J, Carlson JE, DePamphilis CW, Boore JL, Jansen RK. Complete plastid genome sequences of Drimys, Liriodendron, and Piper: implications for the phylogenetic relationships of magnoliids. BMC Evol Biol. 2006;6:1–20.

    Article  Google Scholar 

  16. Thunberg C. Beskrifning paa en ganska besynnerlig och obekant svamp, Hydnora africana. Kongalia Vetenskaps Akademiens Handlingar 1775; 36:69–75.

  17. de Bary A. Prosopanche burmeisteri, eine neue Hydnoree aus Süd-Amerika. HW Schmidt; 1868.

  18. Musselman L. The genus Hydnora (Hydnoraceae). In: 5 International Symposium of Parasitic Weeds, Nairobi (Kenya), 24–30 Jun 1991: 1991. CIMMYT.

  19. Sequeira AS, Rocamundi N, Ferrer MS, Baranzelli MC, Marvaldi AE. Unveiling the history of a peculiar Weevil-Plant interaction in South America: a phylogeographic approach to Hydnorobius hydnorae (Belidae) Associated with Prosopanche americana (Aristolochiaceae). Diversity. 2018;10:33.

    Article  Google Scholar 

  20. Naumann J, Salomo K, Der JP, Wafula EK, Bolin JF, Maass E, Frenzke L, Samain M-S, Neinhuis C, DePamphilis CW. Single-copy nuclear genes place haustorial Hydnoraceae within Piperales and reveal a cretaceous origin of multiple parasitic angiosperm lineages. PLoS ONE. 2013;8:79204.

    Article  Google Scholar 

  21. Jost M, Samain MS, Marques I, Graham SW, Wanke S. Discordant phylogenomic placement of Hydnoraceae and Lactoridaceae within Piperales using data from all three genomes. Front Plant Sci. 2021;12:586.

    Article  Google Scholar 

  22. De Jdt C, Markus C, Junior AM, Zachia RA, Schuch M, Sulzbach E, Mariath JE. Prosopanche cocuccii (Hydnoraceae): a new species from Southern Brazil. Phytotaxa. 2021;521:177–92.

    Article  Google Scholar 

  23. Bolin JF, Lupton D, Musselman LJ. Hydnora arabica (Aristolochiaceae), a new species from the Arabian Peninsula and a key to Hydnora. Phytotaxa. 2018;338:1.

    Article  Google Scholar 

  24. Mkala EM, Mutungi MM, Mutinda ES, Oulo MA, Wanga VO, Mwachala G, Hu G-W. Understanding the ethnobotany, chemistry, pharmacology, and distribution of genus Hydnora (Aristolochiaceae). Plants. 2021;10:494.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Musselman L, Visser J. The strangest plant in the world. Veld Flora. 1986;71:109–11.

    Google Scholar 

  26. Maass E, Musselman LJ. Hydnora triceps (Hydnoraceae)—first record in Namibia and first description of fruits. Dinteria. 2004;29:1–10.

    Google Scholar 

  27. Tennakoon KU, Bolin JF, Musselman LJ, Maass E. Structural attributes of the hypogeous holoparasite Hydnora triceps Drège & Meyer (Hydnoraceae). AJB. 2007;94:1439–49.

    Article  PubMed  Google Scholar 

  28. Beentje H, Luke Q. Flora of Tropical East Africa-Hydnoraceae. CRC Press; 2002.

  29. Mkala EM, Jost M, Wanke S, Ngarega BK, Hughes A, Mutinda ES, Waswa EN, Mwanzia VM, Oulo MA, Wanga VO. How vulnerable are holoparasitic plants with obligate hosts to negative climate change impacts? Ecol Inf. 2022;69:101636.

    Article  Google Scholar 

  30. Harms H, Hydnoraceae. Vol 282–95. Wilhelm Engelmann, Leipzig, Germany.; 1935.

  31. Hatt S, Thorogood C, Bolin J, Musselman L, Cameron D, Grace O. A taxonomic revision of the genus Hydnora (Hydnoraceae).

  32. Bolin JF, Maass E, Musselman LJ. A new species of Hydnora (Hydnoraceae) from Southern Africa. Syst Bot. 2011;36:255–60.

    Article  Google Scholar 

  33. Bolin JF, Musselman LJ. Epitypification and ecological notes for the Malagasy holoparasite Hydnora esculenta (Hydnoraceae). Nord J Bot. 2013;31:286–90.

    Article  Google Scholar 

  34. Seymour RS, Maass E, Bolin JF. Floral thermogenesis of three species of Hydnora (Hydnoraceae) in Africa. Ann Bot. 2009;104:823–32.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Denius HR Jr, Homann PH. The relation between photosynthesis, respiration, and crassulacean acid metabolism in leaf slices of Aloe arborescens Mill. Plant Physiol. 1972;49:873–80.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Funez LA, Ribeiro-Nardes W, Kossmann T, Peroni N, Drechsler-Santos ER. Prosopanche demogorgoni: a new species of Prosopanche (Aristolochiaceae: Hydnoroideae) from southern Brazil. 2019.

  37. Machado RF, Queiroz L. A new species of Prosopanche (Hydnoraceae) from northeastern Brazil. Phytotaxa. 2012;75:58–64.

    Article  Google Scholar 

  38. Martel C, Fernandez-Hilario R, Tello JA, Arteaga RG, Gerlach G. Prosopanche panguanensis (Aristolochiaceae), a new species from central Peru. Phytotaxa. 2018;364:241–9.

    Article  Google Scholar 

  39. Nickrent DL, Blarer A, Qiu YL, Soltis DE, Soltis PS, Zanis M. Molecular data place Hydnoraceae with Aristolochiaceae. AJB. 2002;89:1809–17.

    Article  CAS  PubMed  Google Scholar 

  40. Bolin JF. Ecology and molecular phylogenetics of Hydnora (Hydnoraceae) in Southern Africa. Old Dominion University; 2009.

  41. Sablok G, Nayak KC, Vazquez F, Tatarinova TV. Synonymous codon usage, GC3, and evolutionary patterns across plastomes of three pooid model species: emerging grass genome models for monocots. Mol Biotechnol. 2011;49:116–28.

    Article  CAS  PubMed  Google Scholar 

  42. Stenico M, Lloyd AT, Sharp PM. Codon usage in Caenorhabditis elegans: delineation of translational selection and mutational biases. Nucleic Acids Res. 1994;22:2437–46.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Crick FH. Codon—anticodon pairing: the wobble hypothesis. J Mol Biol. 1966;19:548–55.

    Article  CAS  PubMed  Google Scholar 

  44. Rogalski M, Karcher D, Bock R. Superwobbling facilitates translation with reduced tRNA sets. Nat Struct Mol Biol. 2008;15:192–8.

    Article  CAS  PubMed  Google Scholar 

  45. Porebski S, Bailey LG, Baum BR. Modification of a CTAB DNA extraction protocol for plants containing high polysaccharide and polyphenol components. Plant Mol biology Report. 1997;15:8–15.

    Article  CAS  Google Scholar 

  46. Jin JJ, Yu W-B, Yang JB, Song Y, Yi TS, Li D-Z. GetOrganelle: a simple and fast pipeline for de novo assembly of a complete circular chloroplast genome using genome-skimming data. BioRxiv. 2018;4:256479.

    Google Scholar 

  47. Muller K. Phylogenetic Data Editor. v0. 995. 2005.

  48. Greiner S, Lehwark P, Bock R. OrganellarGenomeDRAW (OGDRAW) version 1.3. 1: expanded toolkit for the graphical visualization of organellar genomes. Nucleic Acids Res. 2019;47:W59–W64.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Katoh K, Misawa K, Kuma Ki, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30:3059–66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30:772–80.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Larsson A. AliView: a fast and lightweight alignment viewer and editor for large datasets. Bioinformatics. 2014;30:3276–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics. 2014;30:1312–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Miller MA, Pfeiffer W, Schwartz T. The CIPRES science gateway: a community resource for phylogenetic analyses. In: Proceedings of the 2011 TeraGrid Conference: extreme digital discovery. 2011; 1–8.

  54. Miller MA, Pfeiffer W, Schwartz T. The CIPRES science gateway: enabling high-impact science for phylogenetics researchers with limited resources. In: Proceedings of the 1st Conference of the Extreme Science and Engineering Discovery Environment: Bridging from the extreme to the campus and beyond 2012; 1–8.

  55. Lanfear R, Calcott B, Kainer D, Mayer C, Stamatakis A. Selecting optimal partitioning schemes for phylogenomic datasets. BMC Evol Biol. 2014;14:1–14.

    Article  Google Scholar 

  56. Lanfear R, Frandsen PB, Wright AM, Senfeld T, Calcott B. PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Mol Biol Evol. 2017;34:772–3.

    CAS  PubMed  Google Scholar 

  57. Stöver BC, Müller KF. TreeGraph 2: combining and visualizing evidence from different phylogenetic analyses. BMC Bioinformatics. 2010;11:1–9.

    Article  Google Scholar 

  58. Sharp PM, Tuohy TM, Mosurski KR. Codon usage in yeast: cluster analysis clearly differentiates highly and lowly expressed genes. Nucleic Acid Res. 1986;14:5125–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Stothard P. The sequence manipulation suite: JavaScript programs for analyzing and formatting protein and DNA sequences. Biotechniques. 2000;28:1102–4.

    Article  CAS  PubMed  Google Scholar 

Download references


We thank Flory Mkangombe Kilingo, Benjamin Muema Watuma, Mwihaki John Karichu, Mwarabu and Bosswell Ramaliah MnJala, for their help during fieldwork at the Taita Hills region and their positive contributions towards success of this project. In addition, we acknowledge the support from John Nzei, Jacinta Ndunge Munyao, Milicent Akinyi Oulo, Vincent Okelo Wanga, Josphat K. Saina, Cornelius M. Kyalo, Peng shuai, during the lab analysis at Wuhan Botanical Garden. We acknowledge Chinese Government Scholarship program granted to Elijah Mbandi Mkala to pursue his PhD studies on this research through National Museums of Kenya.


This work was supported by grants from the National Natural Science Foundation of China (31970211, 31961143026), International Partnership Program of Chinese Academy of Sciences (151853KYSB20190027), from the Sino-Africa Joint Research Center, CAS (SAJC202101) and Chinese Government Scholarship Program University of Chinese Academy of Sciences. The authors affirm no conflicts of interest. The ERASMUS + funds provided by the European Union for the implementation of international mobility activities are highly appreciated, as they have enabled the exchange of students and staff for training and teaching. The funding body only provided the financial means to allow the authors to carry out the study.

Author information

Authors and Affiliations



GWH, MJ ,SW, QFW, GM and PMM designed the experiment and drafted the manuscript. EMM, MJ, and XD, performed the experiment ,analyzed the data and contributed reagent and analysis tools. EMM did the field collection and taking of pictures. All authors have read and approved the manuscript.

Corresponding author

Correspondence to Guang-Wan Hu.

Ethics declarations

Ethics approval and consent to participate

This research was carried under Wuhan Botanical Garden, Chinese Academy of science, project SAJOREC under supervision of Prof. Guang Wan Hu. The institution involved and all other parties agree to publish this work.

Consent for publication

Not Applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Mkala, E.M., Jost, M., Dong, X. et al. Phylogenetic and comparative analyses of Hydnora abyssinica plastomes provide evidence for hidden diversity within Hydnoraceae. BMC Ecol Evo 23, 34 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • Heterotrophy
  • Kenya
  • Taita hills
  • Plastome condensation
  • Monophyly
  • Piperales