Parasitic Plants and Fungi

"Alternatively, parasitic plants may have been commensal or mutualistic before the fall and degenerated through mutation and genomic reduction to become parasitic after the fall." — Harry F. Sanders III, Answers in Genesis, 2019¹

"It is clearly challenging to prove the complete absence of a plastid genome" — Jeanmaire Molina and colleagues, on Rafflesia, 2014²

Summary

Parasitic plants and pathogenic fungi are the best place to test the young-earth claim that parasites are broken versions of once-harmless organisms, because here the loss can be counted gene by gene against non-parasitic relatives in the same family.  The count comes out as the loss model predicts, and then it stops short of explaining the parasite.

• The young-earth position, in its own words. Answers in Genesis offers two models for parasitic plants: plants are not alive in the biblical sense, so one plant feeding on another before the Fall breaks no "very good" rule; or parasitic plants "may have been commensal or mutualistic before the fall and degenerated through mutation and genomic reduction." It calls the haustorium, the organ that breaches the host, designed. Creation Ministries International's interface model says parasitism results from "the removal/dysfunction/mutation" of interface elements. For fungi, the only young-earth statement found says plant-fungus partnerships "have deteriorated into parasitic and/or pathogenic relationships."³¹⁴
• The loss is real, measured, and ordered. Susann Wicke's 2016 survey of the broomrape family, the only family holding every stage from free-living to fully parasitic, found that of 113 plastid genes in the non-parasitic relatives, only 16 protein genes, 4 ribosomal RNAs and 14 transfer RNAs survive in every parasite. The ndh genes go first, then photosynthesis genes, then ATP synthase and rbcL, in the same sequence across independent lineages. Dodder has lost 11.7% of the genes conserved in green plants; Sapria, a relative of Rafflesia, 44%; Rafflesia itself has no recognisable plastid genome.²⁵⁶⁷
• The loss follows dependence; it does not create it. Wicke's data tie the rate shifts and gene losses to the loss of photosynthesis, "whereas parasitism per se, that is the ability to tap into the vascular tissue of another plant, are of subordinate importance." Parasitism arose 12 separate times, and each origin needed three gains: host recognition, an invasive haustorium, and control of transfer. The haustorium's genes are duplicated and repurposed root genes; a growing share of parasite genes came from hosts by horizontal transfer; and dodder takes up "nearly half the expressed transcriptome" of a host plant across the junction.^8910511
• Fungi show the same split. Rust and smut genomes have lost nitrogen and sulfur assimilation and cell-wall enzymes, but have gained large families of secreted effector proteins. The frog-killing chytrid has a non-pathogenic sister lineage, and what separates them is 1,974 genes unique to the pathogen and protease families at least four times larger, expansions that happened after the two lineages split. The panzootic lineage emerged in the early twentieth century out of Korea and spread by trade. Its toll, 501 declining species and 90 presumed extinctions, is disputed in count but not in direction.^1213141516
• Verdict. Where the loss model makes a testable prediction, it passes. But it explains the part of the parasite that a free-living plant already had, not the part that makes it a parasite. The design answer is weakest exactly where Answers in Genesis places it: the haustorium and self/non-self recognition are "designed," which concedes that the parasitic organ is the created part.

Part 1: What young-earth writers propose

The claim has to be stated before it is priced, and the young-earth literature on this subject is thin, so it can be quoted almost entire.

Harry F. Sanders III's 2019 Answers in Genesis article is the only young-earth treatment of parasitic plants found.  It gives two options and allows both.  The first rests on plants not being alive in the biblical sense, so that eating one is not death:  "Thus a plant parasitizing another plant in the pre-fall world would not violate God's 'very good' statement."  The second is degeneration:  "Alternatively, parasitic plants may have been commensal or mutualistic before the fall and degenerated through mutation and genomic reduction to become parasitic after the fall."  He closes with both on the table:  "Parasitic plants could have existed in the pre-fall world as inhibitors on rapidly growing plants or could have devolved in the post-fall world as a result of the curse or both."  On the organ itself he is definite:  "The haustorium and the ability to distinguish between self and not-self would certainly seem to have been designed."¹¹¹¹

Tom Hennigan, Randy Guliuzza and Grace Lansdell's 2022 paper in the Journal of Creation proposes that organisms were built with engineered "interface systems" for cooperation, and uses dodder on potato as its example of parasitism.  Their account of how parasitism arises is a loss account:  "In a MOSR, parasitism and commensalism may be the result of the removal/dysfunction/mutation of one or more of the above elements of interface systems.  In this view, parasitism is the violation of distinct boundaries of one entity upon another."⁴

For fungi, the 2007 Microbe Forum published by Answers in Genesis carries a one-page abstract on plant pathogens under a pseudonym:  "Since the Fall, many plant-fungal interactions have deteriorated into parasitic and/or pathogenic relationships: biotrophic, hemibiotrophic, and necrotrophic interactions."  Its proposal for animal-infecting fungi is not loss but reuse:  "Recent research demonstrates a possible mechanism for the rise of animal pathogenesis in fungi: co-opting of environmentally relevant survival mechanisms for survival in the host environment."³³

Warren Shipton's 2019 paper in the same journal states the general model in one sentence:  "Parasitism arose through gene loss, the expression of front-loaded genes now facilitating pathological life cycles, or other scenarios."  His examples are insects and worms;  he says nothing specific about plants or fungi.¹⁷

No young-earth treatment of rusts, smuts, or the amphibian chytrid was found on the Answers in Genesis, Institute for Creation Research or Creation Ministries International sites, nor in the Creation Research Society Quarterly.  The statements above are the whole of the position, and they make two distinct claims:  that a parasite differs from its free-living relatives mainly by what is missing, and that the haustorium is built.  The evidence below tests the first and prices the second.

Part 2: Parasitic plants shed photosynthesis in a repeatable order

Twelve origins, one percent of flowering plants

James Westwood, John Yoder, Michael Timko and Claude dePamphilis wrote in 2010 that "In angiosperms, parasitism has independently evolved at least 12 or 13 times, and by recent estimates approximately 1% of angiosperm species are parasitic".  Daniel Nickrent's 2020 census in Taxon settled the count:  plants that attach to other plants by a haustorium "have evolved 12 times independently resulting in 292 genera and ca. 4750 species."  Only the broomrapes (Orobanchaceae) contain every stage, from the free-living Lindenbergia through green hemiparasites that photosynthesise while tapping a host, to holoparasites with no chlorophyll, so that family is where the loss can be read as a series.⁸¹⁸

The broomrape series

Wicke and five colleagues sequenced the plastid genomes of 17 broomrapes and compared them with the non-parasitic Lindenbergia in the same family and two relatives outside it.  Their result:  "genes for only 16 proteins, 4 ribosomal RNAs, and 14 transfer RNAs of the 113 unique genes found in the nonparasitic Lindenbergia philippensis (Orobanchaceae), Erythranthe guttata (Phrymaceae), and Sesamum indicum (Pedaliaceae) are commonly present in the plastid genomes of all of the hemiparasitic and holoparasitic plants".  The holoparasites keep between 42 and 71 intact genes each.⁵

The order matters more than the total.  The ndh genes, which encode a complex that adjusts photosynthesis under stress, are lost first, in hemiparasites that still photosynthesise:  the authors call this "rendering ndh genes the first, to our knowledge, to be functionally lost from the plastome".  Next go the photosystem and cytochrome genes and the plastid's own RNA polymerase, then the ATP synthase genes and rbcL, then the remaining housekeeping genes.  Their figure 3 lays out five steps and marks the boundary of the data:  "The associations of evolutionary rates, genomic traits, and functional losses between 1 and 3 have been observed, whereas those after 3 are hypothetical."  Each loss of a functional complex is followed by a burst of faster evolution in what remains, which the authors describe as "roller-coaster rate variation" along the road to full parasitism.⁵⁵⁵

This is the strongest confirmation the loss model gets anywhere:  independent lineages within one family lose the same genes in the sequence in which a host's sugar makes them dispensable.

Dodder and the nuclear genome

The plastid holds about a hundred genes;  the nucleus holds tens of thousands, and the same pattern appears there.  Guiling Sun and colleagues sequenced the Australian dodder, Cuscuta australis, in 2018:  19,671 protein-coding genes, and "11.7% of the conserved orthologs in autotrophic plants are lost".  Of the gene groups lost only in dodder, "839 orthogroups specifically lost their members in C. australis, and these genes are mainly involved in response to light, photosynthesis, chloroplast RNA processing, and adventitious root development".  A further 1,168 pseudogenes carry frame shifts or premature stops, among them 13 photosynthesis genes still recognisable as wreckage.  Alexander Vogel's group sequenced the field dodder, Cuscuta campestris, the same year:  it "lacks genes for pathways superfluous to a parasitic lifestyle," those for high photosynthetic activity and for taking nutrients from soil.⁶⁶⁶¹⁹

Patrick Lyko and Wicke reviewed every sequenced parasitic plant in 2021 and found the losses convergent:  in Sapria, "Gene losses are enriched in gene families associated with photosynthesis, plastid biology, defense, and stress response genes (as described in the sequenced Orobanchaceae species and Cuscuta spp.)."  Three families, three separate origins, the same categories missing.²⁰

Rafflesia: how outright is "outright"?

The catalogue for this article says Rafflesia "has lost its chloroplast genome outright."  The primary source is more careful.  Molina, Nickrent and colleagues sequenced Rafflesia lagascae in 2014 from about 440 million reads.  They assembled the mitochondrial genome at about 350-fold depth, but recovered only small plastid fragments at under twofold depth, and "None of the identified plastid gene sequences had intact reading frames."  Their conclusion is hedged twice in one sentence:  the data "suggests that the parasitic plant genus Rafflesia may be the first plant group for which there is no recognizable plastid genome, or if present is found in cryptic form at very low levels."  In the discussion they say why:  "It is clearly challenging to prove the complete absence of a plastid genome".  And the organelle itself is not gone:  "Despite the possible loss of the chloroplast and its genome, there still remain plastid-like structures in Rafflesia."²²²²

Later work dropped the hedge.  Liming Cai, Charles Davis and colleagues sequenced the related Sapria himalayana in 2021 and reported that "44% of the genes conserved in eurosids are lost in Sapria, dwarfing previously reported levels of gene loss in vascular plants."  A 2023 BMC Biology genome of the same species by Xiaodong Guo's group states flatly that "Plastid genome has been completely lost in Rafflesiaceae plants," citing both papers, and its own methods record that "because of very few plastid genome-specific Illumina short reads, the plastid genome could not be assembled."  The authors conclude that Rafflesiaceae "represent a plastid-bearing lineage, without a plastid genome, that is capable of performing plastid-based metabolic functions".  So "outright" is now what the field says, on the strength of absence in three deep sequencing projects;  Molina's caution still stands in logic, and the organelle itself survives, running on nuclear genes.^7212121

Part 3: What the loss model does not explain

The haustorium is a gain

Every one of the 12 origins had to produce the same three things.  Westwood's group listed them:  "the ability to recognize host plants, form an invasive haustorium, and regulate the transfer of nutrients and other molecules between two different plants."  None of these is a loss.  A plant that has merely lost photosynthesis genes is a dead plant, not a parasite.⁸

Where the haustorium's genes come from is now partly known.  Zhenzhen Yang's 2015 transcriptome study of three broomrapes identified a core set of parasitism genes switched on when the haustorium forms and found that "Most of these parasitism genes are derived from gene duplications in a common ancestor of Orobanchaceae and Mimulus guttatus, a related nonparasitic plant."  Sun's dodder genome reached the same conclusion from a different family:  haustorium formation "requires mostly genes normally involved in root development".  The raw material is old and shared with non-parasites;  what changed is where and when it is expressed.  Nickrent's census makes the same point from the other side:  the closest non-parasitic relatives of holoparasites are known, and it is the holoparasites that show "reduced morphologies, increased substitution rates, and frequent horizontal gene transfers".¹⁰⁶¹⁸

Genes and messages cross the junction

The second gain is imported.  Yang's 2016 study of three broomrapes and the free-living Lindenbergia found "52 high-confidence horizontal transfer events were detected, often from lineages of known host plants and with an increasing number of HGT events in species with the greatest parasitic dependence."  The transferred genes are not junk:  "HGT-acquired genes are preferentially expressed in the haustorium—the organ of parasitic plants—and are strongly biased in predicted gene functions".  Vogel's field dodder genome found "evidence for horizontal gene transfer by way of genomic DNA integration from the parasite's hosts".  In Sapria the figure is "at least 1.2% of the Sapria genome, including both genic and intergenic content," derived from hosts, "and includes genes potentially adaptive for parasitism."^1111197

Claims of gene transfer between eukaryotes are often overstated, as William Martin warned in 2017.  The Orobanchaceae cases meet that objection:  Yang's team traced intron sequences in donor and recipient and found genomic fragments integrated far more often than processed RNA copies, the signature of DNA transfer rather than contamination or convergence.¹¹

Traffic across the haustorium is not only DNA.  Gunjune Kim and Westwood's 2014 Science paper sequenced dodder growing on Arabidopsis and tomato and found host messenger RNAs moving into the parasite in bulk:  "The mobile transcripts represented thousands of different genes, and nearly half the expressed transcriptome of Arabidopsis was identified in Cuscuta."  Only the abstract of that paper could be obtained.⁹

The order of events runs the other way

The loss model needs loss to come first.  The data put it second.  Wicke's analysis found the rate accelerations and relaxed selection tied to obligate parasitism and the loss of photosynthesis, "whereas parasitism per se, that is the ability to tap into the vascular tissue of another plant, are of subordinate importance".  Hemiparasites with fully working haustoria have lost almost nothing;  loss begins once the host's sugar makes photosynthesis optional.  Sanders's degeneration model can accommodate this only by supposing that the haustorium existed first, in a commensal or mutualistic plant, and that is exactly what he proposes.  The difficulty is that no haustorial plant is known to be commensal or mutualistic toward its host.⁵

The nearest candidate is mistletoe, and its benefit runs to the community, not the host.  David Watson and Matthew Herring cut every mistletoe out of 17 Australian woodlands and compared them with 11 controls and 12 woodlands without mistletoe.  Three years on, "treatment woodlands lost, on average, 20.9 per cent of their total species richness, 26.5 per cent of woodland-dependent bird species and 34.8 per cent of their woodland-dependent residents".  Mistletoe is a keystone resource, as Watson had argued in 2001;  the host trees still lose water and wood to it.  A parasite can be good for a forest and bad for its host at once, and only the second relation is what the Fall model has to explain.²²²²

Part 4: Rusts, smuts and chytrids

Rusts and smuts: losses beside arsenals

Fungi that feed only on living tissue, the obligate biotrophs, are the fungal analogue of holoparasites, and their genomes show the same two-sided pattern.  Sébastien Duplessis and colleagues sequenced the poplar leaf rust (101 million bases) and wheat stem rust (89 million) in 2011.  The features tied to their lifestyle "include expanded lineage-specific gene families, a large repertoire of effector-like small secreted proteins, impaired nitrogen and sulfur assimilation pathways, and expanded families of amino acid and oligopeptide membrane transporters."  The impaired assimilation is what the loss model predicts;  the expanded effector families and transporters are not.¹²

The corn smut Ustilago maydis is the same story in a small genome.  Jörg Kämper's 2006 Nature paper reported that "The 20.5-million-base U. maydis genome assembly contains 6,902 predicted protein-encoding genes and lacks pathogenicity signatures found in the genomes of aggressive pathogenic fungi, for example a battery of cell-wall-degrading enzymes."  What it has instead is specific to infection:  "we found 12 clusters of genes encoding small secreted proteins with unknown function," induced in infected tissue, and deleting single clusters changed virulence in five cases.  Pietro Spanu's 2010 survey of powdery mildews, read only in abstract, found "massive retrotransposon proliferation, genome-size expansion, and gene losses," the lost genes being metabolic enzymes and transporters made redundant by living inside a host, beside 248 candidate effectors.²³²³²⁴

So the Microbe Forum's sentence that plant-fungus relations "deteriorated into" biotrophy describes half of each genome.  The other half is an effector arsenal, gene families that grew rather than shrank, and no young-earth source addresses where those came from.

The frog chytrid and its harmless sister

Batrachochytrium dendrobatidis (Bd) is the cleanest test:  a pathogenic fungus with a close non-pathogenic relative.  Suzanne Joneson, Jason Stajich, Shin-Han Shiu and Erica Bree Rosenblum sequenced "the genome of Bd's closest known relative - a non-pathogenic chytrid Homolaphlyctis polyrhiza (Hp)" in 2011 and compared the two.  The difference is not what Bd lacks.  "We identified 1,974 Bd-specific genes, a gene set that is enriched for protease, lipase, and microbial effector Gene Ontology terms."  In the results:  "The Bd genome contained 38 metalloproteases, 32 serine-type proteases, and 99 aspartyl proteases, in all cases at least 4 times as many family members as Hp".  The expansions "occurred after the divergence of Bd and Hp from their common ancestor".  Matthew Fisher and Trenton Garner's 2020 review adds the salamander chytrid discovered in 2013:  the skin-digesting M36 metalloproteases are "strikingly expanded in both Bsal and Bd with 110 and 35 of theses proteases, respectively, compared to the free-living saprobic chytrids Spizellomyces punctatus and Homolaphlyctis polyrhiza, which have 2 and 3, respectively".^1313132526

This is the opposite of the plant result.  A holoparasite is its relative minus a hundred genes;  Bd is its relative plus two thousand.  Whatever made a saprobe into a skin pathogen, it was not deterioration, and the Microbe Forum's second sentence, co-option of "environmentally relevant survival mechanisms," is the young-earth proposal that fits it.

A dated event

The chytrid panzootic is also the one parasite story here with a date, and the date is recent.  Rhys Farrer's 2011 population genomics found three deeply diverged lineages, of which "isolates belonging to one clade, the global panzootic lineage (BdGPL) have emerged across at least five continents during the 20th century", with a molecular-clock emergence of "35 to 257 y before present".  Simon O'Hanlon's 2018 Science paper, with 234 isolates, traced the source to Korea:  "We date the emergence of this pathogen to the early 20th century, coinciding with the global expansion of commercial trade in amphibians".  In the parent article's terms this is displacement, not alteration:  a lineage carried out of its home range by people into hosts that had never met it.²⁷²⁷¹⁴

The toll, and the dispute about it

Ben Scheele and 41 co-authors put a number on the damage in 2019:  chytridiomycosis contributed to "the decline of at least 501 amphibian species over the last half century, including 90 presumed extinctions."  Only 12% of declined species show signs of recovery and 39% are still declining, and the authors call it "the greatest recorded loss of biodiversity attributable to a disease."  Their own methods say that "Most evidence is retrospective because many species declined before the discovery of chytridiomycosis".¹⁵¹⁵¹⁵

Max Lambert and 13 colleagues, including Rosenblum, answered in Science in 2020 that "the data and methods implicating the disease chytridiomycosis in 501 amphibian species declines are deficient."  They counted:  "For 11.6% of included species, chytridiomycosis' role in a decline was based solely on expert opinion".  Rebuilding the evidence categories from Scheele's own data table, they "found that more than half of the species' assignments changed, many dropping from category 3 (multiple lines of evidence) to category 1 (expert opinion only", while adding that "We are in no way diminishing the role of chytridiomycosis in amphibian declines".  Scheele's group replied that expert judgement is standard epidemiology, that Lambert's team had "treat[ed] expert knowledge and retrospective evidence as unreliable (at best) and suspicious (at worst)," and reaffirmed "unambiguous evidence that chytridiomycosis has impacted at least 501 amphibian species."^161616162828

The honest reading is that 501 and 90 are upper bounds built partly on expert opinion, that the critics could not reproduce more than half the evidence grades, and that nobody in the exchange doubts the fungus has driven declines and extinctions.  The dispute is about evidence standards, not about whether frogs died.

Objections and responses

"The gene-loss data confirm the creationist model."  They confirm one prediction of it:  a parasite differs from its relatives by missing genes, in a functional order.  They refute another reading of it:  that loss made the parasite.  Wicke's data show loss beginning after dependence is established, in plants whose haustoria already work.⁵

"The haustorium is only a modified root, so it is not a gain."  Its genes are largely root genes, duplicated and re-deployed, and that is the point:  the change is in regulation and in the ability to recognise a foreign plant and breach it.  A root that grows into soil and one that grows into a neighbour's stem and draws its sugar are not the same organ.  Answers in Genesis agrees, since it calls the haustorium designed;  either way it is not a loss.¹⁰¹

"Rafflesia's plastid genome may still be there."  Molina said so in 2014, and it remains formally possible.  But the mitochondrial genome came out of the same reads at 350-fold coverage while plastid genes came out as fragments without reading frames, and two later projects on Sapria found too few plastid reads to assemble anything.  The evidence is absence at a depth where presence would show.²²¹

"Twelve independent origins of the same organ point to design."  Westwood's group draws the opposite inference, that many plant lineages "possessed the developmental flexibility to meet the requirements of a parasitic life style".  The count assumes common descent among flowering plants, so a reader who rejects the tree cannot use it either.  The stronger design point is the content:  each origin required recognition, invasion and transfer control, and nobody has watched any of them arise.⁸

"Plants are not alive in the biblical sense, so parasitic plants need no Fall."  That dissolves the theological problem for plants, and Sanders offers it.  It does nothing for fungi that kill frogs, and it leaves the genomic question untouched:  if parasitic plants were made parasitic, why do they carry the pseudogenes of photosynthesis?⁶¹

"The chytrid toll is exaggerated."  The count is contested by competent critics and the 90 extinctions are "presumed."  But the disagreement is over grading evidence for individual species, and the critics themselves affirm the fungus's role.  A young-earth reader gains nothing here in any case:  a lineage that emerged in the twentieth century and spread by trade is an invasion, not a Fall.¹⁴¹⁶

Issues and unanswered questions

• No young-earth source addresses the fungal cases. The Microbe Forum abstract is two sentences, and nothing was found on rusts, smuts or chytrids from any creationist organisation. The fungal section therefore tests general statements, not a worked model.
• No haustorial plant is known to be commensal or mutualistic with its host. Sanders's degeneration model requires one to have existed. Mistletoe's keystone role is a benefit to birds and insects, not to the tree it grows on.
• The timescale of the losses has not been computed here. A young-earth reading needs 42 to 71 plastid genes to be lost, and 1,168 nuclear pseudogenes to decay to their present state, in under 6,000 years. Settling that needs the per-generation plastid mutation rate, the generation time of each holoparasite, and the observed pseudogene divergence; nobody on either side has published that calculation for these plants.
• Two load-bearing papers were read only in abstract. Kim 2014 on messenger RNA exchange and Cai 2021 on Sapria are paywalled; the article's claims about them go no further than their abstracts.
• The 501-species count is disputed. The article reports the dispute rather than adjudicating it; treat 501 and 90 as upper bounds.
• Whether Bd's protease expansions preceded pathogenicity or followed it is not settled by two genomes. Joneson's group dated the expansions to after the split from Hp; what selected for them, and in what host, is unknown.

Verdict

The loss model is confirmed where it is testable and silent where it matters.  Across three independent families, parasitic plants lose photosynthesis and plastid genes in the same functional order, dodder has shed 11.7% of the genes green plants share, and the Rafflesia family has no plastid genome that three deep sequencing projects could find.  Every one of those losses is downstream of the haustorium, which no lineage lost anything to acquire:  its genes are duplicated root genes and genes taken from hosts, and dodder draws half a host's messages across it.  On the fungal side, obligate biotrophs have lost assimilation pathways and gained effector arsenals, and the frog chytrid is its harmless sister plus two thousand genes, a twentieth-century invader whose toll is real and whose count is contested.  Young-earth writers have the losses right and the haustorium conceded as designed, which leaves their model explaining how a parasite decays and not how it became one.

Sources