Research The nexian phalaris breeding programme

[Ivan Illich]

Thanks! I appreciate the quick answer.
This is a quite a solid recurrent selection setup, Some things remain unclear to me however. A few follow-ups on what I think are some key things to point out/find out here:

How many plants enter each breeding cycle, and how many contribute pollen and seed, are these sampled?
Since the plant is self-incompatible and tetraploid you can't inbreed to fix traits, so a population size is really your main defense against losing traits. Removing the bottom 50% low producers each round is a good intensity, but population has to be kept high enough. Please note here!!!: If you keep 100 plants but only a handful end up as pollen donors (because a few produce copious pollen and dominate, or they produce ripe pollen first), your Ne is set by that handful, not by 100.

Have you tracked whether group means for metabolite are actually moving across cycles, or is the gain assumed? What measures have been taken to minimize the effect of environment on alkaloid content? This is to make sure that the effects you are seeing actually have a genetic foundation instead of an environmental one.

The selected plants are then grouped according to their chemotype profiles using the following classification:
Type 1: DMT-dominant
Type 2: DMT and 5-MeO-DMT
Type 3: 5-MeO-DMT and 5-MeO-NMT

Plants are subsequently cross-pollinated within their respective groups. Where, exceptional individuals may be used for introgression between groups to combine desirable traits.

Since the metabolite profiles likely represent the final branches in the metabolic pathway, it is probably not very useful to separate your plants in these chemotype groups yet and just aim for higher total alkaloid content instead. As this represents the total flux capacity of the metabolic pathways that can produce these compounds. If the effective breeding population is sufficiently large, you will keep all these chemo types at the end of that ride, then you select from there. Even keeping gramine in might actually be good, that is if it is yet another tip of the fork of the N,N-DMT or 5-Meo-DMT biosynthesis pathway. However it is super promising to see that it can be selected out relatively quickly from a population! I'll see if I can dive into some literature to find out what the bio synthetic pathways look like for these compounds.

It makes sense that the lineages are unknown from your wild samples. However, have population level lineages been kept, as soon as you started breeding and crossing?
moreover:
- The amount of generations for example,
- which plants were taken for the next generation,
- how many plants were kept? (Although getting the actual pollinators would be amazing).
Basically anything that will help me see important aspects of the breeding process

Really looking forward to hearing what you guys think! It's been a while for me when I had to use good old breeding genetics, so if I made some errors, or you just have general disagreements, let me know!

Thanks a lot for initiating this wonderful project.

Much love

 

[TransistorBass]

I'm simply asking out of interest.

I don't understand why you don't like acid base? Tnt tek is just another form of alcohol Tek.

 

[Grasshoppers]

The Evolution of a Underground Project
This project began years ago, sparked by decades-old rumors in the psychedelic community and literature on dead sheep on Phalaris pastures. It started with gardeners, hobbyists, and plant enthusiasts collecting seeds, exchanging ideas, and conducting positive bioassays that induced life-changing psychedelic journeys.

Eventually, we moved toward Thin-Layer Chromatography (TLC). By testing what we grew, we began exerting selective breeding pressure, leading to even more intriguing discoveries.
There is no written genealogical tree—no standardization, no formal documentation. Only anecdotes. It isn’t standardized today, and perhaps it never will be. The number of plants entering each breeding cycle is strictly limited by the testing capacity of myself and my fellow grasshoppers.

This work has evolved far beyond average gardening. Today, chemical testing makes up the majority of the breeding effort. While the exact number is unknown, a few hundred plants enter a new cycle each year.


The Current State & Breeding Dynamics
Progress is undeniable. Potency means are way up, and low-yielding plants have become rare in the breeding line. In fact, some of our selections are approaching the potency of Psychotria viridis. However, this plant is notoriously difficult to work with.

Here is what we are learning and navigating right now:

• Environmental Variables: Plants must be compared locally—grown in the same conditions and sampled during the same season—to limit the massive impact of environmental factors.
• Testing Rigor: Repeated testing is highly recommended before selecting a single plant. We hope to reduce the recommended number of tests per plant once we better understand the alkaloid fluctuations.
• Genetic Challenges: Self-incompatibility is not a reliable trait in these hybrids (Putievsky, Oram, Malafant 1980). Furthermore, extreme profiles (Type 1 and Type 3) could easily get lost in the massive Type 2 middle ground of the tetraploid Hardy-Weinberg equilibrium. Because these extremes are vital for future exploration, separating them into distinct breeding groups is essential, even if temporary.
• Keep moving forward: We currently recommend a heavy culling rate of about 80%.

Where We Are Headed
Right now, our focus is to deeply understand these chemical profiles: how they respond to climate cues, how to streamline the testing/growing process, and how to map the exact psychedelic effects of the different types and subtypes. Of course, we are also continuously acquiring wild germplasm to expand the gene pool.
In the future, we will certainly see a more structured, planned approach—though it can never be completely steered, given the independent nature of the grasshoppers involved.


Join the Search
Every gardener is welcome to join us. We need more hands plowing the soil to raise promising seedlings, and more eyes finding wild germplasm to enrich our gene pool.
There are still so many unanswered questions, and so many things left to try.
I am curious to hear your thoughts and deeply appreciate your contributions to this space. Your posts will be read.
Welcome to the journey if you choose to join us.

The project is just beginning.

 

[aizoaceous]

Since the plant is self-incompatible and tetraploid you can't inbreed to fix traits,

To emphasize @Grasshoppers's note, the genus Phalaris is reported to contain both consistently self-compatible species (e.g. P. brachystachys) and consistently self-incompatible species (e.g. P. arundinacea). From previous discussion P. aquatica appears to be inconsistent, with examples of both.

I assume you're not emasculating for your crosses, so it's possible that some of them are actually selfed? Have you tried deliberately selfing any of your plants, both as potentially interesting descendants and as a test of whether emasculation would have an effect? Manual emasculation seems impossibly tedious, but hot water was reported to work on P. canariensis.

This is really interesting work. @Grasshoppers potentially has the most diverse collection of Phalaris anywhere in the world, if commercial breeding programs (for forage grass or alpiste) discard most of that diversity when they first select for low alkaloids.

 

[Grasshoppers]

Thank you.

We did not perform emasculation. Self-fertilization is likely occurring. If no additional germplasm is introduced, self-incompatibility should be restored over successive generations. However, this topic would certainly merit a more thorough investigation.

I was not aware of the hot water method for emasculation, so thank you for bringing it to my attention.

Emasculation will also be essential for future attempts to hybridize P. aquatica and P. arundinacea, which I believe is one of the most important directions for future research.

 

[neurobloom]

Yes hot water emasculation is well covered up in agronomic literature.

Arudinacea was hybridized with aquatica using hot water emasculation and the offspring was crossed back with aquatica again to obtain water logging tolerant Aquatica cultivar according to CSIRO but the cultivar was dismissed for causing high rates of sheep staggers.

I think the reverse can be done to migrate aquatica genetics into arudinacea. I think this is better approach than maintaining the original hybrid which would almost certainly have issues one way or another like very low seed count per panicle or completely absent seeds or poor seed viability.

It could also exhibit other genetic flaws like poorly formed panicles or mutated weird looking leaves.

 

[Ivan Illich]

Nice! I'll look into the water method, as well as that self-fertilization paper!
I am also making an overview of the bio synthetic pathways for the metabolites that are found in Phalaris spp.
I am still affiliated with my university, so if you guys need something more specific like papers or some database, let me know.

Genetic Challenges: Self-incompatibility is not a reliable trait in these hybrids (Putievsky, Oram, Malafant 1980). Furthermore, extreme profiles (Type 1 and Type 3) could easily get lost in the massive Type 2 middle ground of the tetraploid Hardy-Weinberg equilibrium. Because these extremes are vital for future exploration, separating them into distinct breeding groups is essential, even if temporary.

For you current breeding line, do you have numbers of the ratio's between type 1, 2 and 3? Or all they all kind of equally occurring?
As for the seeds you have for that line, they are at the population level, or from a single individual?

Also @Grasshoppers, I appreciate the thorough message, however I notice a lot of LLM-style embellishments in the text. The content is most important, but I think a human touch would be nice . I suppose being a bit more direct with communication can't go wrong with projects like these!

 

[Ivan Illich]

There is quite a good genome available for P. arundinacea! Could be really useful.

Phalaris arundinacea genome assembly lpPhaArun1.hap1.1

- Phalaris arundinacea

www.ncbi.nlm.nih.gov

 

[Transform]

appreciate the thorough message, however I notice a lot of LLM-style embellishments in the text. The content is most important, but I think a human touch would be nice . I suppose being a bit more direct with communication can't go wrong with projects like these!

AFAIK, @Grasshoppers is a collective, none of whom have English as their native language. (In fact, you might be surprised at the number of ESOL Nexians who are active here…) Ergo, if there's any user who I'd give generous slack with use of LLM "AI" as an aid to composition, it's these guys

 

[Ivan Illich]

I understand, whatever works best of course! I am not against use of it.
The text just felt a bit off to me, perhaps I shouldn't have commented on it. I hope I didn't rub anyone in the wrong way. Sorry!

I am not surprised on the amount of ESOL's here, I am one myself too.

 

[Transform]

I understand, whatever works best of course! I am not against use of it.
The text just felt a bit off to me, perhaps I shouldn't have commented on it. I hope I didn't rub anyone in the wrong way. Sorry!

I am not surprised on the amount of ESOL's here, I am one myself too.

It's fine, quite understandable, in fact!

Also, it's very exciting to have you here with your background in plant bioscience - and I can barely wait to see what you'll manage to bring to the project! Do you anticipate you'll be applying TLC testing to your specimens as well?

 

[Ivan Illich]

It's fine, quite understandable, in fact!

Also, it's very exciting to have you here with your background in plant bioscience - and I can barely wait to see what you'll manage to bring to the project! Do you anticipate you'll be applying TLC testing to your specimens as well?

I have done a few TLC's in the past for a course where we are allowed to extract a compound of choice from a plant of choice (most fun course ever, we achieved 99% pure mitragynine from kratom). However I have never done it for tryptamines and at home, so perhaps I can learn the proper technique here from the experts!
My main worry for this project however, is that I do not have a garden. I live in a student dormitory. However I am a nearly graduated student (A few more months to go!!!) so my living situation is likely going to change rather soon.

It's a shame because I love gardening, my parents allowed me to grow a cannabis plant in their garden and I grew this monstrosity of a plant!

 

[Ivan Illich]

on a sidenote, am I allowed to share books in here? (Copyrighted material)
I have a nice book pdf called "Principles of Plant Genetics and Breeding"

 

[Ivan Illich]

I guess your love emoji means yes, so here it is:

 

[blig-blug]

Share away, they can issue a takedown request if they dislike it being shared.
And thank you because it looks like a very interesting book!

 

[Ivan Illich]

I found a paper, where they discuss breeding for a non-toxic P. Aquatica cultivar. AKA less alkaloids.
This R.N Oram seems to have a lot of experience in breeding Phalaris Aquatica.

No time to read it currently though. I will get back to it sooner or later.

 

[Grasshoppers]

@Transform is correct. This account is primarily intended for external communications—to document the breeding process and provide information about the project.

Type 2 alkaloid profiles occur most frequently in Phalaris aquatica populations. There is also a non-psychedelic, gramine-dominant type, although it has largely disappeared from our breeding populations.

Thank you for the e-book.

CSIRO conducted extensive research on Phalaris aquatica toxicology. You can find CSIRO publications here:

CSIRO Research Publications Repository

Several Phalaris aquatica cultivars have been bred for low alkaloid content to improve their suitability as forage, including:

• ADVANCED AT
• ATLAS PG
• CONFEDERATE
• SIROLAN
• SIROSA
• HOLDFAST
• HOLDFAST GT
• HORIZON
• LANDMASTER
• MATÉ
• MARU

There are also several older cultivars that have retained their original alkaloid profiles but are no longer used commercially, including:

• AUSTRALIAN
• AUSTRALIAN II
• UNETA

These cultivars differ in a range of agronomic characteristics, including seed retention, winter growth, summer dormancy and tolerance to waterlogging. They are of obvious value for our breeding efforts.

 

[Ivan Illich]

I'm loving diving back into plant breeding again! It's been 4 years but this is such a good refresher. While I am reading the book, there are sections which provide important information for our species. Giving us clues to what effective breeding strategies might look like. I'm going to a little book review/discussion that applies the information to our species, and what the implications are. But before that we need to know with what karyotype we are working with. Let's make a little overview first:

What kind of genetic inheritance are we working with?

Our lovely plant:

Family:	Poaceae
Subfamily:	Pooideae
Species:	Phalaris aquatica

chromosomal constitution: 2n = 4x = 28
(a somatic cell (2n), has a sets of 4 chromosomes, with 28 total chromosomes, meaning there are 7 chromosome types e.g x = 7)
The tetraploidy level is sometimes also denoted as 4n = 28, confusing, I know...

There is some discussion on how this sets of 4 chromosomes are composed. The main takeaway that I get from these two papers is that it's genome is compose of two ancestral genomes that have combined into a new one. Here is an image depicting what that means:




The chromosomes in P. Aquatica appear to form two distinct ancestors, B and C. This is what we mean when we say that P. Aquatica is an allopolyploid species. Roughly speaking, you can kind of think of it as two diploid genomes living in the same organism. The genomes usually stay seperate, each behaving as their own diploid, but occasionally they do pair with each other and exchange segments through recombination. Some chromosomes between B and C are more likely to do this than others. Which ones can pair between B and C is not known.

B1 chromosomes = homologs
B1 and C1 = homeologs

*AI assisted part*
The reason they usually stay separate isn't that they can't pair. it's that the plant has an active control system that suppresses B–C pairing, forcing each chromosome to pair only with its true partner. We know the required similarity to pair is still there because of what happens when you make a haploid, (one B set + one C set. e.g B1 now only is a single chromosome, same for C1, B2...), and now, with no true partners, the B and C chromosomes are forced to either pair with each other or not. About 4 of the 7 possible pairings still form, which shows the two genomes remain similar enough to pair the control system just normally keeps them apart.
*End*

In wheat there is actually a similar mechanism, and the molecular foundation has been found: ZIP4 is required for normal progression of synapsis and for over 95% of crossovers in wheat meiosis - PMC
There is a gene that makes the pairing more strict, so B1 and C1 don't pair. But if that gene is knocked out, the organism is suddenly a lot less strict about pairing between B1 and C1. Likely a similar mechanism is at work here. This system can also be broken.

Anyways, lets not go to deep on all this , I got a bit carried away!
What does this mean for breeding anyway?

Well, the conclusion is here is that we can breed it as a regular diploid instead of a complex tetraploid. PHEW! Standard stuff: recurrent selection, tracking allele frequencies, truncation selection all work on the assumption of orderly mendelian inheritance. There are caveats of-course
- Selecting a recessive chemotype allele is harder
- Wide crosses carry a meiotic-abnormality tax
- Just like the selfing that can be broken by wide crosses, the inhibition of homeologous pairing can be broken too by wide crosses. (
- This can lead to all kinds of finnicky meiotic abnormalities like partial sterility, seedling lethals, male sterilty and aneuploids

So plants that look messed up on the first or second generation might still be valuable!

I also found this important section in the book (section 3.11.1), as it highlights a trap we might fall for.


In self-pollinated species, individuals are homozygous and when used in a cross their genotype is precisely reproduced in their progeny. Hence, a progeny test is adequate for evaluating an individual’s performance. However, open-pollinated species are heterozygous plants and are further pollinated by other heterozygous plants growing with them in the field. Progeny testing is thus not adequately evaluative of the performance of individual plants of such species. A more accurate evaluation of performance may be achieved by using pollen (preferably from a homozygous source – inbred line) to pollinate the plants. As previously described, the method
of evaluating the performance of different mother plants in a comparative way using a common pollen source (tester line) is called a test cross. The objective of such a test is to evaluate the performance of a parent in a cross, a concept called combining ability.

This means that doing a TLC or a simple alkaloid concentration test on an individual does not say a lot about that individuals genetic capabilities for providing a high yielding cultivar for example. I think this is a really important takeaway! This effects also compounds with the environmental factors that cause a significant variation in alkaloid content, even with the same genetics. How do we separate environmental noise from the genetics?

A tester line therefore might be important to have, but I am not sure yet if it would be a requirement for genetic improvement towards a high yielding cultivar. It would complicate and slow down the process. So it is a tradeoff, but it might be worth it! If we decide to do so, the community would need to share this same tester line (clones). Relying on a single line is also not good so if we do so I would propose 2 tester lines.

TO BE CONTINUED!!!

Feel free to give feedback and correct mistakes (Including spelling and grammar), the eventual goal, for me at least, is to collect all available information and make a single overview of crucial information that would serve as the foundation for our at home community breeding program. It seems to be a delicate balance, it needs to be reasonably effective, while keeping complexity, time and effort to a minimum. If we make it too complex, but effective in theory, it might not be achievable for everyone making it ineffective in practice. If we make it easy and accessible, but rather ineffective in theory, we might not get anywhere.


edit:
At the end of chapter 6, another good to know:

By the same token, if a cultivar is produced by a process in which controlled cross-pollination is enforced, the only way to prevent the cultivar from returning to the natural way of being susceptible to cross-fertilization is to continue to enforce restricted pollination in its maintenance.

We luckily have a species that happily clonally propagates! This mainly concerns seed production

 

[Ivan Illich]

Section 16.5, page 307 is a good read for the most simple method called "Mass selection"

Key takeaways:

1. Mass selection is most effective if the expression of the trait of interest is conditioned by additive gene action. (chemotype is not)
   
   
2. in cross-pollinated populations, gene frequencies are expected to remain unchanged
   unless the selection of plants was biased enough to change the frequency of alleles that control the trait of interest.
   
   
3. Mass selection is based on plant phenotype. Consequently, it is most effective if the trait of interest
   has high heritability.
   
   
4. Cultivars developed by mass selection tend to be phenotypically uniform for qualitative (simply inherited) traits that are readily selectable in a breeding program. This uniformity notwithstanding, the cultivar could retain significant variability for quantitative traits. It is helpful if the selection environment is uniform. This will ensure that genetically superior plants are distinguishable from mediocre plants.
   

The case for mass selection

Advantages

1. It is rapid, simple, and straightforward. Large populations can be handled and one generation per cycle can be used.
   
   
2. It is inexpensive to conduct.
   
   
3. The cultivar is phenotypically fairly uniform even though it is a mixture of pure lines.

Disadvantages

1. To be most effective, the traits of interest should have high heritability. (As we know, the trait 'alkaloid content' has low heritability, as the effect of environmental variance seems to be rather high, see chapter 4. This points back to point 3 of the advantages, and the next point here)
   
   
2. Because selection is based on phenotypic values, optimal selection is achieved if it is conducted in a uniform environment.
   
   
3. Phenotypic uniformity is less than in cultivars produced by pure line selection. (So if we want to shoot for the stars, other methods are better)
   
   
4. With dominance, heterozygotes are indistinguishable from homozygous dominant genotypes. Without progeny testing, the selected heterozygotes will segregate in the next generation. (If there are cases for where this matters, and this forms a problem progeny testing can be done to determine what the homozygote is.)

I think that this breeding method is really nice for improving general quantative traits, and for traits that are causes by addative gene action, like cold tolerance for example. However for making stable chemotypes. This method is less effective, here controlled crosses would provide the fastest, and probably most easy method.

 

[Ivan Illich]

In this case, I think mass selection (which is what I suspect @Grasshoppers has been doing), is a good method for improving quantitative traits, like total alkaloid content for example. And it seems like you have already succeeded in this!!!! Amazing job.

All three types have been preselected for high yield—there are no low-yielding genetics in this population. If you grow and extract from this seed population, you'll obtain a mix of all three types.

I am proposing a theory for the chemotypes that we are seeing, and things we could do to test that theory. Note that this relies a lot on heavy speculation, and probably some errors of thought. Please be fierce in discussing this.
I suspect that the synthesis of 5-meo-DMT, which has not been uncovered in Phalaris is a dominant trait, which likely involves 2 steps https://www.science.org/doi/10.1126/sciadv.aeb3034. DMT --> Bufotenin --> 5-meo-DMT. Alternative routes seem possible too, but perhaps less likely. Since there has only been a single paper reporting bufotenin in Phalaris I think that these 2 genes are tightly linked and perhaps are close together on the chromosome. Here the rate of the second reaction that yields 5-meo-DMT could be a lot higher than the reaction that yields Bufotenin from DMT. Which would make the reaction rate limited at the first step, leaving Bufotenin in a chronically depleted state.

For simplicty of research we can treat the two genes that yield 5-meo-DMT as one locus, with two alleles, and see if that works. for chemotype 3, I think the case is here is that were seeing incomplete dominance of that trait. I will further explain what I mean:

I will denote the allotetraploid as G₁g₁ / G₂g₂. where G₁g₁ cannot combine with G₂g₂, so two seperate diploid systems that are linked. I am assuming that both G1 and G2 carry these genes.
Lets call the allele M, for 5-MeO-DMT, but keep in the back of your mind that it also contributes to the bufotenin middle step. and the DMT phenotype has alle m.

type 1:
m₁m₁ / m₂m₂.
The genes to yield 5-MeO-DMT are absent. producing the chemotype 1. DMT dominant.

Type 3:
M₁m₁ / m₂m₂
m₁m₁ / M₂m₂
The genes are present, but in a sufficiently low dosage that only a partial amount of DMT is converted, leaving the plant with both.

Type 2:
M₁M₁ / m₂m₂
m₁m₁ / M₂M₂
M₁m₁ / M₂m₂
M₁M₁ / M₂m₂
M₁m₁ / M₂M₂
M₁M₁ / M₂M₂

as soon as a double dosage is reached, the rate of conversion is too high for DMT and 5-meo-DMT to co-exist.

I suspect, that if you cross two DMT producing plants, only DMT producing progeny will be yielded, there are no dominant alleles.

If you cross Type 3, the resulting segregation pattern will be 1/4 DMT, 1/2 both, and 1/4 5-meo-DMT

Crossing type 2 does not yield a distinguished segregation pattern, but the segregation pattern of F1 can tell you if you had a double heterozygous.

If types reproduces among their own groups:
Type 3 will never be able to be true to seed.
Type 2 will only be true to seed if the the plant had a double dominant, or more.
Type 1 will always be true to seed

I think this might serve as a good theory to test out, but that is only after we rigorously check the possibly awful assumptions that I am making here.
I thought it fit nicely as we see that type 2 is the dominating chemotype, and the lack of bufotenine in Phalaris.

N,N-DMT and 5-MeO-DMT appear to occur mutually exclusively within individual Phalaris aquatica plants—high concentrations of both compounds have not been observed in the same specimen.

This also fits relatively nicely, as Type 3 is relatively rare.

Interestingly, if bufotenin is actually used as an intermediate step, it could be possible to breed for it. But if its responsible gene is tightly linked to a functional O-methyltransferase gene. It would be brutally hard to do so. especially if you don't see the phenotype already within you population.