Chromosome-level baobab genome illuminates its evolutionary trajectory and environmental adaptation

Kitony, Justine K.; Colt, Kelly; Abramson, Bradley W.; Hartwick, Nolan T.; Petrus, Semar; Konozy, Emadeldin H. E.; Karimi, Nisa; Yant, Levi; Michael, Todd P.

doi:10.1038/s41467-024-53157-w

Download PDF

Article
Open access
Published: 12 October 2024

Chromosome-level baobab genome illuminates its evolutionary trajectory and environmental adaptation

Nature Communications volume 15, Article number: 8833 (2024) Cite this article

7058 Accesses
2 Citations
10 Altmetric
Metrics details

Subjects

Abstract

Baobab (Adansonia digitata) is a long-lived tree endemic to Africa with economic, ecological, and cultural importance, yet its genomic features are underexplored. Here, we report a chromosome-level reference genome anchored to 42 chromosomes for A. digitata, alongside draft assemblies for a sibling tree, two trees from distinct locations in Africa, and A. za from Madagascar. The baobab genome is uniquely rich in DNA transposons, which make up 33%, while LTR retrotransposons account for 10%. A. digitata experienced whole genome multiplication (WGM) around 30 million years ago (MYA), followed by a second WGM event 3–11 MYA, likely linked to autotetraploidy. Resequencing of 25 trees identify three subpopulations, with gene flow across West Africa distinct from East Africa. Gene enrichment and fixation index (Fst) analyses show baobab retained multiple circadian, flowering, and light-responsive genes, which likely support longevity through the UV RESISTANCE LOCUS 8 (UVR8) pathway. In sum, we provide genomic resources and insights for baobab breeding and conservation.

The rise of baobab trees in Madagascar

Article Open access 15 May 2024

Analysis of 427 genomes reveals moso bamboo population structure and genetic basis of property traits

Article Open access 15 September 2021

Reshuffling of the ancestral core-eudicot genome shaped chromatin topology and epigenetic modification in Panax

Article Open access 07 April 2022

Introduction

The African baobab (Adansonia digitata) is a deciduous tree belonging to the Malvaceae family, specifically within the Bombacoideae subfamily. Hereafter, it will be simply referred to as “baobab” with other species like the Australian or Malagasy baobab mentioned as needed in the text. One of the earliest references to baobab was made by Ibn Batuta in the 14th century, who described it as a food and a large, long-living tree in Africa^1,2. Colloquially, the baobab is referred to as the ‘upside down tree’ since when it loses its leaves, the branches look like roots; in addition, due to the Hollywood blockbuster “The Lion King”, baobab is also referred to as the “Tree of Life”.

Baobab offers various edible parts, i.e., seeds, leaves, roots, flowers, and powdery fruit pulp. The fruit is particularly rich in vitamin C, antioxidants, anti-inflammatory compounds, minerals, and fiber. Beyond its dietary benefits, the bark is used in crafting robes and mats, adding to the economic importance of the baobab tree. Furthermore, the seeds yield oil used in cosmetics³. Baobab seeds contain phytic acids, just like legume seeds; however, proper processing can reduce these acids⁴. The recent approval of baobab as a food ingredient by the European Commission and the United States Food and Drug Administration (FDA) has significantly increased demand for baobab products outside of Africa. The estimated value of baobab products was US$8.2 billion in 2022 and is anticipated to reach US$12.1 billion by 2030⁵. Thus, there is economic interest and social need for genomic resources to study, preserve, and increase baobab yields⁶.

Baobabs are among the oldest and largest non-clonal organisms, living over 2400 years with canopies exceeding 500 m³ and trunks about 35 meters wide⁷. Unlike most large trees, baobabs are succulents, lacking “growth rings” or true wood⁸. Achieving maturity in the wild is challenging due to predation by caterpillars and larger animals⁹. Despite having bisexual flowers, baobabs are mostly self-incompatible and rely on external pollinators like bats, bush babies, and hawkmoths^{10,11,12,13,14,15}. As an obligate outcrosser, A. digitata shows high genetic diversity, seen as heterozygosity in the genome (Fig. 1a, b; Table 1)^11,15. Understanding baobab ploidy is key for breeding, as polyploidy influences growth and adaptability. In the wild, baobabs can take over 100 years to flower^16,17. Additionally, genome size knowledge guides sequencing projects^18,19. Despite thriving in harsh, low-rainfall environments, recent reports show rising baobab mortality^7,20. For instance, the 1400-year-old Chapman baobab died suddenly in 2016²¹, with others like Panke, the oldest, dying in recent years^7,9. Suspected causes include rising temperatures, pathogens, soil compaction, and overexploitation, but limited research and genome resources hinder clear conclusions^{7,16,22,23,24}.

**Fig. 1: Characteristics of baobab genomes.**

Table 1 K-mer-based genome size estimates across the eight Adansonia species

Full size table

The Adansonia genus has eight recognized species, with six species endemic to Madagascar: A. grandidieri, A. perrieri, A. rubrostipa, A. madagascariensis, A. za, and A. suarezensis, while A. digitata and A. gregorii are indigenous to mainland Africa and Australia, respectively^15,20,24. Madagascar is considered the center of origin for baobabs, with ongoing interspecific hybridization when species are in sympatry²⁴. While seven baobab species are widely acknowledged as diploids (2n = 2x = 88)^15,25,26, A. digitata is a tetraploid²⁷. Pettigrew et al. claimed to have found a diploid species in mainland Africa, A. kilima, but this hypothesis has not been supported by subsequent research^15,28. Historical reports for African baobab suggested chromosome numbers of 2n = 96–168, accompanied by genome sizes ranging from 3 to 7 pg 2 C/holoploid^15,25,29. However, the ploidy of A. digitata remains uncertain, with questions about whether it is diploid, autotetraploid, or allotetraploid²⁰.

Here, we report genome size estimates for all eight recognized baobab species using a K-mer-based method from short-read genomic sequences. This method provides independent estimates comparable to flow cytometry^20,25. Our primary objectives are to create a reference genome, explore genomic features, and examine evolutionary history. We generate haploid chromosome-scale assembly of A. digitata (Ad77271a; originally from Tanzania) alongside draft assemblies for sibling tree (Ad77271b), two trees from distinct locations in Africa [Kord Bao (AdKB) from Sudan and Okahao Heritage Tree from Namibia (AdOHT)], and a related species A. za from Madagascar (Aza135). Additionally, we resequence 25 A. digitata trees representing different regions of Africa to assess genetic diversity. Our findings reveal: (a) a high proportion of DNA transposons over LTR retrotransposons, indicating unusual genomic structure in baobab; (b) four DNA Mutator-type transposons possibly contributing to centromere formation; (c) retention of nearly complete chromosomes after whole-genome multiplication 30 million years ago, leading to gene expansion for flower development, chromatin regulation, and exocytosis; (d) expansion of UV RESISTANCE LOCUS 8 (UVR8) genes, suggesting a unique genome protection mechanism in long-lived trees; and (e) genetically distinct Namibian baobab populations, highlighting the need for targeted conservation. These insights could facilitate baobab breeding, conservation, and domestication.

Results

Adansonia species genome sizes and heterozygosity

Historically, Feulgen staining and flow cytometry have been used to estimate genome size³⁰. Cytological methods suggested A. digitata has 42 chromosomes, a haploid genome size of 920 Mb, and a 2C-DNA value of 3.8 pg^25,30. We skim-sequenced all eight recognized baobab species and used K-mer frequency analysis to estimate genome sizes, which range from 646 Mb in A. perrieri to 1.5 Gb in A. grandidieri. Genome sizes of A. grandidieri and A. rubrostipa were double the predicted haploid size, with K-mer peaks indicating potential hybridization, as has been reported between A. rubrostipa and A. za²⁴. However, sampling or methodological artifacts may affect these estimates, requiring further validation. The K-mer-based estimates were largely consistent with the previous data^23,24,25,31. Repeat content varied from 25.8% to 57%, with A. rubrostipa having the highest repeats. Heterozygosity ranged from 1.1% to 1.8%, with A. digitata showing one of the highest levels (1.6%), likely due to its outcrossing nature^11,15 and autotetraploidy²⁷ (Table 1).

Baobab genome assembly, annotation, and ploidy estimation

The GRIN conserves A. digitata trees named “PI77271” (Supplementary Fig. 1), which we chose because it’s publicly accessible for research. Using long-read Oxford Nanopore Technologies (ONT), we sequenced two sibling seedlings, “Ad77271a” and “Ad77271b”, whose genome assemblies produced 1780 and 2430 contigs, respectively (Supplementary Data 1, Supplementary Fig. 2). Ad77271a was scaffolded into 42 chromosomes using Hi-C (Fig. 1c; Fig. 1d), consistent with the haploid chromosome number (2n = 4x = 168) in A. digitata²⁵. The final haploid genome size for Ad77271a had a higher N50 of 15 Mb (Supplementary Data 1). This assembly aligns with two recently published A. digitata genomes: one at 686 Mb using short reads²³ and another at 668 Mb using PacBio CLR, which had an N50 of only 1 Mb²⁴.

Additionally, using ONT reads, we assembled draft genomes for two A. digitata trees: Kord Bao (AdKB) from Sudan and Okahao Heritage Tree (AdOHT) from Namibia, as well as A. za (Aza135) from Madagascar (23°12'44.0“S, 44°02'32.2“E) (Supplementary Data 1 and 2; Supplementary Fig. 3). These assemblies had modest N50 lengths (600–800 kb) compared to Ad77271a and Ad77271b genomes (Supplementary Data 1), but sufficient for analyzing structural variations, nucleotide diversity, and polyploidy (Supplementary Fig. 4). BUSCO analysis showed ~98% completeness for Ad77271a, Ad77271b, and AdKB, with slightly lower scores for AdOHT (94%) and Aza135 (93%) (Supplementary Fig. 3a). The LTR Assembly Index (LAI) further confirmed assembly completeness (Supplementary Fig. 3).

Several chromosome rearrangements were found between Adansonia species, with the predominant differences occurring in the centromere region (Fig. 1e; Supplementary Figs. 5 and 6). These unaligned regions may result from unassembled repeat sequences or reflect true centromere differences, especially between A. digitata and A. za. The inversion count varied from 28 between Ad77271a and Aza135 to 135 between Ad77271b and AdKB. The highest translocation count was between Ad77271b and AdKB (359), followed by AdKB and AdOHT (333) (Supplementary Data 3). These structural differences were observed despite the high nucleotide similarity (Supplementary Fig. 7a), as seen in other autotetraploid genomes³².

The only large (~1 Mb) translocation identified was between A. digitata and Aza135 on chromosome 23, moving approximately 120 genes from near the centromere to positions closer to the telomere (Fig. 1e). Gene ontology (GO) enrichment linked this region to RNA splicing and included genes like HISTONE-FOLD COMPLEX 2 (MHF2) and REPLICATION PROTEIN A 1B (RPA1B), which are involved with DNA replication, repair, recombination, and transcription, as well as the latter being involved in telomere length^33,34,35,36. In Arabidopsis thaliana, loss of the RPA1B increases sensitivity to UV-B light, causing DNA damage and inhibiting root growth³⁵. While telomere protection^37,38 and DNA repair are key in long-lived organisms^39,40, the translocation itself doesn’t directly link to longevity, as A. za can live for 900 years⁴¹. Additionally, gene flow among baobab species indicates ancient introgression followed by WGM in A. digitata²⁴.

K-mer frequency analysis estimated a haploid genome size of around 659 Mb (Table 1) and high heterozygosity (Fig. 1a, b), though the single K-mer peak suggested a homozygous diploid genome. This single peak, often seen in autopolyploids like coast redwood^42,43, led us to test the tetraploid designation of A. digitata²⁷. A heatmap of heterozygous loci coverage showed two peaks, typical of autotetraploids (Fig. 1d; Supplementary Fig. 4). SNP analysis between Ad77271a and Ad77271b identified 3.9 million SNPs (0.65%) (Figshare raw datasets), reflecting low per-haplotype heterozygosity and underscoring the challenge of fully resolving haplotypes. Additional methods will be required to achieve a complete haplotype-resolved autotetraploid assembly, as elucidated in potato³².

Baobab genome organization: centromere, telomere, rDNA, and DNA methylation

The 42 chromosomes of baobab were small, ranging from 9 to 23 Mb. The Hi-C contact map suggested the centromere sequence was highly conserved across the chromosomes, and consistent with this, we found a putative centromere repeat with a base unit of 158 bp, and a higher order repeat (HORs) of 314 and 468 bp (Fig. 2a–e). In general, the centromere arrays assembled well into 1–2 Mb regions that were both metacentric as well as acrocentric (chr12, chr24, chr26, chr28, chr35, chr38, chr39 are acrocentric; Fig. 2f). Unlike the allotetraploid Eragrostis tef, which has distinct centromere repeats per subgenome⁴⁴, we identified one centromere repeat in the A. digitata genome consistent with autotetraploidy. In addition, telomere sequences (AAACCT) were assembled on the ends of most chromosomes, and these telomeres were longer compared to other plants with maximum sequences spanning 30 kb as compared to the model plants Arabidopsis thaliana and Zea mays that have 3–5 kb telomeres (Fig. 2b; Supplementary Data 4)³⁷. In the Ad77271a genome, we assembled a total of 300 Kb of the ribosomal DNA (rDNA) with only one 26S array that included 35 copies on chr38, and one 5S array that included 442 copies on Ad77271a chr01 (Supplementary Table 1).

**Fig. 2: Genomic organization of baobab.**

A consistent marker of age and longevity in animals is DNA methylation^45,46. In plants, the situation is more complex, but it has been shown that DNA methylation is linked to genome stability and epimutation⁴⁷. DNA methylation is sometimes referred to as the fifth base because it is a chemical modification to the cytosine base of DNA that is known in plants to specify cell fate, silence transposable elements, and mark environmental interactions⁴⁸. ONT reads enable the direct detection of 5-methylcytosine (5-mC)⁴⁹. We leveraged these reads to look for global DNA methylation in baobab, finding average levels of 54.74% in Ad77271a, 54.94% in Ad77271b, and a higher level of 62.52% in AdKB (Supplementary Fig. 8). Consistent with previous findings⁵⁰, we observed increased DNA methylation in the putative centromere arrays (Fig. 2c–f). Additionally, we found that TEs exhibit hypermethylation, while genes display hypomethylation (Supplementary Fig. 9). These observations align with expected methylation patterns seen in other angiosperms⁵⁰.

High accumulation of DNA transposons and the potential role of mutators in centromere creation or repositioning

We performed an ab initio TE prediction for Ad77271a, identifying 378,634 TEs spanning 296 Mb (~43% of the genome) (Fig. 3a; Supplementary Data 1). In most plant genomes, long terminal repeat retrotransposons (LTR-RTs) constitute the largest fraction of TEs due to their copy-and-paste mechanism of proliferation, which leads to genome bloating⁵¹. However, in the Ad77271a genome, LTR-RTs comprised only 10% of the TEs complement, while DNA TEs, which proliferate by a cut-and-paste mechanism, made up 33% of the genome (Supplementary Data 1). Among DNA TEs, ‘Mutator-type’ elements were predominant, a pattern observed across all baobab genomes (Fig. 3a; Supplementary Data 5). Moreover, Copia LTR and CACTA elements showed signs of recent bursts. Investigation of TE insertion times using synonymous substitution (Ks) values showed that most TEs were inserted earlier than 10 million years ago (MYA). We also found more recent TE insertions, particularly in the autotetraploid A. digitata, peaking around 2–3 MYA (Fig. 3b; Supplementary Fig. 6).

**Fig. 3: Insights into transposable elements (TEs) in baobab.**

Further analysis of the relationship between TEs and centromeres across the 42 chromosomes of A. digitata revealed that most centromeric regions were flanked by four Mutator elements (TE_00000631, TE_00000845, TE_00000927, and TE_00000967) (Fig. 3c and Supplementary Fig. 5). A secondary analysis of the chromosomes and genes after WGM suggested that Mutator element expansion likely occurred approximately 11 MYA (Fig. 3b). While it is more common for LTR-RTs to nucleate centromere arrays⁵², here the proliferation of the Mutator element is likely associated with the establishment of centromeres. Conversely, LTR-RTs exhibited a high ratio of solo to intact elements in Ad77271a chromosomes (Fig. 3d), indicating that genome-purging mechanisms are adopted to counterbalance the genome size amplification⁵¹.

Gene expansion, contraction, and comparative orthogroup analysis

Baobab is unlike most large, long-lived trees because it is succulent, and when it dies, its “wood” seems to “deflate” or mush; wood soaked in water will completely disintegrate after several days, leaving only fibers that are used as packing materials⁸. We predicted and annotated genes across the baobab genome assemblies to compare against other tree genomes and closely related species like cotton, cacao, and bombax. Leveraging both ab initio and protein homology gene prediction, we estimated an average of 44,000 genes in A. digitata (Ad77271a, Ad7721b, AdKB, and AdOHT), and a slightly higher number in Aza135 (46,890).

Gene family comparisons between baobab and fifteen plants identified a total of 40,312 orthogroups (OG) (Fig. 4a–c). In the Malvaceae, including baobab, cotton, cacao, durio, and cotton tree, alongside representatives from various plant families, identified 3,851 expanded gene families shared between baobab and cacao, 799 gene families expanded in baobab only, and 232 gene families expanded in cacao only (Fig. 4d). Among contracted gene families, 596 were shared between baobab and cacao, with 1465 specific to baobab and 5678 specific to cacao (Fig. 4d). While there were 107 and 283, Ad77271a and Ad77271b specific OG respectively (Supplementary Table 2), the largest shared specific OGs (4626) were baobab-specific (both Ad77271a and Ad77271b), confirming a shared evolution event between the baobab siblings (Fig. 4a). The second largest shared OG was between monocots, while cotton genomes were the third largest shared OG group with 1723 and 1263 respectively (Fig. 4a). Further gene family comparisons between A. digitata (Ad77271a, Ad77271b, AdKB, AdOHT) and A.za (Aza135) with 15 other plant species revealed 93% of genes in orthogroups and 6399 orthogroups with all the twenty species (Supplementary Fig. 10)

**Fig. 4: Orthogroups, gene ontology, longevity pathway, and evolutionary dynamics of African baobab tree.**

Leveraging GO enrichment analysis, we asked which biological categories were specific to baobab. We identified 490 significant (FDR < 0.01) GO terms that could be clustered into three broad categories of metabolic processes (GO:0008152; right top), response to stimulus (GO:0050896; left top), and a more dispersed group that included growth (GO:0040007), chromatin (GO:0006325) immune system process (GO:0002376) and circadian rhythm (GO:0007623) (Fig. 4b; Supplementary Data 6; Supplementary Fig. 11). The third grouping represents genes associated with the long-lived nature of baobab. For instance, there are six baobab UV RESISTANCE LOCUS 8 (UVR8) genes, represented in four OG families, three of which are only found in baobab genomes. UVR8 is a UV-A/B photoreceptor that interacts with the E3 ubiquitin-ligase CONSTITUTIVELY PHOTOMORPHOGENIC1 (COP1) and SUPPRESSOR OF PHYA-105 (SPA) to stabilize and destabilize two central growth-regular transcription factors ELONGATED HYPOCOTYL 5 (HY5) and PHYTOCHROME INTERACTING FACTOR 5 (PIF5) respectively^53,54,55,56 (Fig. 4c). UV-B radiation has the potential to damage macromolecules such as DNA and impair cellular processes, suggesting the baobab UVR8 proteins may play a broad signaling role to protect the genome of this long-lived tree. It has been shown that UVR8 may interact directly with chromatin based on its orthology to REGULATOR OF CHROMATIN CONDENSATION 1 (RCC1) and nucleosome binding assays⁵⁷. Consistent with this, we observe enriched OG for DNA repair (GO:0006281), DNA damage response (GO:0006974), chromatin organization (GO:0006325), and remodeling (GO:0006338), consistent with baobab actively protecting its genome through UVR8-chromatin regulation (Fig. 4c).

Evidence of an ancient WGM event in baobab genome

We performed comparative genomics of the baobab genomes Ad77271a, Ad77271b, AdKB, AdOHT and Aza135 with three closely related Malvaceae species, cotton (G. raimondii), bombax (B. ceiba) and cacao (T. cacao), as well as grape (V. vinifera) that only has one whole-genome triplication (WGT) and amborella (A. trichopoda) that is sister to the eudicot lineage and lacks WGM^58,59. Synteny-based and rates of synonymous substitutions (Ks) were used to estimate WGM as well as to understand its relationship with other species, such as A.za (Aza135). The Ks analysis revealed a consistent timing of the separation of baobab-amborella and baobab-grape at 128 and 96 MYA. In contrast, bombax and Aza135 genomes diverged from Ad77271a around 20 MYA and 17 MYA (Fig. 5a–c)^58,59. Following the split, Aza135 experienced chromosome rearrangements, resulting in smaller chromosomes (Supplementary Fig. 12). The Aza135 genome also showed a second peak around 30 MYA, indicating remnants of WGM. Further alternative analysis using a calibrated gene-tree species-tree confirmed that the WGM for bombax and Aza135 occurred after their divergence from cacao (Supplementary Fig. 13). Additionally, the calibrated time tree revealed that cotton species diverged ~5.0 MYA, congruent with other studies^60,61.

**Fig. 5: Syntenic relationships and the Ks distribution suggest an ancient whole-genome multiplication (WGM) event in baobab.**

The self-self alignments (paralogs) of genomes provided insights into the timing of WGM events. This analysis clarified that all of the baobab, bombax, and cotton genomes experienced WGM around 30 MYA⁶², while cacao underwent its last WGM of approximately 118 MYA, consistent with it only experiencing the WGT-γ event, as previously reported (Fig. 5c)⁶³. Focusing on the baobab complex, we observed a minor Ks peak around 30 MYA across all genomes, along with additional peaks at 4, 6, 11, and 17 MYA for Ad77271b, AdKB, AdOHT, and Aza135 respectively (Supplementary Fig. 6d). We hypothesize that these peaks represent speciation and polyploidization events. The recent WGM in A. digitata resulted in an autotetraploid genome, as further evidenced by minor allele frequency coverage showing two peaks, indicating a chromosomal cluster from the two pairs of homologous chromosomes (four subgenomes) (Fig. 1d). This WGM in A. digitata is estimated to have occurred between 4 AND 11 MYA, with additional evidence provided by a burst of TEs and time-calibrated gene tree (Fig. 3b; Supplementary Fig. 13).

The self-self (paralog) and pairwise (ortholog) alignment of the haploid baobab genomes revealed a 4:4 syntenic depth like the pairwise alignments of cotton and bombax consistent with them sharing the WGM (Fig. 5a, c; Supplementary Fig. 14). In contrast, cacao displayed a 4:1 syntenic depth with the baobab and cotton genomes, and a 5:1 syntenic depth with bombax, confirming it did not share the recent WGM with these genomes as seen by Ks and time tree analysis (Fig. 5a, d; Supplementary Fig. 14). It was postulated that the cotton genome experienced a WGM that was a decaploidization^64,65, and this WGM history has been clarified with a high-quality balsa (Ochroma pyramidale) genome to suggest an allohexaploid lineage hybridized to form an allodecaploid lineage sister to the Bombacoideae and Malvoideae⁶⁶. Recently, a higher quality chromosome-scale bombax genome as well as the closely related kapok (Ceiba pentandra) genome were found to have a syntenic depth of 5:1 with cacao consistent and a shared WGM of about 30 MYA⁶⁷. While we find a syntenic depth of 4:1 between cacao and cotton/baobab, 8.2 and 6.4% of genes are found in a 5:1 syntenic depth, respectively, which is similar to the 14.6% of genes having a syntenic depth of 5:1 between cacao and bombax, suggesting the underlying WGM in baobab could be a decaploidization event⁶⁶.

Plants generally return to diploidy over time through a process called fractionation, which involves gene loss and chromosomal fusions^68,69. However, in all baobab genomes, as well as bombax, kapok, and balsa, the last WGM remains mostly unfractionated, with up to five chromosomes retained compared to one in cacao (Fig. 5a; Supplementary Fig. 15; Supplementary Data 7). In Ad77271a, we found that 15%, 24%, 30%, 13%, and 6% of genes were retained in 1, 2, 3, 4, and 5 copies respectively post-WGM. Since the predominant syntenic depth was 4:1 in baobab, we looked at GO enrichment analysis of quadruplicated genes, which revealed a significant (bonferroni FDR < 0.05) set of overlapping GO terms that were predominantly associated with gene regulation/chromatin, exocytosis, and flower timing/development (Fig. 5e). In contrast, cotton has undergone extensive chromosomal rearrangements, fractionation, and reshuffling, resulting in the loss of the Malvaceae or Malvadendrina protochromosomes⁶⁶.

Across most plant genomes analyzed, circadian, flowering, and light-related genes are reduced back to one or two copies in the genome during fractionation, presumably to ensure the correct gene dosage^68,69,70. However, in Bombacoideae subfamily, with baobab as an example, we found that among circadian, flowering, and light orthologs⁷¹, only six (out of 35) orthogroups had one gene copy, while more than half (18/35) contained three or four copies (Supplementary Data 7; Supplementary Data 8). In contrast to baobab, a grape that has only the WGT-γ and cotton that has both the WGT-γ and WGM have 50% and 75% of the circadian, flowering, and light orthologs respectively, while bombax has retained orthologs similar to baobab (Supplementary Data 7). One pair of genes retained in four copies in the baobab genomes were an evolutionarily conserved syntenic gene pair ZEITLUPE (ZTL) and VERNALIZATION 3 (VIN3), which are involved with flowering time, thermomorphogenesis, photomorphogenesis, and the circadian clock, and are linked in the genome across the eudicot lineage back to amborella (Fig. 5d)⁷⁰. Plants that leverage Crassulacean Acid Metabolism (CAM) photosynthesis to ensure stomata are open at the correct time of day (TOD)^72,73,74,75, as well as the crop soybean that highly specific latitude maturity groups⁷⁶, have retained multiple copies of circadian, light, and growth genes⁷⁰; these results indicate that baobab may leverage CAM photosynthesis and a tightly regulated TOD process for flowering. These genes offer insight into how baobab has adapted to various environments across Africa (see variation analysis below).

Diversity in baobab across Africa

A. digitata is endemic to mainland Africa and little is known about the genetic diversity across the continent. We resequenced 25 A. digitata trees from across Africa using Illumina short reads, yielding an average sequencing depth of 20x per accession (Supplementary Data 2). Mapping these sequences to Ad77271a reference produced 58.9 million SNPs and 446 thousand INDELs, accessible at https://resources.michael.salk.edu/baobab/index.html. Analysis of ploidy using these genotypes confirmed that A. digitata is an autotetraploid (Supplementary Data 9). Principal component analysis (PCA) of 25 accessions revealed that more than 30% of the genetic variance was due to geographical origins, mostly east and north/west (Fig. 6a; Supplementary Data 9). Population 1 encompasses germplasms from north of the equator up to approximately 16 degrees north (and mostly west), while populations 2 and 3 span from the equator to 26 degrees south. The trees from the northern desert region of Namibia, clustered into population 3 (Fig. 6a; top left; green circles), were distinct from the trees collected closer to Botswana (population 2). These results indicate there are geographical or environmental factors that have limited gene flow between these populations.

Fig. 6: Diversity and structural variations in *A. digitata.*

We leveraged the fixation index (Fst) across the three populations to determine their relatedness. The analysis revealed that population 1 and 3 were more related (lower Fst; average <0.1), while populations 2 and 3, as well as population 1 and 2, were less related (higher average Fst > 0.4) (Fig. 6b). These results were consistent with Namibian population 3 being closer to population 1, in contrast to population 2, despite their relative geographic closeness. We looked at the genes with high Fst SNPs (>0.8) across the three population comparisons to identify genes that are under selection for their environment, and only found enriched GO terms between populations 2v3 and 1v3 that could be summarized into pollination, organelle localization, and chromatin (Supplementary Data 10 and 11). Consistent with our findings that baobab-specific genes were related to longevity through the UVR8-chromatin connection, these genes also played a role in an environment-specific fashion. The identification of pollination genes having high Fst suggests there may be some selective pressure on synchronizing flowering and pollination, which is thought to be one of the factors that led to the ‘rise of angiosperms^77,78. Similar to CAM plants that have expanded circadian clock genes to ensure TOD-specific stomatal opening, it has been shown that coordination of flower development and opening must be highly TOD-specific to pollinators⁷⁰. This explains why many circadian clock genes have been retained in three or four copies after fractionation (Fig. 5d; Supplementary Data 7), and we find that many of these genes have high Fst. For instance, the circadian-linked gene pair ZTL and VIN3 that were retained in four copies also had high Fst (Fig. 5d; Supplementary Data 12). Baobab flowers open in a tightly regulated manner, starting to bloom at sunset. They remain in bloom and receptive to pollinators throughout the night but begin to wilt by early the next morning. In East and West Africa, bats have been identified as the primary pollinators, while in Southern Africa, hawkmoths also contribute to pollination^13,79. Taken together, the regulation of chromatin and pollination play important roles in shaping the diverse populations of baobab in Africa.

Variation and selection can be specific to regions of the chromosome; therefore, we employed a local PCA technique to identify patterns of relatedness from SNP frequencies across the genome⁸⁰. Along the chromosomes, we observed striking regions of shared relatedness that contrasted greatly from surrounding regions. One region on chromosome 23 between 6 and 7 Mb overlaps with the predicted centromere arrays (6–9 Mb) (Fig. 1e; Fig. 2f; Fig. 6c; Supplementary Fig. 16). This region is present in Ad77271a, Ad77271b, and AdKB (all East Africa), but lost in AdOHT (Northwest Namibia), suggesting that region could be important for the diversification of baobab across Africa (Fig. 1e). Many of the putative centromere regions in A. digitata had translocation, inversions, and duplications, or regions that didn’t map (Supplementary Fig. 17; Supplementary Data 3), as well as high Fst with GO enrichment of chromatin, consistent with these regions being under active change in the baobab genomes. Baobab has a large number (2n = 4x = 168) of small chromosomes (9–23 Mb), suggesting that selection on chromatin stability and arrangement may play a role in longevity and environmental specificity. In a nutshell, baobab has adapted to specific environments across Africa and achieved extreme longevity by actively protecting and regulating its genetic material while retaining accurate TOD acuity through the retention of circadian, light, and flowering time genes to ensure environment-specific pollination.

Discussion

Baobab, an iconic succulent tree emblematic of Africa’s savannas, bears nutritious fruits fueling its global demand and bolstering income for rural communities across the continent. However, some of the oldest known baobabs are dying across Africa^7,81 which makes it imperative that we better understand the baobab genome to initiate conservation and breeding. Baobab belongs to the Malvaceae family, which comprises 4,225 species across 243 genera and nine subfamilies, including economically and ecologically important plants such as cotton (Gossypium spp.), the cotton tree (Bombax ceiba), cacao (Theobroma cacao), and baobab (A. digitata)⁶⁶. WGM has played a crucial role in the evolution of these plants. For example, following the ancient core eudicot whole-genome triplication (γ) event, cotton plants have undergone repeated polyploidization and hybridization, resulting in polyploid offspring with combined genomes that diverge significantly from their diploid parents, thereby initiating new evolutionary trajectories. The decaploid ancestor of cotton (a 5× multiplication) significantly contributed to the development of spinnable fibers^24,65,66. Similarly, polyploidy in Bombax ceiba has facilitated its robust growth and vibrant blooms in tropical regions⁶⁷, while ancient WGM in cacao has led to the emergence of diverse and flavorful varieties. Here, we demonstrate that baobab has a unique proportion of DNA transposons, experienced WGM, and presents a high-quality chromosome-level haploid assembly of A. digitata (Ad77271a) alongside draft genomes from a sibling tree (Ad77271b), two geographically diverse A. digitata (AdOHT and AdKB) and distinct species A. za (Aza135) from Madagascar. We also show a distinct baobab population in Namibia.

Gene dosage and gene balance hypotheses are well-established concepts in plant genetics, emphasizing the conservation of the functionality of specific gene pathways^68,82,83. Baobab has undergone several rounds of WGM starting with the WGT-γ followed by what is thought to be a decaploidization^64,65,66,67, the latter of which has remained largely unfractionated with between four and five full chromosomes retained as compared to cacao (Fig. 5a; Supplementary Figs. 14g and 18). Consequently, baobab has retained evolutionarily conserved linkages between the circadian/temperature genes LHY/PRR9 (3 copies) and ZTL/VIN3 (4 copies) (Fig. 5d; Supplementary Data 7)^70,84,85, similar to patterns observed in soybean (Glycine max)^70,86, which provide environmental specificity and are correlated with maturity⁷⁶. Our analysis also identified unique UV RESISTANCE LOCUS 8 (UVR8) genes, suggesting mechanisms for UV-B light reception and DNA repair, crucial for baobab’s adaptation to arid conditions. Upon UV-B absorption, UVR8 undergoes monomerization, leading to a structural change that initiates downstream signaling events through the circadian clock and CONSTITUTIVELY PHOTOMORPHOGENIC1 (COP1) to regulate DNA damage response and repair^53,87. Further, genes with a high fixation index (Fst) showed enriched terms in flower development, regulation/chromatin, immunity, and exocytosis (Supplemental Fig. 11; Supplementary Data 6). We note that Fst is a population genetic estimation based on a diploid model. Thus we interpret our polyploid Fst results cautiously⁸⁸. Baobab flowers bloom late in the afternoon and wilt within 24 h, indicating tightly controlled timing for flower development, opening, and pollinator schedules^{70,84,85,89,90}. These findings posit that the retention of gene copies in baobab, especially in circadian, light, and flowering time pathways, contributes to its adaptation and longevity.

This study further confirms baobab’s autotetraploid nature through analysis of the sibling genome (Ad77271b) and the 25 resequenced trees (Figs. 1d and 3; Supplementary Fig. 5; Supplementary Data 9). Initially, the K-mer analysis profiled a single peak with an estimated haploid genome assembly size of 659 Mb, which is in line with what was found through Feulgen staining and flow cytometry²⁵, suggesting a diploid genome. However, since chromosome counting indicated that it is tetraploid²⁵, we looked more closely at the allele frequencies. Leveraging both the Ad77271b and the resequenced 25 tree reads, we analyzed coverage at heterozygous sites, which provided evidence that A. digitata possesses an autotetraploid genome. In addition, we identified only one centromeric repeat (base unit = 158 bp) supporting autotetraploidy, since allotetraploids like Eragrostis tef have been shown to have distinct centromere arrays representing the different subgenome origins⁴⁴. Finally, our Ks analysis found a recent WGM 3–11 MYA consistent with an autotetraploidy event. However, it is possible that autotetraploidy didn’t immediately lead to disomic inheritance, instead, there may have been a long period of tetrasomic inheritances that could still be happening now, as has been seen in the coast redwoods⁴².

Our study also revealed baobab genome is unlike any published plant genome to date in that DNA TEs are dominant (3x) compared to LTR-RT that typically result in the bloating of plant genomes⁵¹. Compared to other plant genomes, baobab has an average total TE content of A. digitata (~45%) having more than A. za (35%), which was also in line with the 50-60% DNA methylation levels (Supplementary Data 1; Supplementary Fig. 8). The A. digitata TE composition was unusual compared to other plant genomes in that the proportion of DNA TEs was 33% while the LTR-RTs were 10% (Supplementary Data 1). Typically, LTR-RTs are the predominant transposon in a plant genome since they proliferate in a “copy-and-paste” mechanism, while in contrast, DNA TEs accumulate through a “cut-and-paste” mechanism⁵¹. In terms of LTR-RT TEs, a big rise in Copia and CACTA TEs was seen around 10 MYA, which sets baobabs apart from its relative in the Malvaceae family, cacao^59,63. In contrast, the DNA TEs burst around 3–11 MYA, coinciding with the putative tetraploidy event in baobab, suggesting that the tetraploidy event may have played a role in the accumulation of the cut-and-paste DNA TEs that have shaped the baobab genome. This unusual genomic composition raises the possibility that DNA transposons, particularly the four sets of mutator-type elements (Fig. 3c), play roles in recruiting centromeric histones and potentially contributing to new centromere formation, suggesting an adaptive mechanism⁹¹.

Finally, the resequencing of 25 A. digitata accessions from across Africa revealed that Adansonia trees situated north of the equator exhibited some divergence from those located southward, with approximately 6% of the variation attributed to this distinction. However, the most partitioning, about 30% of the variation, occurred along an east-west axis, as exemplified by the distinctiveness between populations 1 and 2 (Fig. 6; Supplementary Fig. 7). Intriguingly, Namibian baobab trees clustered into two populations (2 and 3), which correspond to different watersheds, suggesting limited gene flow due to possible geographic barriers and adaptation to distinct climate regimes³¹. Another study conducted in Niger and Mali supported this notion, indicating variations in baobab species across the continent; specifically, it found that West African germplasms exhibited faster growth and better adaptation to arid environments compared to their East African counterparts⁹². A recent study on the evolutionary history of baobab trees suggests that a diploid progenitor dispersed from Madagascar to Africa, where autopolyploidy occurred in West Africa, creating the current tetraploid A. digitata. This neo-polyploidy, with its ecological advantages, likely contributed to A. digitata’s spread eastwards and southwards²⁴. Additionally, the observed complex collinearity between Ad77271a from mainland Africa and Aza135 from Madagascar underscores the extent of chromosomal restructuring and divergence within the species (Supplementary Fig. 12).

In summary, this work illuminates the evolutionary history of baobab, confirming A. digitata as an autotetraploid (2n = 4x = 168) with 42 chromosomes (Fig. 1c). The retention of multiple copies of circadian, light, and flowering-related genes following WGM highlights their roles in baobab’s strategies for longevity and pollination⁷⁰. The identification of UVR8 as key for stress resilience opens new avenues for breeding programs to enhance adaptation. The baobab genome’s high proportion of DNA transposons, compared to LTR retrotransposons, is unique among plant genomes, highlighting its unusual structure. Additionally, the identification of specific Mutator transposons (TE_00000631, TE_00000845, TE_00000927, and TE_00000967) suggests they may play a role in centromere formation or repositioning. The isolation of genetically diverse baobab populations in Namibia underscores the need for targeted conservation efforts. These findings highlight the key role of genomics in conserving baobab, advancing domestication, and optimizing agricultural potential^60,93,94. Our study also provides a detailed view of the genome’s structural organization, including centromeres, telomeres, ribosomal DNA, and patterns of DNA methylation.

Methods

Plant materials

Seeds were obtained from the USDA Germplasm Information Resource Network (GRIN) from three trees grown in the USDA-Agriculture Research Service, Subtropical Horticulture Research Station, Miami, FL, USA, under the accession number PI77271. The original seed came from Dar es Salaam, Tanganyika Territory, Tanzania, Africa, in 1928. The PI77271 tree was chosen to generate the reference genome to enable broad access to the baobab germplasm through GRIN. Seed for the Adansonia species to estimate genome sizes was ordered from Le Jardin Naturel (https://www.baobabs.com/). N.K. and E.H.E.K. (co-authors of this paper) collected leaf and seed material for the resequencing of 25 baobab species. The seed was cleaned, soaked in boiling water for three days, and then planted in well-drained soil. Seedlings were grown to the first true leaf stage and then dark-adapted for two days for DNA and RNA extraction (Supplementary Fig. 1a).

DNA and RNA extraction

Baobab has been a difficult species to extract high molecular weight (HMW) DNA due to the large amounts of polysaccharides that it produces. Therefore, we employed two different methods to obtain HMW DNA from baobab: first, seedlings at the two true leaf stages were used for DNA extraction, and second, they were dark-adapted for two days to deplete the polysaccharides. After two days of dark adaptation, two PI77271 seedlings were chosen for genome sequencing and named “Ad77271a” and “Ad77271b”; These sibling seedlings were chosen for sequencing to enable analysis of reported autotetraploidy. HMW DNA was extracted from Ad77271a and Ad77271b, as well as “AdOHT” from Namibia and “AdKB” from Sudan, along with A. za (Aza135) from Madagascar using a modified protocol⁹⁵ (Supplementary Fig. 2). For the 25 A. digitata resequencing, DNA was extracted from dried leaf samples⁶⁰. For RNA, total RNA from young-fresh leaves was extracted using the QIAGEN RNeasy Plus Kit according to the manufacturer’s instructions. The RNA was quantified using the Qubit RNA Assay and TapeStation 4200. Before library preparation, DNase treatment was performed, followed by AMPure bead cleanup and rRNA depletion with QIAGEN FastSelect HMR.

Genome and transcriptome sequencing

Unsheared HMW DNA (1.5 µg) from Ad77271a, Ad77271b, AdOHT, AdKB and Aza135 was used for ONT ligation-based libraries (SQK-LSK109). Final libraries were loaded on an ONT flowcell (v9.4.1) and run on the PromethION. Illumina 2 × 150 paired-end reads were also generated for genome size estimates and polishing genome sequences. Libraries were prepared from HMW DNA using Illumina NexteraXT library prep kit and sequenced on NextSeq High Output 300 cycle using Illumina 2 × 150 paired-end kit (Illumina, San Diego, CA). The resulting raw sequence was only trimmed for adapters, resulting in >60× coverage. For the transcriptome, NovaSeq6000 platform with a 2 × 150 bp configuration was utilized.

Hi-C library preparation

Hi-C library was prepared for Ad77271a using Phase Proximo Hi-C (Plant) kit (V.3.0) and run on Illumina NextSeq P3 300 cycle, paired-end 2 × 150 kit.

Genome assembly analysis

ONT reads were assembled using Flye (v2.9.2)⁹⁶ and then polished using Racon (v1.5.0)⁹⁷ and Pilon (v1.24)⁹⁸ with Illumina reads. Hi-C data and Juicer version 1.6.2, 3ddna (v180419), and JBAT (v1.11.08) built the final assembly. The completeness of the genome assembly was assessed through BUSCO (v. 5.4.3), utilizing the ODB10 eudicots dataset⁹⁹.

K-mer-based genome size estimates

Applying K-mer-based techniques to Illumina short reads from Illumina sequencing libraries allowed us to estimate the genome’s size, repeat, and heterozygosity. Jellyfish in combination with GenomeScope2¹⁰⁰ were employed to assess parameters such as haploid genome length, repeat content, and heterozygosity. For analysis of ploidy, nQuire Tool¹⁰¹ in conjunction with statistics of variants from Illumina short reads were utilized (Supplementary Data 9).

Scaffolding long-read assembly contigs

Ragtag (v2.1.0) was used to scaffold the contigs of Ad77271b, AdOHT, AdKB, and Aza135 with the Hi-C scaffolded Ad77271a.

Repeats and gene prediction

EDTA (v1.9.6)¹⁰² was used to construct a repeat library and softmask complex repeats. Tandem Repeats Finder (v4.09)¹⁰³ was employed to identify centromere and telomere sequences, as well as softmask simple repeats. The ONT cDNA library was mapped against the reference using Minimap2 with splice presets and then assembled using Stringtie (v2.2.1) with the long reads processing flag. Genes in baobab were predicted via the Funannotate (v1.8.2; https://github.com/nextgenusfs/funannotate) pipeline with the Stringtie transcript models included as evidence in addition to the UniProt protein set. Predicted proteins were characterized using Eggnog-mapper v2.0.1¹⁰⁴.

Solo long terminal repeat (LTR) and intact LTR retrotransposons

We used the EDTA (v1.9.6) 98 pipeline as described by Lynch et al.¹⁰⁵. Briefly, solo-LTRs were identified by first collecting the set of LTRs that were not assigned as intact LTR-RTs, which are retrieved on the basis of “method = homology” in the attribute column of the TEanno.gff3 file. We applied thresholds to isolate solo-LTRs from truncated and intact LTRs, as well as internal sequences of LTR-RTs. These thresholds include a minimum sequence length of 100 bp, 0.8 identity relative to the reference LTR, and a minimum alignment score of 300. We also required that the four adjacent LTR-RT annotations not to have the same LTR-RT ID. Further, we required a minimum distance of 5000 bp to the nearest adjacent solo-LTR, intact LTR, or internal sequence. Last, we kept solo-LTR sequences that fell within the 95th percentile for LTR lengths. LTR insertion time was estimated using a substitution rate of 4.72 × 10⁻⁹ per year¹⁰⁶.

DNA methylation analysis

LoReMe (Long Read Methylation) (Oxford Nanopore Technologies · GitHub) was used to infer DNA methylation patterns. In brief, the process involved the conversion of ONT FAST5 data to POD5 format using the Loreme Dorado-convert tool v.0.3.1. Subsequently, super-high-accuracy base calling was performed, aligning the sequences to the reference genome of A. digitata (Ad77271a). Modkit v0.1.11 was employed to generate a bed file containing comprehensive methylation data, enabling us to create visual representations of methylation profiles for further investigation and interpretation.

Orthology analysis

OrthoFinder (v2.5.5)¹⁰⁷ was used for comparative genomics of Malvaceae: A. digitata (Ad77271a and Ad77271b), cotton (Gossypium raimondii and Gossypium hirsutum), cacao (Theobroma cacao), durio (Durio Zibethinus) and cotton tree (Bombax ceiba). Additionally, we examined representatives from Vitaceae (Vitis vinifera), Brassicaceae (Arabidopsis thaliana) Salicaceae (Populus trichocarpa), Fagaceae (Quercus rubra), Myrtaceae (Syzygium grande), Crassulaceae (Kalanchoe fedtschenkoi), Acoraceae (tetraploid Acorus calamus and diploid Acorus gramineus), Amborellaceae (Amborella trichopoda) and Ginkgoaceae (Ginkgo biloba). Except for the baobab, the primary proteins were downloaded from phytozome (v13)¹⁰⁸ and websites (Supplementary Table 2). Gene family size in the context of phylogeny was analyzed using CAFE v5.0¹⁰⁹. Orthogroups with lots of genes in one or more species (100 genes) and only present in one species were excluded¹¹⁰. Results were then visualized using CafePlotter (https://github.com/moshi4/CafePlotter).

Synteny analysis

CoGe was used to make syntenic region dot plots for intergenomic and intragenomic alignments (https://genomevolution.org/CoGe/). MCscan (https://github.com/tanghaibao/jcvi/wiki/MCscan; Python version) was used for interspecies syntenic analysis, enabling the identification of homologous gene pairs, gene blocks, and the creation of syntenic plots that depict the relationships between homologous gene pairs between baobab and other species.

Structural variation and rearrangement identification

Structural variations (SVs) were profiled using Syri v1.6.3¹¹¹.

Variants analysis

In order to compare the sibling baobabs, we performed short variant calling and structural variant calling using two distinct pipelines: a short read (Illumina) based small variant calling workflow, and a long read (i.e. ONT) based structural variant calling workflow. Short reads were used for small variant calling due to their high accuracy at the nucleotide level, permitting high confidence SNP and short indel calls. Long reads were used for structural variant calling because their increased length allows for covering large structural variants and verifying their structure. For each of our sibling baobabs, we ran both variants calling pipelines using both baobabs as a method of sanity-checking our results. Results were consistently symmetric regardless of which of the two baobabs was used as a reference.

Our short-read-based short variant calling pipeline involves four primary steps: read trimming, read alignment, variant calling, and quality filtering. Read trimming is helpful for filtering out low-quality portions of reads that could reduce variant calling accuracy and run time as erroneous base calls when aligned to the reference have to be processed as potential variants. Read trimming was conducted using Trim Galore (v0.6.6). Read alignment was conducted using minimap2 (v2.20) with short-read-appropriate settings and was then sorted using Samtools (v1.12). Each sample was mapped independently and then processed by freebayes (v1.3.5) collectively using tetraploid settings in order to call variants. Subsequent variant calls were then filtered to Q20 before being manually inspected and summarized with several stats tools: vcftools (v0.1.16) and rtg tools (v3.12). We also used an in-house developed stats tool (available from https://gitlab.com/NolanHartwick/bio_utils), which includes functionality to process coverage stats as output by freebayes in order to verify ploidy.

Molecular dating and whole-genome multiplication (WGM)

A total of 492 single-copy orthologues, 365,810 genes, and 31,169 orthogroups from nine species: T. cacao, B. ceiba, G. arboreum, G. raimondii, A.za (Aza135), A. digitata (Ad77271a), Carica papaya, Arabidopsis thaliana and Oryza sativa were used in the analysis. Secondary calibration of molecular clocks was performed using priors from the http://www.timetree.org/ database for monocot (Oryza sativa) and dicot (Arabidopsis thaliana) (163 MYA). For alternative WGM prediction, synonymous substitution rates (Ks values) were calculated using the MCScan software (Python version) with default parameters. A substitution rate of 6.56 × 10⁻⁹ was then utilized to estimate the time of duplication events¹¹².

Ultraviolet-B radiation (UV-B) photoreceptor (UVR8) gene analysis

Baobab-specific orthologs were subjected to gene ontology enrichment (GO) analysis using Python GOATOOLS¹¹³, and subsequently visualized using REVIGO¹¹⁴. The phylogenetic tree was analyzed via GeomeNet (https://www.genome.jp/en/about.html).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The genome assemblies and annotations for Ad77271a, Ad77271b, AdOHT, AdKB, and Aza135 are available through the Michael Lab genome portal [https://resources.michael.salk.edu/baobab/index.html], and are also uploaded to CoGe under ID 67790 – 67801. The raw sequencing reads can be accessed via BioProject 1022505. Source data are provided with this paper and via Figshare [https://doi.org/10.6084/m9.figshare.26039878]¹¹⁵.

References

Gibb, H. A. R. & Beckingham, C. F. The Travels of Ibn Battuta, AD 1325–1354 (Routledge, 2017).
Baum, D. A. A systematic revision of Adansonia (Bombacaceae). Ann. Mo. Bot. Gard. 82, 440–470 (1995).
Article Google Scholar
Asogwa, I. S., Ibrahim, A. N. & Agbaka, J. I. African baobab: its role in enhancing nutrition, health, and the environment. Trees For. People 3, 100043 (2021).
Article Google Scholar
Silva, V. M., Putti, F. F., White, P. J. & Reis, A. R. D. Phytic acid accumulation in plants: biosynthesis pathway regulation and role in human diet. Plant Physiol. Biochem. 164, 132–146 (2021).
Article PubMed CAS Google Scholar
Research and Markets. Baobab powder - global strategic business report. https://www.researchandmarkets.com/reports/5029822/baobab-powder-global-strategic-business-report (2024).
Offiah, V. O. & Falade, K. O. Potentials of baobab in food systems. Appl. Food Res. 3, 100299 (2023).
Article CAS Google Scholar
Patrut, A. et al. The demise of the largest and oldest African baobabs. Nat. Plants 4, 423–426 (2018).
Article PubMed Google Scholar
Gebauer, J. et al. Africa’s wooden elephant: the baobab tree (Adansonia digitata L.) in Sudan and Kenya: a review. Genet. Resour. Crop Evol. 63, 377–399 (2016).
Article MathSciNet Google Scholar
Venter, S. M. & Witkowski, E. T. F. Where are the young baobabs? Factors affecting regeneration of Adansonia digitata L. in a communally managed region of southern Africa. J. Arid Environ. 92, 1–13 (2013).
Article ADS Google Scholar
Venter, S. M. et al. Baobabs (Adansonia digitata L.) are self-incompatible and ‘male’ trees can produce fruit if hand-pollinated. S. Afr. J. Bot. 109, 263–268 (2017).
Article Google Scholar
Karimi, N. et al. Evidence for hawkmoth pollination in the chiropterophilous African baobab (Adansonia digitata). Biotropica 54, 113–124 (2022).
Article Google Scholar
Start, A. N. Pollination of the baobab (Adansonia digitata L.) by the fruit bat Rousettus aegyptiacus E. Geoffroy. Afr. J. Ecol. 10, 71–72 (1972).
Taylor, P. J., Vise, C., Krishnamoorthy, M. A., Kingston, T. & Venter, S. Citizen science confirms the rarity of fruit bat pollination of baobab (Adansonia digitata) flowers in Southern Africa. Diversity 12, 106 (2020).
Article Google Scholar
Coe, M. J. & Isaac, F. M. Pollination of the baobab (Adansonia digitata L.) by the lesser bush baby (Galago crassicaudatus E. Geoffroy). East Afr. Wildl. J. 3, 123–124 (1965).
Article Google Scholar
Cron, G. V. et al. One African baobab species or two? Synonymy of Adansonia kilima and A. digitata. Taxon 65, 1037–1049 (2016).
Article Google Scholar
Patrut, A. et al. Radiocarbon dating of two old African baobabs from India. PLoS ONE 15, e0227352 (2020).
Article PubMed PubMed Central CAS Google Scholar
Swanepoel, C. M. Notes and records baobab phenology and growth in the Zambezi Valley, Zimbabwe. Afr. J. Ecol. 31, 84–86 (1993).
Article Google Scholar
Kitony, J. K. Nested association mapping population in crops: current status and future prospects. J. Crop Sci. Biotechnol. 26, 1–12 (2022).
Article Google Scholar
Levin, D. A. The Role of Chromosomal Change in Plant Evolution (Oxford Univ. Press, Oxford, 2002).
Wickens, G. E. The Baobabs: Pachycauls of Africa, Madagascar and Australia (Springer Science & Business Media, 2008).
Chan, E. K. F. et al. Human origins in a southern African palaeo-wetland and first migrations. Nature 575, 185–189 (2019).
Article ADS PubMed CAS Google Scholar
Sanchez, A. C., Osborne, P. E. & Haq, N. Climate change and the African baobab (Adansonia digitata L.): the need for better conservation strategies. Afr. J. Ecol. 49, 234–245 (2011).
Article Google Scholar
Woods, S., O’Neill, K. & Pirro, S. The complete genome sequence of (Malvaceae, Malvales), the African baobab. Biodivers. Genomes https://doi.org/10.56179/001c.72789 (2023).
Wan, J.-N. et al. The rise of baobab trees in Madagascar. Nature 629, 1091–1099 (2024).
Article ADS PubMed PubMed Central CAS Google Scholar
Islam-Faridi, N., Sakhanokho, H. F. & Dana Nelson, C. New chromosome number and cyto-molecular characterization of the African Baobab (Adansonia digitata L.) - ‘The Tree of Life’. Sci. Rep. 10, 13174 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Costa, L., Oliveira, Á., Carvalho-Sobrinho, J. & Souza, G. Comparative cytomolecular analyses reveal karyotype variability related to biogeographic and species richness patterns in Bombacoideae (Malvaceae). Plant Syst. Evol. 303, 1131–1144 (2017).
Article CAS Google Scholar
Baum, D. A. & Oginuma, K. A review of chromosome numbers in Bombacaceae with new counts for Adansonia. Taxon 43, 1 (1994).
Article Google Scholar
Pettigrew FRS, J. D. et al. Morphology, ploidy and molecular phylogenetics reveal a new diploid species from Africa in the baobab genus Adansonia (Malvaceae: Bombacoideae). Taxon 61, 1240–1250 (2012).
Bennett, M. D. & Leitch, I. J. Nuclear DNA amounts in angiosperms: targets, trends and tomorrow. Ann. Bot. 107, 467–590 (2011).
Article PubMed PubMed Central CAS Google Scholar
Henniges, M. C. et al. The plant DNA C-values database: a one-stop shop for plant genome size data. Methods Mol. Biol. 2703, 111–122 (2023).
Article PubMed CAS Google Scholar
Leong Pock Tsy, J.-M. et al. Nuclear microsatellite variation in Malagasy baobabs (Adansonia, Bombacoideae, Malvaceae) reveals past hybridization and introgression. Ann. Bot. 112, 1759–1773 (2013).
Article PubMed PubMed Central CAS Google Scholar
Sun, H. et al. Chromosome-scale and haplotype-resolved genome assembly of a tetraploid potato cultivar. Nat. Genet. 54, 342–348 (2022).
Article PubMed PubMed Central CAS Google Scholar
Aklilu, B. B. et al. Functional diversification of replication protein A paralogs and telomere length maintenance in Arabidopsis. Genetics 215, 989–1002 (2020).
Article PubMed PubMed Central CAS Google Scholar
Aklilu, B. B., Soderquist, R. S. & Culligan, K. M. Genetic analysis of the replication protein A large subunit family in Arabidopsis reveals unique and overlapping roles in DNA repair, meiosis and DNA replication. Nucleic Acids Res. 42, 3104–3118 (2014).
Article PubMed CAS Google Scholar
Ishibashi, T. et al. Two types of replication protein A in seed plants. FEBS J. 272, 3270–3281 (2005).
Article PubMed CAS Google Scholar
Takashi, Y., Kobayashi, Y., Tanaka, K. & Tamura, K. Arabidopsis replication protein A 70a is required for DNA damage response and telomere length homeostasis. Plant Cell Physiol. 50, 1965–1976 (2009).
Article PubMed CAS Google Scholar
Colt, K. et al. Telomere length in plants estimated with long read sequencing. Preprint at bioRxiv https://doi.org/10.1101/2024.03.27.586973 (2024).
Whittemore, K., Vera, E., Martínez-Nevado, E., Sanpera, C. & Blasco, M. A. Telomere shortening rate predicts species life span. Proc. Natl Acad. Sci. USA 116, 15122–15127 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Han, Y. et al. Chromosome-level genome assembly of Welwitschia mirabilis, a unique Namib Desert species. Mol. Ecol. Resour. 22, 391–403 (2022).
Article PubMed CAS Google Scholar
Wan, T. et al. The Welwitschia genome reveals a unique biology underpinning extreme longevity in deserts. Nat. Commun. 12, 4247 (2021).
Article ADS PubMed PubMed Central CAS Google Scholar
Patrut, A. et al. AMS radiocarbon dating of large Za baobabs (Adansonia za) of Madagascar. PLoS ONE 11, e0146977 (2016).
Article PubMed PubMed Central Google Scholar
Scott, A. D., Stenz, N. W. M., Ingvarsson, P. K. & Baum, D. A. Whole genome duplication in coast redwood (Sequoia sempervirens) and its implications for explaining the rarity of polyploidy in conifers. N. Phytol. 211, 186–193 (2016).
Article CAS Google Scholar
Ernst, E. et al. The genomes and epigenomes of aquatic plants (Lemnaceae) promote triploid hybridization and clonal reproduction. Preprint at bioRxiv https://doi.org/10.1101/2023.08.02.551673 (2023).
VanBuren, R. et al. Exceptional subgenome stability and functional divergence in the allotetraploid Ethiopian cereal teff. Nat. Commun. 11, 884 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Bell, C. G. et al. DNA methylation aging clocks: challenges and recommendations. Genome Biol. 20, 249 (2019).
Article PubMed PubMed Central Google Scholar
Wilkinson, G. S. et al. DNA methylation predicts age and provides insight into exceptional longevity of bats. Nat. Commun. 12, 1615 (2021).
Article ADS PubMed PubMed Central CAS Google Scholar
Mira, S., Pirredda, M., Martín-Sánchez, M., Marchessi, J. E. & Martín, C. DNA methylation and integrity in aged seeds and regenerated plants. Seed Sci. Res. 30, 92–100 (2020).
Article CAS Google Scholar
Gallego-Bartolomé, J. DNA methylation in plants: mechanisms and tools for targeted manipulation. N. Phytol. 227, 38–44 (2020).
Article Google Scholar
Naish, M. et al. The genetic and epigenetic landscape of the centromeres. Science 374, eabi7489 (2021).
Article PubMed PubMed Central Google Scholar
Niederhuth, C. E. et al. Widespread natural variation of DNA methylation within angiosperms. Genome Biol. 17, 194 (2016).
Article PubMed PubMed Central Google Scholar
Michael, T. P. Plant genome size variation: bloating and purging DNA. Brief. Funct. Genom. 13, 308–317 (2014).
Article CAS Google Scholar
Comai, L., Maheshwari, S. & Marimuthu, M. P. A. Plant centromeres. Curr. Opin. Plant Biol. 36, 158–167 (2017).
Article PubMed CAS Google Scholar
Tilbrook, K. et al. The UVR8 UV-B photoreceptor: perception, signaling and response. Arabidopsis Book 11, e0164 (2013).
Article PubMed PubMed Central Google Scholar
Tossi, V. E. et al. Beyond Arabidopsis: differential UV-B response mediated by UVR8 in diverse species. Front. Plant Sci. 10, 780 (2019).
Article PubMed PubMed Central Google Scholar
Liu, W. et al. Phosphorylation of Arabidopsis UVR8 photoreceptor modulates protein interactions and responses to UV-B radiation. Nat. Commun. 15, 1221 (2024).
Article ADS PubMed PubMed Central CAS Google Scholar
Jenkins, G. I. The UV-B photoreceptor UVR8: from structure to physiology. Plant Cell 26, 21–37 (2014).
Article PubMed PubMed Central CAS Google Scholar
Bourbousse, C., Barneche, F. & Laloi, C. Plant chromatin catches the sun. Front. Plant Sci. 10, 1728 (2019).
Article PubMed Google Scholar
Amborella Genome Project. The Amborella genome and the evolution of flowering plants. Science 342, 1241089 (2013).
Article Google Scholar
Jaillon, O. et al. The grapevine genome sequence suggests ancestral hexaploidization in major angiosperm phyla. Nature 449, 463–467 (2007).
Article ADS PubMed CAS Google Scholar
Karimi, N. et al. Reticulate evolution helps explain apparent homoplasy in floral biology and pollination in baobabs (Adansonia; Bombacoideae; Malvaceae). Syst. Biol. 69, 462–478 (2020).
Article PubMed Google Scholar
Wang, M. et al. Comparative genome analyses highlight transposon-mediated genome expansion and the evolutionary architecture of 3D genomic folding in cotton. Mol. Biol. Evol. 38, 3621–3636 (2021).
Article PubMed PubMed Central CAS Google Scholar
Conover, J. L. et al. A Malvaceae mystery: a mallow maelstrom of genome multiplications and maybe misleading methods? J. Integr. Plant Biol. 61, 12–31 (2019).
Article PubMed Google Scholar
Argout, X. et al. The genome of Theobroma cacao. Nat. Genet. 43, 101–108 (2010).
Article PubMed Google Scholar
Wang, X. et al. Comparative genomic de-convolution of the cotton genome revealed a decaploid ancestor and widespread chromosomal fractionation. N. Phytol. 209, 1252–1263 (2016).
Article CAS Google Scholar
Paterson, A. H. et al. Repeated polyploidization of Gossypium genomes and the evolution of spinnable cotton fibres. Nature 492, 423–427 (2012).
Article ADS PubMed CAS Google Scholar
Sun, P. et al. Subgenome-aware analyses reveal the genomic consequences of ancient allopolyploid hybridizations throughout the cotton family. Proc. Natl Acad. Sci. USA 121, e2313921121 (2024).
Article PubMed PubMed Central CAS Google Scholar
Shao, L. et al. High-quality genomes of Bombax ceiba and Ceiba pentandra provide insights into the evolution of Malvaceae species and differences in their natural fiber development. Plant Commun. 5, 100832 (2024).
Article PubMed PubMed Central CAS Google Scholar
Cheng, F. et al. Gene retention, fractionation and subgenome differences in polyploid plants. Nat. Plants 4, 258–268 (2018).
Article PubMed CAS Google Scholar
Raju, S. K. K. Gene dosage balance immediately following whole-genome duplication in Arabidopsis. Plant cell 32, 1344–1345 (2020).
Article PubMed PubMed Central Google Scholar
Michael, T. P. Core circadian clock and light signaling genes brought into genetic linkage across the green lineage. Plant Physiol. 190, 1037–1056 (2022).
Article PubMed PubMed Central CAS Google Scholar
Lou, P. et al. Preferential retention of circadian clock genes during diploidization following whole genome triplication in Brassica rapa. Plant Cell 24, 2415–2426 (2012).
Article PubMed PubMed Central CAS Google Scholar
Wickell, D. et al. Underwater CAM photosynthesis elucidated by Isoetes genome. Nat. Commun. 12, 6348 (2021).
Article ADS PubMed PubMed Central CAS Google Scholar
Yang, X. et al. The Kalanchoë genome provides insights into convergent evolution and building blocks of crassulacean acid metabolism. Nat. Commun. 8, 1–15 (2017).
Article ADS Google Scholar
Wai, C. M. et al. Time of day and network reprogramming during drought induced CAM photosynthesis in Sedum album. PLoS Genet. 15, e1008209 (2019).
Article PubMed PubMed Central CAS Google Scholar
Ming, R. et al. The pineapple genome and the evolution of CAM photosynthesis. Nat. Genet. 47, 1435–1442 (2015).
Article PubMed PubMed Central CAS Google Scholar
Greenham, K. et al. Geographic variation of plant circadian clock function in natural and agricultural settings. J. Biol. Rhythms 32, 26–34 (2017).
Article PubMed Google Scholar
Condamine, F. L., Silvestro, D., Koppelhus, E. B. & Antonelli, A. The rise of angiosperms pushed conifers to decline during global cooling. Proc. Natl Acad. Sci. USA 117, 28867–28875 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Soltis, P. S., Folk, R. A. & Soltis, D. E. Darwin review: angiosperm phylogeny and evolutionary radiations. Proc. R. Soc. B. 286, 20190099 (2019).
Article PubMed Central Google Scholar
Chetty, A., Glennon, K. L., Venter, S. M., Cron, G. V. & Witkowski, E. T. F. Reproductive ecology of the African baobab: floral features differ among individuals with different fruit production. Ecol. Manag. 489, 119077 (2021).
Article Google Scholar
Li, H. & Ralph, P. Local PCA shows how the effect of population structure differs along the genome. Genetics 211, 289–304 (2019).
Article PubMed CAS Google Scholar
Wild, S. Africa’s majestic baobab trees are mysteriously dying. Nature 558, 529–529 (2018).
Google Scholar
Birchler, J. A. & Veitia, R. A. Gene balance hypothesis: connecting issues of dosage sensitivity across biological disciplines. Proc. Natl Acad. Sci. USA 109, 14746–14753 (2012).
Article ADS PubMed PubMed Central CAS Google Scholar
Feng, X. et al. Genomic evidence for rediploidization and adaptive evolution following the whole-genome triplication. Nat. Commun. 15, 1635 (2024).
Article ADS PubMed PubMed Central CAS Google Scholar
Marshall, C. M., Thompson, V. L., Creux, N. M. & Harmer, S. L. The circadian clock controls temporal and spatial patterns of floral development in sunflower. eLife 12, e80984 (2023).
Fenske, M. P., Nguyen, L. P., Horn, E. K., Riffell, J. A. & Imaizumi, T. Circadian clocks of both plants and pollinators influence flower seeking behavior of the pollinator hawkmoth Manduca sexta. Sci. Rep. 8, 2842 (2018).
Article ADS PubMed PubMed Central Google Scholar
Garsmeur, O. et al. Two evolutionarily distinct classes of paleopolyploidy. Mol. Biol. Evol. 31, 448–454 (2014).
Article PubMed CAS Google Scholar
Fehér, B. et al. Functional interaction of the circadian clock and UV RESISTANCE LOCUS 8-controlled UV-B signaling pathways in Arabidopsis thaliana. Plant J. 67, 37–48 (2011).
Article PubMed Google Scholar
Monnahan, P. et al. Pervasive population genomic consequences of genome duplication in Arabidopsis arenosa. Nat. Ecol. Evol. 3, 457–468 (2019).
Article PubMed Google Scholar
Bloch, G., Bar-Shai, N., Cytter, Y. & Green, R. Time is honey: circadian clocks of bees and flowers and how their interactions may influence ecological communities. Philos. Trans. R. Soc. Lond. B Biol. Sci. 372, 20160256 (2017).
Fenske, M. P. & Imaizumi, T. Circadian rhythms in floral scent emission. Front. Plant Sci. 7, 462 (2016).
Article PubMed PubMed Central Google Scholar
Klein, S. J. & O’Neill, R. J. Transposable elements: genome innovation, chromosome diversity, and centromere conflict. Chromosome Res. 26, 5–23 (2018).
Article PubMed PubMed Central CAS Google Scholar
Korbo, A. et al. Comparison of East and West African populations of baobab (Adansonia digitata L.). Agrofor. Syst. 85, 505–518 (2011).
Article Google Scholar
Liao, N. et al. Chromosome-level genome assembly of bunching onion illuminates genome evolution and flavor formation in Allium crops. Nat. Commun. 13, 6690 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Budhlakoti, N. et al. Genomic selection: a tool for accelerating the efficiency of molecular breeding for development of climate-resilient crops. Front. Genet. 13, 832153 (2022).
Article PubMed PubMed Central Google Scholar
Lutz, K. A., Wang, W., Zdepski, A. & Michael, T. P. Isolation and analysis of high quality nuclear DNA with reduced organellar DNA for plant genome sequencing and resequencing. BMC Biotechnol. 11, 54 (2011).
Article PubMed PubMed Central CAS Google Scholar
Kolmogorov, M., Yuan, J., Lin, Y. & Pevzner, P. A. Assembly of long, error-prone reads using repeat graphs. Nat. Biotechnol. 37, 540–546 (2019).
Article PubMed CAS Google Scholar
Vaser, R., Sović, I., Nagarajan, N. & Šikić, M. Fast and accurate de novo genome assembly from long uncorrected reads. Genome Res. 27, 737–746 (2017).
Article PubMed PubMed Central CAS Google Scholar
Walker, B. J. et al. Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS ONE 9, e112963 (2014).
Article ADS PubMed PubMed Central Google Scholar
Manni, M., Berkeley, M. R., Seppey, M., Simão, F. A. & Zdobnov, E. M. BUSCO update: novel and streamlined workflows along with broader and deeper phylogenetic coverage for scoring of eukaryotic, prokaryotic, and viral genomes. Mol. Biol. Evol. 38, 4647–4654 (2021).
Article PubMed PubMed Central CAS Google Scholar
Ranallo-Benavidez, T. R., Jaron, K. S. & Schatz, M. C. GenomeScope 2.0 and Smudgeplot for reference-free profiling of polyploid genomes. Nat. Commun. 11, 1432 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Weiß, C. L., Pais, M., Cano, L. M., Kamoun, S. & Burbano, H. A. nQuire: a statistical framework for ploidy estimation using next generation sequencing. BMC Bioinforma. 19, 122 (2018).
Article Google Scholar
Ou, S. et al. Author correction: Benchmarking transposable element annotation methods for creation of a streamlined, comprehensive pipeline. Genome Biol. 23, 76 (2022).
Article PubMed PubMed Central Google Scholar
Benson, G. Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res. 27, 573–580 (1999).
Article PubMed PubMed Central CAS Google Scholar
Cantalapiedra, C. P., Hernández-Plaza, A., Letunic, I., Bork, P. & Huerta-Cepas, J. eggNOG-mapper v2: functional annotation, orthology assignments, and domain prediction at the metagenomic scale. Mol. Biol. Evol. 38, 5825–5829 (2021).
Article PubMed PubMed Central CAS Google Scholar
Lynch, R. C. et al. Domesticated cannabinoid synthases amid a wild mosaic cannabis pangenome. Preprint at bioRxiv https://doi.org/10.1101/2024.05.21.595196 (2024).
Cossu, R. M., Buti, M., Giordani, T., Natali, L. & Cavallini, A. A computational study of the dynamics of LTR retrotransposons in the Populus trichocarpa genome. Tree Genet. Genomes 8, 61–75 (2011).
Article Google Scholar
Emms, D. M. & Kelly, S. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 20, 238 (2019).
Article PubMed PubMed Central Google Scholar
Goodstein, D. M. et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res. 40, D1178–D1186 (2012).
Article PubMed CAS Google Scholar
Mendes, F. K., Vanderpool, D., Fulton, B. & Hahn, M. W. CAFE 5 models variation in evolutionary rates among gene families. Bioinformatics 36, 5516–5518 (2021).
Article PubMed Google Scholar
Padgitt-Cobb, L. K., Pitra, N. J., Matthews, P. D., Henning, J. A. & Hendrix, D. A. An improved assembly of the ‘Cascade’ hop (Humulus lupulus) genome uncovers signatures of molecular evolution and refines time of divergence estimates for the Cannabaceae family. Hortic. Res 10, uhac281 (2023).
Article PubMed CAS Google Scholar
Goel, M., Sun, H., Jiao, W.-B. & Schneeberger, K. SyRI: finding genomic rearrangements and local sequence differences from whole-genome assemblies. Genome Biol. 20, 277 (2019).
Article PubMed PubMed Central Google Scholar
Gaut, B. S., Morton, B. R., McCaig, B. C. & Clegg, M. T. Substitution rate comparisons between grasses and palms: synonymous rate differences at the nuclear gene Adh parallel rate differences at the plastid gene rbcL. Proc. Natl Acad. Sci. USA 93, 10274–10279 (1996).
Article ADS PubMed PubMed Central CAS Google Scholar
Klopfenstein, D. V. et al. GOATOOLS: a Python library for gene ontology analyses. Sci. Rep. 8, 10872 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Supek, F., Bošnjak, M., Škunca, N. & Šmuc, T. REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS ONE 6, e21800 (2011).
Article ADS PubMed PubMed Central CAS Google Scholar
Kitony, J. K. et al. Chromosome-level baobab genome illuminates its evolutionary trajectory and environmental adaptation. Figshare, https://doi.org/10.6084/m9.figshare.26039878 (2024).

Download references

Acknowledgements

We thank members of the Michael Lab for their comments on the genome work and manuscript. We also thank the genome sequencing team at Monsanto for the initial genome size survey of the Adansonia species funded by the Illumina Greater Good program awarded to T.P.M. Additionally, we are grateful to Mike Winterstein, USDA, GRIN for sending seed and leaf material from A. digitata tree PI77271 for genome sequencing. This work was supported by the Tang Genome Fund to T.P.M., a Global Challenges Research Fund (GCRF), Nottingham Interdisciplinary Research Award, and the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme [grant number ERC-StG 679056 HOTSPOT], via a grant to L.Y. The 25 samples used for resequencing were prepared with financial support from the National Science Foundation award DEB-1354268 to N.K. and field collecting from Diana Mayne, Sarah M. Venter, and Achille E. Assogbadjo. Finally, we are very grateful to David Baum for his constructive suggestions during the writing of the manuscript.

Author information

Bradley W. Abramson
Present address: Noblis, Inc., Washington, DC, USA
Semar Petrus
Present address: Cepheid, Sunnyvale, CA, USA

Authors and Affiliations

Plant Molecular and Cellular Biology Laboratory, The Salk Institute for Biological Studies, La Jolla, CA, USA
Justine K. Kitony, Kelly Colt, Bradley W. Abramson, Nolan T. Hartwick, Semar Petrus & Todd P. Michael
Biomedical and Clinical Research Centre (BCRC), College of Health and Allied Sciences, University of Cape Coast, Cape Coast, Ghana
Emadeldin H. E. Konozy
Missouri Botanical Garden, Science and Conservation Division, St. Louis, MO, USA
Nisa Karimi
Department of Botany, University of Wisconsin - Madison, Madison, WI, USA
Nisa Karimi
School of Life Sciences, University of Nottingham, Nottingham, UK
Levi Yant
Department of Botany, Faculty of Science, Charles University, Prague, Czech Republic
Levi Yant

Authors

Justine K. Kitony
View author publications
Search author on:PubMed Google Scholar
Kelly Colt
View author publications
Search author on:PubMed Google Scholar
Bradley W. Abramson
View author publications
Search author on:PubMed Google Scholar
Nolan T. Hartwick
View author publications
Search author on:PubMed Google Scholar
Semar Petrus
View author publications
Search author on:PubMed Google Scholar
Emadeldin H. E. Konozy
View author publications
Search author on:PubMed Google Scholar
Nisa Karimi
View author publications
Search author on:PubMed Google Scholar
Levi Yant
View author publications
Search author on:PubMed Google Scholar
Todd P. Michael
View author publications
Search author on:PubMed Google Scholar

Contributions

T.P.M. and L.Y. designed the research; J.K.K., T.P.M., K.C., B.W.A., N.T.H., S.P., and N.K. performed research or analyzed data; E.H.E.K. and N.K. contributed materials and/or tools; J.K.K. and T.P.M. wrote the manuscript. All authors revised the manuscript.

Corresponding author

Correspondence to Todd P. Michael.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Bohdan Lojka, Tao Wan, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

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

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Supplementary Data 7

Supplementary Data 8

Supplementary Data 9

Supplementary Data 10

Supplementary Data 11

Supplementary Data 12

Supplementary Data 13

Reporting Summary

Source data

Source Data

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 http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Kitony, J.K., Colt, K., Abramson, B.W. et al. Chromosome-level baobab genome illuminates its evolutionary trajectory and environmental adaptation. Nat Commun 15, 8833 (2024). https://doi.org/10.1038/s41467-024-53157-w

Download citation

Received: 14 April 2024
Accepted: 03 October 2024
Published: 12 October 2024
DOI: https://doi.org/10.1038/s41467-024-53157-w