Breeding New Fruit Tree Landraces

Yeah I’d say that all fits with the way I was using the term, except maybe for:

History is a subjective thing of course. But if a person or group of people spend time adapting a population over multiple crop generations to their land, then it has a history. Most landraces have a longer history. But everything starts somewhere. Like, most countries have a very long history, whereas some countries are very new. But they all have some kind of history, just some have a longer one than others.

As for grex, I think of a grex as a population all deriving from the same two original parents, like being siblings, where as a landrace can have potentially many more. Also my impression is that a grex might not have any specific adaptation to a particular land. So, very different from a landrace. And I would think that a grex might be a good population from which to start a landrace project, though I would rather start with something much more diverse, which I would tend to call a ‘hybrid swarm’. I’m not even sure of the strict definition of ‘hybrid swarm’, but I personally use the term to refer to a situation where one has made many F1s from many different parents, and then if possible crossed all those F1s (or maybe F2s or whatever if one wants to select a bit first) with each other, to make a richly diverse population. Welcome to correct me if I am wrong with the way I am using that term.

The kind of landrace I have been discussing is one where we’re trying not to homogenise too many phenotypes. But, my impression is that there can be quite a spectrum of this. Such that some traditional populations labelled by botanists as landraces, might still have some phenotypic variation, but way less than the approach I am advocating for, which some people might consider to be a ‘narrow range of traits’, though ‘narrow’ is of course a very subjective term. That’s my impression anyway but feel free to correct me if I am mistaken.

Actually, on that topic, in the video of Dr. Debal Deb, whose work is absolutely admirable, I did as it happens find myself wondering, when he was talking about the traditional method of maintaining rice landraces in India in the video I shared above. He was talking about a very strict method of maintaining them by selecting for… it sounded like pretty much all their phenotype traits! I couldn’t help wondering, is that really how they were maintained in pre-colonial India? Or has he perhaps been influenced by the more modern Western ‘heirloom’ trend which I have the impression is more fanatical about selecting against ‘off-types’. Like, it could perhaps be that the native Indian landrace culture really was that strict, and their landraces really were that homogenous. But at least that is not the level of strictness I get the impression some landraces were maintained with in other cultures where it seems to me they were more diverse, and more in line with what you just said about that. That part of his lecture did conflict with my own understanding of landraces in general.

By the way what is your source(/es) for your criteria?

I would qualify that by saying that it could do just that on the breeder’s land! Or on their and their neighbours land! That’s after all how so many of the ‘old’ landraces also came into being!

Of course this also brings into question, what about if I buy seeds from that breeder’s relatively new landrace, or for that matter, from a 1,000 year old landrace from some ‘exotic’ area straight from the indigenous people of that area - is it still a landrace now, in my fields? This could perhaps be merely a linguistic issue. But from one perspective, we might have to wait until it adapts to our land (if it does), to call it a ‘landrace’ again, at which point it might be then a new and different landrace! So perhaps we would have to say it’s ‘from’ a landrace until then.

But yeah, my focus is on making new landraces. The only crops I’ve grown ‘from’ old landraces are one tomato landrace from Colombia, and some rice landraces. Both for the purpose of breeding. I do hope to make a nice new rice landrace!

Another thing that comes into this is for example Joseph Lofthouse’s ‘landraces’. When they are community projects bred in various lands which are quite different from each other, I find myself not thinking of them as landraces at all, but rather, more as hybrid swarms. But the ones which are very much focused on his land, but with phenotypic variability considerably more than traditional landraces, I find myself unsure of what term to describe them by. But some of them I really can think of as landraces, really suited to his land and very successful there, often way more than any other variety there is, from the work he has done. But I see it as his kindness to deliberately keep the diversity considerably high specifically to enable the sharing of them over a wide spectrum of lands. So the aim overlaps with the traditional aim of landraces, but has that added factor of consciously wanting to share them over a wide area. in one way almost ‘proto-landraces’, or perhaps the ones bred communally in quite different places as ‘proto-landraces’ which do have some specific adaptation but are basically very highly suited to sharing for other people to use in starting their own landrace projects. But the ones more specifically adapted to his place but with a consciously shareable level of diversity, perhaps simultaneously ‘landraces’ and ‘proto-landraces’. Well anyway, all these labels are us humans just imagining conceptual boundaries in a boundless world. But anyway, there are some of my thoughts on it. It’s a rather uncommon thing to be deliberately making diverse populations to share with people, so no wonder the language isn’t quite settled!

1 Like

I think you’re conflating the idea of a stabilized “heirloom” cultivar that is generally true to seed with a “landrace” (geographically isolated population sharing certain traits).

Landrace has built into it the idea that it’s been basically the only “type” within a bioregion for long enough that all the various traits within the population have relatively stabilized due to a long period of geographic isolation from other regions where the species grows.

Using the same term for a population that has been bred in a single person’s control on a single parcel of land seems like a misuse of the term. That’s just a new cultivar or breeding pool that was selected using a method of breeding that is attempting to borrow from the process by which landraces form. It is not a landrace itself.


I disagree. I believe I have done by best to differentiate the two concepts, including the issue I specified I had with Dr. Debal Deb’s strictness of selection which should show I’m not conflating them. Though I also don’t think the boundary between the two is terribly black and white when it comes to how precisely how stabilised they might be.

That’s what I mean, in terms of ‘relatively stabilised’, and just how stabilised, can vary. But I strongly disagree with the part of that I put in bold. I’ll give some examples. Let’s start with Laotian rice landraces.

From the paper ’ Genetic diversity and population structure of ‘Khao Kai Noi’, a Lao rice (Oryza sativa L.) landrace, revealed by microsatellite DNA markers’ by Koukham Vilayheuang et al.:

Rice (Oryza sativa L.) is the main food for people in Laos, where it has been grown and eaten since prehistory. Diverse landraces are grown in Laos.
From 1995 to 2000, the Lao Ministry of Agriculture and Forestry and the International Rice Research Institute (IRRI) collected 13,192 local accessions of rice throughout the country (Rao et al. 2006c).

In short, there are many many landraces of rice in Laos! And they’re not isolated in their own unique regions.

From the paper ‘Naming of traditional rice varieties by farmers in the Lao PDR’ by S. Appa Rao et al., we see:

The Lao Ministry of Agriculture and Forestry (MAF) and the International Rice Research Institute (IRRI) jointly
collected rice germplasm in the Lao PDR between 1995 and 2000. Among the 13192 samples collected, there are
3169 distinct variety names.

Copy and paste in that document is an issue but for example they found 462 distinct variety names just within Luang Prabang.

In the paper ‘Rice Landrace (Oryza sativa L.) Field Survey and Collection in Xiengkhouang and Houaphan Provinces of Laos in 2014’ by Koukham VILAYHEUANG et al., we see:

The provinces of Xiengkhouang and Houaphan, located in the north eastern region of Laos, are
well known as major producers of Khao Kai Noi (KKN), a traditional rice variety. In order to observe the
diversity of rice, especially of KKN in these two provinces, we conducted a survey and sample collection
during the harvest season from October 16th to November 1st, 2014. We visited paddy fields where KKN
was found. We identified and interviewed the farm owners and requested for rice panicles. We majorly
focused on the collection of KKN but also collected other traditional rice varieties that were grown in the
same farmer’s paddy field.
Genetic diversity and population structure may change with time and human
activities and for efficient conservation of such valuable rice, recent diversity of its on-farm population
should be studied to allow comparison with genebank materials. Therefore, in this survey, we collected KKN and other traditional rice varieties (landraces) from farmers’ fields during the harvest season in Laos.
Most farmers grew more than one variant of KKN in their single-family paddy field because they
can serve different purposes, for example, KKN Lai could be sold for a higher price, but other KKN forms
could be produced in higher yield (according to a farmer in Xiengkhouang). Some families grew KKN and
other rice on separate plots in the same field.

From the paper ‘Genetic diversity of the wx flanking region in rice landraces in northern Laos’ by Chiaki Muto et al.:

The rice fields in northern Laos consist mostly of upland fields and partly paddy fields. Recently, some modern rice varieties have been introduced to paddy fields as cash crops, but not for private consumption (Ishikawa et al. 2002). However, the majority of the fields are cultivated for glutinous landraces mostly on 1.5–2 ha farms in hilly areas that are owned by single families. These landraces are sown as seeds harvested by the farmers themselves. Multiple landraces recognized from their morphological appearance are cultivated in single fields and later identified molecular bases (Ishikawa et al. 2002, Muto et al. 2010).

In Yunnan, China - in the paper ‘Estimating Genetic Diversity of Rice Landraces from Yunnan by SSR Assay and Its Implication for Conservation’ by ZHU Ming-Yu et al.:

Eighty-five rice (Oryza sativa L.) varieties, including 82 rice landraces collected from 17
villages in Yunnan Province of China and three standard varieties representing typical Indica and Japonica
ecotypes, were studied using simple sequence repeat (SSR) markers to estimate their genetic diversity for
the purpose of strategic conservation
A total of 85 rice (Oryza sativa L.) accessions were used
in this study for the SSR analysis, including Indica and
Japonica ecotypes, as well as glutinous and non-glutinous varieties. Most of the varieties were local landraces
commonly cultivated in Yunnan Province. Of these, 74 accessions were collected directly from farmer households
scattered randomly in 16 villages

This is about naming but it gives you a glimpse into the number of varieties grown within single villages in Gambia, from the paper ’ Mechanisms Explaining Variety Naming by Farmers and Name Consistency of Rice Varieties in The Gambia’, by Edwin Nuijten & Conny J. M. Almekinders:

Later, in 2002, a questionnaire was administered in 10 villages (Fig. 1) which provided data on number of varieties grown, cultivated area, seed loss, seed distribution, seed sources, and variety names. In conjunction with this questionnaire, farmers were asked for seed samples. As many rice samples as possible were collected (in total 297). The number of samples provided per farmer varied from zero to five.
Given an early observation during fieldwork that different farmers sometimes give different names to the same variety, the assumption was that the number of variety names given by farmers would be much higher than the number of identified varieties. The total number of varieties identified in this study on the basis of seed sample comparisons was 102, which is 80% of the total number of variety names (129) (Table 2). However, the situation varied between villages. In Jiroff, for example, 11 varieties were identified, while 15 names were linked to these varieties. In Massembe, 11 varieties were identified and 10 names given (Table 2).

From Nepal - in the paper ‘Genetic diversity study in landraces of rice (Oryza sativa L.) by agro-morphological characters and microsatellite DNA markers’ by Bajracharya, Jwala:

The genetic diversity of 632 rice landraces was studied from three ecosites representing three agro-ecozones of Nepal. There were 147 landraces from Jumla (2240-3000 m altitude), 291 from Kaski (668-1206 m) and 194 from Bara (80-90 m).

I was actually searching for the paper which went more into detail about Laotian rice landrace practice which I mentioned in an above comment but couldn’t find it. It was something like around 6 or so landraces being grown by each farmer and something like 12 or 20 per village, something like that anyway. But you can see from the various rice landrace examples from different countries that in fact many landraces are grown even in the same village or even same field, let alone bioregion. And you will find the same with other crops - I mentioned potato landraces in the Andes for example. I don’t have time to find examples but it should be fairly easy to look up if interested.

To your next point:

Why? What in your concept of landrace requires that there need be more than 1 individual involved in the originating of a landrace? You don’t think that for example many potato landraces and rice landraces originated from a single farmer in Laos or the Andes?

I’ll give you a modern Western example in case it helps - see this page for ‘Balcaskie Landrace wheat’, developed by Anders Borgen:

It explains:

Balcaskie Landrace wheat is a mixture of long-strawed wheat varieties grown together, harvested and re-sown in order to harness the adaptive power of natural selection in a particular landscape, in this case the Balcaskie Estate in Fife, Scotland.

Let’s examine this criticism. Suppose we go to Laos, and find a 70 year old woman who has been growing many rice landraces all her life. And among the 12 landraces in her fields, she has 1… let’s for now use a broader term ‘variety’ which is as genetically diverse as her other landraces, but is distinct in some way, and she has named this population 30 years ago and her village love it and eat it, and they consider is as a variety of the same standing as the others, not thinking all her other ones are 1 category and that one different just because she is the one who brought it into being. This is actually quite common, you can see this in at least one of the papers I quoted from above. So then, will you call all of them landraces except for that one because she’s the one who made it, named it, and is growing it?

If so, then at what stage to do consider it a landrace? Does she have to die first? Or does she have to give seeds to her neighbour and only the neighbour’s population you will consider as a landrace, not the population on her own land? Or does it have to go to the next village first? Or do you put a time limit on it, insisting 100 years has to pass?

And to your point “That’s just a new cultivar”, well no, the examples and principles I gave in prior comments to which you were giving that conclusion, would be too diverse to be called a ‘cultivar’. You say “or breeding pool that was selected using a method of breeding that is attempting to borrow from the process by which landraces form” - well, yes. But ‘borrowing’… no I would disagree. I wonder if this is a race issue? That might be resolved if you answer the Laotian example question - does it make a difference to you if the subject is non-Western? I do not think it should. There’s no requirement that a landrace can only be a landrace if it comes from a non-Westerner.

1 Like

It appears that there has been much debate among scholars about what is and isn’t a landrace, with many arguing that the term has been overly broadened and misused, while others argue for new definitions even broader in scope. It seems you are on the side that wishes to broaden the term.

Even in the camp arguing for broad meaning, the “historical origin” is almost always emphasized, meaning a new “variety” that was selected based on being well adapted to local conditions would not be a landrace. For example:

A landrace is a dynamic population(s) of a cultivated plant that has historical origin, distinct identity and lacks formal crop improvement, as well as often being genetically diverse, locally adapted and associated with traditional farming systems.

Defining and identifying crop landraces

While some recent papers have argued for a ridiculously broad definition of landrace that might encompass what you suggest, it is by far not the accepted definition as even the paper authors admit:

We propose a more inclusive definition of landraces as plant materials consisting of cultivated varieties that have evolved and may continue evolving, using conventional or modern breeding techniques, in traditional or new agricultural environments within a defined ecogeographical area and under the influence of local human culture.

Toward an Evolved Concept of Landrace


It might be a “variety” definition you’re going for, Justin. Modern breeding for stable, seed-born, similar individuals. “Hansen’s Bush Cherry” comes to mind, Mr Hansen bred prunus besseyi by the 10,000’s discarding unwanted traits each generation until he could release the stable , genetically similar line he named the “Hansen’s Bush Cherry” variety.
And I could easily be wrong, been over 50 years since I was in school for the subject! :blush:


Well I personally feel that I am using the term in conformity with the way it is used in the many academic papers I have read on the subject. Admittedly, my reading is biased towards rice, potatoes, and a few other crops. So who knows, perhaps there are different crops (or animals) for which scholars tend to have a somewhat different usage of the word. But what I would say is that words’ meaning is dependent on their usage. For example, nobody (I know of anyway) says ‘gay’ doesn’t mean ‘homosexual’, because in some contexts it means ‘happy’. The word has meaning by the way the word is used. And I feel as if I have provided ample evidence for the usage of the word from many sources in botanical academia to support the way I am using it. By which I mean to say, if my use is in conformity with its already widespread usage at least in those circles I referenced, there might not be any need to ‘broaden’ it. If it is already sufficiently broad. Don’t you agree that I refuted some of your points quite satisfactorily?

So then what is your cutoff point? Exactly how many years has to pass before it qualifies for your criteria? And I ask that seriously, because without quantifying that, it’s kind of like saying nothing, right? Like, it’s not as though the 13,192 identified Laotian rice landraces all come from the ancient past - they are constantly making new ones! So if you are going to insist on a time requirement, then we should be open about what exactly that time requirement is.

What’s that meant to mean, ‘formal crop improvement’? Improving crops is very much a concern to traditional farmers. I can bet you that countless landraces have come into being because the farmer saw something in the plants that they considered an improvement, and consciously selected for that and made a new population on that basis. Let’s look again at that paper ‘Naming of traditional rice varieties by farmers in the Lao PDR’:

Some names indicate higher yield potential (broken store Lav tek, heavy grain Phannak, twelve grains yield one million Moon lan). Farmers use names like Gnod nang (superwoman) and Gna thao (grandmother) to indicate high yielding varieties.

I mean, it should be really obvious that farmers select for improvements. But to give concrete examples, those landrace names make it quite explicit that yield was one such criteria that they were specifically interested in, and named new landraces after. And I know of examples from India that were selected for their beauty - beauty which is only seen by the farmer and not even there in the final product, the naked grain - which most modern breeders would not be interested in since it doesn’t fit the capitalist model, but for sure some traditional farmers will be interested in beautiful grains in the fields and consider such things as ‘improvements’ even if a cold economist would not.

To give another example from that paper:

Good grain quality and aroma is called by such romantic names as Nang nuan (sweet and soft), or Hom nang nuan (sweet smelling soft lady).

These things are not random. They’re deliberately selecting for ‘improvements’. Here are some more examples:

Lao farmers have selected varieties that have resistance to commonly occurring stresses such as drought or flooding. Drought is important in both the rainfed lowlands and uplands. Drought tolerant varieties are called Bo ngo nam (does not care for water), Khok (upper terrace), Ea phon or phon (ant hill), among many others. Floods are a common occurrence along the Mekong River and its tributaries. Variety names indicate adaptation to flooding: for rapid stem elongation there are names such as Loy loy or Phou (floating or swimming rice), and Loy pa or Louk pa (floating fish). Many of the traditional varieties grow very tall and lodging is a major constraint.Varieties that do not lodge are called by names such as Kok lek (iron stem), Bong (bamboo stem), Aev (flexible stem) and Tia or Tam (short plant) for example. They also indicate adaptation to poor soil conditions, such as Bo hina (any field). Resistance to biotic stress like birds (hidden against birds Li nook), and good competition with weeds (win over weeds Phae nya) are common.

Maybe it’s about the word ‘formal’? Maybe this is not considered ‘formal crop improvement’ because these skilled farmers don’t wear lab coats or suits or write academic papers?

Also this bit about “associated with traditional farming systems” seems totally arbitrary to me. Suppose an expert Laotian farmer moves to France, and develops their own varieties over decades there which qualify as landraces in every respect except that France does not have the tradition of growing rice in the way she’s growing it. So then is it disqualified because she’s not using ‘French’ farming methods? Or, is she passing the criteria because her method is in fact traditional Laotian farming, just that it is in France? If so, what if she isn’t Laotian - a French woman using Laotian methods, is the population then disqualified because the ethnicity of the farmer is what is critical? If so, what if a French person develops a landrace in Laos using Laotian farming, are they disqualified on the single basis that they are French?

Or what if Italians make a new wheat landrace but use a no-dig method, are they disqualified because the well established mainstream Italian method is to use ploughing and toxic chemicals?

Also from that paper ‘Mechanisms Explaining Variety Naming by Farmers and Name Consistency of Rice Varieties in The Gambia’, this part I put in bold is worth remembering:

Millet versus Rice (Outbreeding versus Inbreeding Crops)

While in The Gambia rice varieties are often named after people, millet varieties mostly have ecotype names (Nuijten 2005). This difference is related to the fact that rice varieties are plentiful and are replaced relatively frequently, while millet varieties are few and not often replaced. Because of the small number of available varieties, millet farmers use the ecotype names (which have become part of the language) for naming millet varieties, whereas ecotype names cannot be applied to rice, which has many distinct varieties per ecology.

And I don’t have time to search for the part in one of the above papers about the frequent new creation of rice landraces in Laos, but, my most concise response to this idea of " that has historical origin , distinct identity and lacks formal crop improvement , as well as often being genetically diverse, locally adapted and associated with traditional farming" is that, like I said before, all breeding has a history. And deliberate improvement is a fundamental aspect of the development of new landraces in ‘traditional’ cultures, like the ones I have referenced above.

Also it’s worth noting that many of the landraces in Laos that these eminent rice research organisations define as landraces, are grown in quite varied conditions, like some of them being grown both in upland and lowland conditions, the two contrasting rice cultivation methods. So, as I have clearly demonstrated above, it’s not just that it’s false to think there can only be 1 landrace of a crop in a single bioregion - it’s also the case that one landrace can be grown in multiple quite different bioregions. And while that might seem somewhat contradictory to the idea of landraces being characteristically adapted to specific conditions, it makes sense when we realise that some landraces can of course do very well in a variety of conditions.

No I don’t think so. That’s precisely why I just gave so many references specifically to the academic use of the term ‘landrace’ from various different cultures and countries. I could give more from Andean potato landraces but I feel like I have made my point enough just with rice. I would say ‘variety’ is a broader term than ‘landrace’ and can be applied to landraces, but ‘landrace’ is more specific. Not all varieties are landraces. Also ‘variety’ is more easily understood by the public so sometimes that word might be used instead when referring to landraces. I will quote again from that paper ‘Rice Landrace (Oryza sativa L.) Field Survey and Collection in Xiengkhouang and Houaphan Provinces of Laos in 2014’ to give an example:

Also these are not from random unqualified sources - that was published in the Annual Report on Exploration and Introduction of Plant Genetic Resources.

I would really hope that it’s obvious that with regard to the breeding methods I have been advocating for, I have been talking quite specifically about populations that are quite deliberately genetically diverse. That’s been the very core of my focus throughout this whole discussion.

1 Like

Justin, you are missing - deliberately I think - the key part of a landrace. A landrace has been grown in a region long enough to have lost many traits not needed in that region. In other words, it is more fit for growing in the region of origin than other breeds or landraces. What you are breeding is not a landrace.


FYI, “Isolated” need not mean it is the only form of a species available within a certain country. It need only mean that it is reproducing isolated from other forms; at least in so far as to allow it to persist as identifiably distinct from them (it wouldn’t mean that no members of the landrace ever came in contact with other types, but rather this is a generalization of the overall population). Take for example the many ancient types of corn in North America. Different groups maintained different landraces of corn. They couldn’t be attributed to specific breeders, but rather to peoples of differing locations and cultures. These unique differences between landraces would have undoubtedly developed as a result of cultural preferences in combination with environmental pressures. It’s a bit different than a breeder trying to push a population to reach their personal goals.


I would say that’s a generalised observation rather than a rule - would you not? I mean, you say “what you are breeding” - what makes you think what I am personally breeding doesn’t fit that anyway? I have been talking about breeding specifically for one land, but with enough diversity to be able to adapt to the conditions of that land as our climate continues to change, so I don’t know where you are seeing contradiction there in the method I have been advocating for regarding fruit tree projects, nor about my own work on tomatoes. Naturally when you start with a highly diverse populations such as a hybrid swarm made from landraces gathered from around the world, or my interspecies breeding method, then yes as I keep saying, the initial extreme diversity quickly settles down to much less diversity, in that period of rapid adaptation as it transitions from a hybrid swarm towards becoming a landrace - a deliberately fast way of creating a landrace - so there will be a rapid loss of many traits as it adapts to the local conditions and then a further loss of traits when you start selecting for phenotype after or during that process. So I can’t see how you’re saying I missed that, I think I covered that in detail. But, we have the option of how far to take that. Landraces are still adaptive, and we have the choice how far to narrow the diversity, and how limited they are to regions/conditions.

I’ll give you an example from one of the papers I quoted from above, ‘Genetic diversity study in landraces of rice (Oryza sativa L.) by agro-morphological characters and microsatellite DNA markers’ to show how some landraces in this paper are only grown in single conditions whereas some are suited to more varied conditions:

Farmers in Kaski identified four different domains for rice. These were Mule
khet/Kule khet (irrigation by seasonal canal), Sim khet (marshy wet land), Tari khet
(rainfed good fertile land) and Pakho tari (completely rainfed marginal uplands). Kule
khet ranked the first in production potential and followed by Sim khet, Tari khet and
Pakho tari (Rana et al., 2000b). The distribution and diversity of landraces varied greatly
between these domains (Appendix 4.3a, Figure 4.6). Kule khet and Sim khet were the
most favourable domains for rice, and had the greatest diversity of rice. Tari and Pakho
tari were two domains where water was limiting and they were both less productive and
less diverse. Out of the 69 landraces, 38% (26 landraces) were specific to a particular
> domain while 62% were grown in two or more adjacent domains. Jhinuwa, a small grained, aromatic rice, was the only one reported to be grown in three domains: Tari,
Mule khet and Sim khet.
Out of 21
landraces reported in the survey, 13 (62%) were specific to domains while 38% were
grown in adjacent domains (Appendix 4.3b and Figure 4.7).

Also to your point

Let’s think about that. We know for a fact that many landraces for one crop are grown in the same region. I have shown that they’re often grown even in the same field! By your logic then, for all those (which would be many) that originate from the same region, only the single fittest landrace in that region would qualify as being called a landrace, and all the others would be disqualified! But evidently this is absolutely not the case, as demonstrated by the above papers, and countless papers on other crops such as potatoes and so on. So I ask you, where are you getting that idea from? And what is your explanation for that severe contraditions - are you saying all of these botanists and landrace experts and top international organisations are wrong in their use of the term ‘landrace’?

Furthermore, according to your logic, even the most ancient landrace of a region would be disqualified from being called a landrace the moment any local farmer were to come up with a new landrace that were more fit than the old one, or even to import one from elsewhere. I find it hard to believe that that is your view.

For whose information? I can’t tell if you are directing that at me or at @swincher . And I don’t know why you mentions countries - that is an arbitrary political boundary, so I will not focus on that category but anyway to you point… They specifically wrote:

So I agree with you, I’ve demonstrated multiple landraces being traditionally grown even in the same fields, and many landraces per village or farm.

I don’t know about corn, but I would not be surprised if also individual groups or families also grew multiple corn landraces.

Sure. Especially after the genocide. But let’s see that paper ‘Mechanisms Explaining Variety Naming by Farmers and Name Consistency of Rice Varieties in The Gambia’ again:

To study consistency in variety naming, the logical starting point is the character of variety names. Various researchers reported for a range of crops in different parts of the world that farmers name their crop varieties after plant traits or for the person who introduced or first encountered the variety.
From these examples, we conclude that new rice varieties introduced into a village which are adopted by many other farmers of the same village in a short period of time are named after the person who found or brought the rice variety to the village. Naming a variety after the person who introduced it can be perceived as giving credit to that person. After a period of time, maybe 20 years or so, the variety gets a new name based on its distinctive morphological, agronomic, or culinary traits. Possibly, when varieties get more widely diffused, the actual origin loses its meaning and is forgotten.

That’s just one example. But even in cultures were new landraces are not named after the person, the person instead naming it by its traits or some romantic name like some examples I gave from Laos, for a while at least, many people still might be aware of who started it off.

How can you explain traditional farmers deliberately coming up with new landraces by selecting for traits that please them, as we continually see in traditional landrace cultures, if you ban the application of ‘personal goals’ from the equation? Would you really strike from the list of landraces, any that came about because the farmers brought them into being because they fit with their goals for taste, aroma, disease resistance, yield, performance in drought, and so on? I can’t see logic in that.

1 Like

So apparently this was indeed about lemons and the story is reported by ‘David the Good’ in his book ‘Push the Zone’. I don’t have the book so was unable to find the name of the person.


A landrace is s landrace. How people use them does not define what a landrace is. Seems use is being used to redefine the meaning.
Looking at the subject of this thread the last thing I would want is a landrace I’m looking for diversity in flavor and ripening times. I want a mix of flavors that ripen throughout the growing season. This is a huge influence in what I breed.
I like growing white, yellow and red fleshed nectarines. The diversity is very appealing. I have a red fleshed nectarine that ripens in august and a red fleshed peach that ripens in
October I would like to cross and back cross to get a red fleshed nectarine that ripens in September. At the least it would be fun trying.
It seems to me the only good reason to stabilize genes in fruit trees would be for use as a rootstock. I know citrus is different. I can’t grow them so have no interest. Even if I could I’m not that interested in citrus which is nothing like the fruit trees I do grow.
What would be the point of a fruit tree landrace? What’s not covered by the multitude of cultivars out there? Seems to me finding a cultivar with desired traits and cloning it would be the way to go. I guess just about every tree nursery does just that. Again the only need for stabilized genes is for rootstock like Lovell peaches you can grow as seed for rootstock.


Well if we take the same approach that dictionaries generally take, then actualy usage defines meaning. Take the article ’ The Oxford English Dictionary, like the English language, is all about change’ from the National Post for example, which states:

Samuel Johnson once said a dictionary should aim to “not form, but register” the language. Indeed, a dictionary should “not teach men how they should think,” he continued, “but relate how they have hitherto expressed their thoughts.”

We tend to think of our dictionaries as tools of instruction; as books that set the standard, with a certain air of the definitive, for how our words ought properly to be used. But that was never the intention. The great lexicographers understood that the ideal campaign of a worthy dictionary was descriptive, not prescriptive. It would simply record what we’ve already agreed upon socially. It would, like the English language, live and breathe and change.

That’s why I have been making my argument against those who are saying I am using the word incorrectly, by refuting their points with direct examples from all of these professional publications showing their use of the term.

It sounds to me like you might therefore not want a landrace that doesn’t have those qualities, but that doesn’t have to mean you would not want to make your own landrace within those parameters. Or if some would consider that ‘too diverse’ to be called a landrace, no problem, let them call it whatever they want but you could still use this method I have been advocating for but calibrate your phenotypic selection process to allow for that variation you want, while still ending up with a population that is specifically well suited to your land.

Or if that would be too unpredictable for you and you might not like the various fruits that are in between the types you specified, various different blends of traits, then you could take the more common landrace approach of making (or acquiring) landraces that individually suit those goals and grow them all on your land, just like rice farmers grow different rice for the different dishes they love to cook etc… Similarly people can grow or create multiple tomato landraces, for example having some cherry tomatoes, some beef tomatoes, some paste tomatoes, some storage tomatoes, and on another variable, earlier and later ones to spread out the harvest.

I’m actually trying to work not just with SI systems but UI systems, so that outcrossing domesticated tomato landraces can be grown in groups neighbouring each other but remain distinct from each other, for example. Though one could also use larger distances to isolate them if they didn’t have inbuilt separation via UI (unilateral incompatibility), like is done with various other SI crops for seed production. I just prefer if they have inbuilt protection from mixing so that it makes things easier for other growers.

I don’t know much about SI and UI systems in fruit trees, but if there are crossable species, this might be possible - for example if fruit tree A can be a pollen donor to fruit tree B but the cross can’t go the other way, it could be possible several BxA crosses, and through selection, breed a population from them that can cross with each other but not receive pollen from B. Such a population would then be possible to grow as an interbreeding population alongside populations of B but still being genetically isolated from them. In my tomato case, that means my BxA tomato population would be protected from dilution from B, B being domestic tomatoes. So far, I have an F1 generation that is behaving exactly like this, so the next step will be to try to select in F2 for that characteristic. If there are a great many that pass that test, I might also select for other traits at the same time, maybe environmental performance and/or fruit characteristics; or if there are too few, I will wait until later generations before doing so, to not prematurely limit the diversity. And in parallel I am working on a different population with a different isolation system being brought over from another wild species, let’s call it BxC (though it’s more like Bx[DxC] since it uses a bridge species) with the hope that that can be lead to a population that is genetically protected both from B populations and from BxA populations, just like A and C are protected from B.

But it also might be the case that none of these options would suit you, and you’d rather grow from grafts or from seed but not have any continually evolving seed-grown population ongoing, adapting to your land, and that would be fine too, of course. In particular, growing multiple highly or totally outcrossing fruit tree landraces in one area which are cross-compatible, would sound like it would only be normally applicable if there were suitable separation distances between populations, so that might rule that out for various people.

1 Like

From my vantage point, There are sort of two tracks to this discussion here.

One is just broadly can and should you breed for purposes of producing seedlings as an end in itself- not rootstock, but seedling “varieties”.

The second is the semantics piece- what constitutes a landrace vs variety vs subspecies vs species. That type of analysis can go around and around in large part because words are always fuzzy at the edges of their meaning. We all know there are reptiles, amphibians, and mammals, but damned if there isnt a duck billed platypus to stick in the craw of our otherwise perfect framework.

Heck, as Ive pointed out before, the species concept itself, useful as it is, is still not flushed out and an area of active debate.,-sexually%20re-%20producing%20organisms.

Just as with grammar, there are many “special cases”, and if you poke at them a bit you realize the whole thing is a construct, a useful shorthand. So working backwards from the definition as though it were immutable will never produce totally satisfactory analysis

Not to derail the conversation with a linguistic deep dive, but properly speaking, ALL words are metaphors. Etymology makes that clear. Its only usage that leads to a concept becoming reified as though it were an actual object. We may think those meanings are bounded, but just like zooming in on a cell wall, it turns out theyre permeable.


This seems like a good, reasonable way to approach the issue, and we could probably have a great thread if we focused on the first question instead of the second. But I suspect the next 100 posts will still lean heavily toward arguments over the second part.

I think it is a great idea to create a breeding population where the long-term goal is to allow people to propagate it by seed and usually get the same fruit trees, or something close enough. I’m imagining things like Indian Blood peaches, or Antonovka apples.

The problem with one person trying to do this with fruit trees in their lifetime is even with all the tricks you can concoct to speed up maturation, or even picking the most precocious fruit species you can find, that level of stabilization will simply require too many generations of trees for a single person’s lifespan. You could start with an already stable population and select out a few specific traits, but that’s not really the same thing as creating that stable population from a diverse starting gene pool, which is really the thing that was proposed, I believe.


Yeah, that was after all the real topic. Though I’d add that it was not merely about being seed propagated, but also about being genetically diverse, and both adapted to a land (or on the way to that) and adaptive to further climate change. And the inter-species aspect of the topic was a part of that diversity aspect. The ‘landrace’ label I believe fits quite perfectly to this but those uncomfortable with that term can just use another, like ‘adaptive population’ or whatever.

I’m glad you think so. As mentioned above, from my view, that can be a great aim, and having reliably delicious but still varied fruit can be good too. That can even add a really fun aspect to it. Right now I am imagining several neighbours getting together in excitement to taste the fruits of a tree’s first harvest, or excitedly swapping fruit from their all unique trees. What fun! Maybe even inventing hyper-local food dishes or wines/beers/ciders specific for the fruit of individual trees!

That really depends on what ‘level of stabilisation’ you are aiming for! Though also, how long did you guys say it takes for a graft to bear fruit if made from a seedling? Say it’s 2 years, you could do 8 generations in 16 years. So for example (I have already given this example but I will give it again), you could make a very diverse… ok I will put this in bullet points and I will gear this to a speedy method:

  1. Start with a selection of parents, which you might already have on grafted trees, or you might even use an already distributed collection among a network of friends etc., then cross them all. You could even do this as a community, even an international community if sending pollen by post or transporting it in person.
    These would be faster but the slower option would be to obtain diverse seeds from diverse places, go to step 2.
    And for this step, this could be selecting these many parents all from 1 species, or from different compatible species, or even from already existing interspecies hybrids, though making your own would give you more control, and potentially much more diversity since you could make many different crosses with the same 2 species.
  2. End of season, plant all those seeds, and grow them under good lights in warm conditions, so you might be able to graft them by next spring or whenever the grafting season is.
  3. When those are flowering - would that be that same year, i.e. year 2, or would it be year 3? Let’s suppose it takes another year before flowers, so in year 3, then cross them all, and repeat step 2.
  4. Year 5, you have the double crossed seeds. Lots of them. Repeat step 2.
  5. Year 7, you now have a large very diverse collection of grafts, all unique in the world. It could be hundreds, or more even. Now eliminate all that don’t have acceptable fruit.
  6. With the remaining ones, plant as many seeds as you like, from all of them. You’re now growing them in the ground, so the environmental natural selection will be starting.

You might still get fruits from the next generation that’s not acceptable, but you’ve increased your chances by eliminating those ones you didn’t like. This saves space. And using the grafting method up to this point, as well as indoor seedling growing, has saved time.

  1. If you want to save more space, to rush the project you could also take a cutting from each seedling as soon as they are big enough to do that, and graft them, making it such that you will be able to sample their fruit earlier, so if you want to, for those you don’t like the taste of, you can dig up the rooted plant.
  2. If you want, from those new grafts, for those you especially like, you could even cross them and grow them out, putting them in the spaces made free from eliminating the ones you didn’t like the taste of.
  3. Just keep repeating this cycle. Once they are deliciously stable, you don’t need to do any more grafts, though you might still just out of impatience at wanting to taste each new unique fruit.

It would be better to be as lenient as you can be with the taste tests (and especially so for the first year of tasting), since keeping more diversity will mean better changes of keeping environmental/disease capabilities. And, if one doesn’t quite meet your taste criteria but is already showing some other advantages such as vigour or cold resistance more than the others, you might consider keeping it in no that basis, or you might even deliberately cross it with some delicious ones and plant those seeds, and perhaps even remove that parent, and a couple of years on taste the offspring and they might be worth keeping and have some of that added benefit.

  1. Some stage after step 4, an additional way of speeding things up would be to deliberately expose your entire population to specific diseases you need them to have tolerance to, or environmental tests such as cold. Now, if it were me I would rather do this by cloning them all via cuttings, and using those cuttings once established as the test subjects. You could then remove from your population all of those whose clones failed the worst in such tests.

Of course you could just do it all simply, just make the crosses and plant the seeds and not use grafting at all, but yes, that will take a long time and would probably be a multigenerational project, which can also be great, but if you are in a rush, the above method could dramatically speed things up. You’d just need to keep an eye on the total diversity you are keeping, and if for whatever reason your population diminishes too much (for example if in your haste for specific disease resistance you eliminated a high proportion of your population), you could add new members from new sources and cross them into the population. You might even hear of a wonderful tree that suits your needs, perhaps an old neighbour heard about your project and shows you her tree - so great, add it in, see how that goes. In the first several cycles, this attitude could be quite helpful.

1 Like

Your idealism is palpable @Justin. Your ideas bespeak the likelihood that you are in a particular stage of things. Pardon my presumption, but Im guessing you may be in your 20’s, childless, unmarried. You’re considering mass plantings, which would suggest you have access to a piece of land, but if so, you havent owned it long enough to realize how much time and energy it can and will suck up just to keep it in stasis. If you dont yet have land, youve not realized the full extent of the BS and compromises you’ll inevitably go through to attain it.

If true, that doesnt diminish the veracity of your arguments, but I suspect some of the pushback youre getting comes from a place of having lived a little (or a lot!) longer, having less discretionary time, more responsibility, and having had expectations tempered significantly by experience. The older you get, among other things, the less long range you tend to think. Similarly, having a family and a home of ones own tends to temper ones tendency to consider the greater good vs tending to ones own. Both come from similar impulses, theyre just different expressions based on what stage of life youre at and how your lived experience evolves as you age.

Theres nothing about what you’re describing and planning that seems wrong or impossible, and I dont think you should worry too much about justifying its worth. As I said earlier, the proof is in the pudding. I see that statement as equally encouraging and cautionary. Since you are clearly driven to do so, I say pursue it and let us know of your successes and failures along the way. The hurdles others have mentioned are real for sure (and your optimism may be clouding your estimation of them) , but they could be far worse. You could be trying to re-speciate extinct tortoises from limited hybrid genetics! Id wager the age to sexual maturity of a tortoise alone makes tree breeding look like child’s play.


Glad to hear it :slight_smile:

Ha ha ha :slight_smile: These days I often finish work around 11pm or later. But I figure it’s worth it. So far this year I made (well this is just for the seeds I have processed so far so the actual number is larger) 72 different crosses between domestic tomatoes (some of which are already domestic-wild hybrids) and 6 different wild species (104 in total if not only including the crosses with pure wild species), and will soon be harvesting the seeds from those crosses, some of which have been crossed again - plus I have more crosses to make before the year’s end - it should be 3 cycles in total this year, though more than that may be possible). Not to mention today having spent all day doing interspecial pepper crosses. So yes you’re right, this kind of work can take a lot of time, but that depends on how much one wants to do and how much one shares the work out amongst a community, which can be a great way. I just decided that I wanted to do what would usually take several years, in one, and figured having an intense year would be worth it. Well, it might be an intense couple of years. But the way I see it, if you start strong, especially if making a hybrid swarm, the whole project has much better chances. So I don’t expect it to be like this once I switch from speed breeding, to environmental selection. I can let nature do most of the work then, like I have done with some populations this year also. Much more relaxing at that stage, mainly just about preparing the ground, and processing the seeds at the end of the season. I imagine with the fruit tree idea also the main work would be seasonal, like the grafting stage, and the time planting and tending to the seeds. Also not forgetting that the speedy method I proposed above for fruit trees requires a lot more work than slower methods - I was deliberately giving a speedy method to address that concern. But sure, what I’m talking about does require work!

Though it’s all scalable also. I could have done a lot less work with tomatoes had I chosen lower diversity, for example. And continuing, I can chose to work with as much or as little as I like, with the seeds I’ve created up to now at my disposal for whenever I want to germinate them. I’ve already filed some away and am focusing right now on a subset of them.

But yeah I guess it takes an unusual kind of person to consider the kind of thing I’m suggesting.

Anyway thanks for the encouragement. I started this discussion really just out of curiosity and to learn more about the possibilities of fruit trees. I have no intention of making a fruit tree landrace right now. But for sure I keep an open mind about doing it in future. That’s dependent on a few external conditions. Or this discussion might even inspire some others to try!

Perhaps a good way of bringing this topic forward would be:
Which species could suit an interspecial fruit tree landrace project?

I’m thinking Prunus salicina could be a good candidate.
What other diploid species would be suitable to cross with it?
And what other interspecial combinations would you folk recommend?


Yeah I must admit I don’t get it? But hey it’s me so no worries. I just don’t see the point of it? All tomatoes grow here in the summer. Some better than others but most of the hybrids grow extremely well already. And the variety is amazing. For example I’m growing two hybrid and one heirloom cherry tomato. Each have produced around 300 fruits. The heirloom is about my 8th time I planted it. Using seed from the previous year. I decided to let it go as I prefer the hybrids although I do grow a few heirlooms I like of regular tomatoes. I like the black tomatoes and only grow them. I also grow fig and pomegranates here in Michigan but have to protect the plants. I get around 500 figs a year. Only one fig will actually grow in ground here and produce before winter. But it’s not very good. I only keep it around for ornamental reasons.
I plan to breed new pluots as the flavor range is amazing. One can get some wild crazy looking fruit too.
I also work with brambles and have five cultivars I developed so far. I met most of my goals so not sure what I will do next? Mother Nature helped me out and a volunteer purple raspberry came up in my yard. Which is a cross between reds and blacks. In this case it was a primocane fruiting black. So I have the first primocane fruiting purple raspberry. Discovered in 2022.
Unlike other purples this one has excellent flavor. Tastes a lot like a boysenberry, just outstanding!
I’m going to try and keep it going. I’m cloning two plants via tip rooting now to have a couple backups. I also developed a yellow that is very productive. Most are not. I’m trying to have various colors and I’m working on developing an orange colored Rubus idaeus. So far I have only developed a pink but I have faith! One was developed in Great Briton but not offered here else I would just buy one.
I mostly grow California developed pluots nectarines and peaches as any prunus species grows just fine here. I like what Zaiger has done so use his cultivars in my breeding. I do avoid those susceptible to bacterial spot as it’s not an issue in California so they never bred it out. Still only a few are so no big deal as we know which ones are susceptible. One could always spray if they really wanted to grow it.
Making a landrace sounds difficult. And I’m not sure when I would call it done?
With tomatoes between the amazing tasting heirlooms and the disease resistant hybrids I would not want to compete in that market. I was a member of tomatoville and although I didn’t participate, but enjoyed reading about the dwarf tomato project from the start to the finish. An amazing project that developed over 90 new stabilized varieties. Members grew out the seeds all over the world. They found it took about 8 generations to stabilize a variety. A few seed companies are now selling them.


The point with tomatoes for me is about making it easier in this environment, in which many people struggle with tomatoes outdoors. The main issue is late blight, and only some specialist hybrids are free from that problem though as late blight adapts, their resistance will likely be overcome. Those hybrids also keep costing money each year, you can’t use them in the traditional way of saving seeds, so that’s a big negative. Plus they don’t taste good. But the seed thing is twofold - the cost is real for many people here so that is important. Sustainability is important too, not to have to depend on seed companies, and we have already seen that situation’s weakness with a dramatic pause in the availability of seeds, and I expect that to happen more drastically in future. But above all, saving seeds because that’s the essential to give people the ability to adapt the populations to their land and growing habits.

Almost all of the interspecies tomatoes I grew this year tasted better than all of the fully domestic tomatoes I grew, which included some very delicious varieties including some heirlooms and some new crosses that I was using for their traits regarding climate adaptation, size, growth habit etc. And the very best tasting were all interspecies crosses.

So, this is about disease resistance (not just late blight); cold tolerance; flavour; ease of management (less de-suckering and less supporting); earliness of production in this specific climate; resistance to pests; ability to bear the new hot dry spells without being watered; and the ability to keep adapting to changing climate, pest, and disease pressures as time goes on. There are a lot of very valuable traits in wild tomato species. And a whole world of new flavours in their crosses.


Thanks for the follow up.

Yeah, planting out a bunch of lemon seeds and hoping for hardiness to miraculously appear, in this case at least, is the wrong approach. Lemons have basically no genetic potential for cold hardiness, no matter how your arrange those genes around, a few hours of more then a few degrees below zero is going to kill the tree. To say nothing of the fact that most lemon seeds will be clones of their mother plant.

There are some cold hardy “lemons” that aren’t true lemons, either rangpur (mid twenties I think), Meyer lemon, rough lemon (both low twenties), or ichang lemon (teens, though this one has no citron genetics so it really isn’t a lemon even in the loosest sense) and just enjoy that lemon. Planting out those seeds en mass still would be pretty pointless, since they’d mostly be clones, and citrus take several years to develope much cold hardiness.

If someone were really committed to breeding cold hardy lemons, and had a big greenhouse, they might could get some citrons and cross them with an appropriate cold hardy citrus. You’d need something that gave mostly zygotic seed, was quite juicy, and sour enough that the progeny would be sour enough to be good lemons. Smooth Flat Seville is hardy to the upper part of the upper teens and zygotic. But that’s not likely to get you much more hardiness than a rough lemon, even on the F2 generation. Might could try crossing with a poncirus hybrid, preferably one of the zygotic ones like US-1279, SuperSour 1 and 3, Bishop citanderin, or one of the Conestogas if any of them are zygotic. But going that route, while you’d have cold hardiness and zygotic seed, the juice content would probably be mediocre, and the poncirus yuck might come through. So maybe a second round of crosses, with the best of citron x Smooth Flat Seville and the best of citron x poncirus hybrid, at which point you’d have something ½ citron, about ¼-⅛ mandarin, about ⅛ pomelo, and about ⅛ poncirus (and a gigantic greenhouse). Given enough money, time, and luck, a true or nearly true lemon hardy to the low twenties or maybe even into the teens sounds possible. My gut feeling is it’d take about three decades, though ideally you’d tack on another five to ten years to properly trial the final generation.

I really like this.

Making really wide crosses like this, or like in my citrus example, is what gets me excited. While more challenging and failure prone, the potential reward seems higher as a general rule. And tomatoes, with the huge variety of wild relatives, fast generational turnover, and ease of working with, together with the fact that so many people grow them and that there are a lot of disease and other problems that need overcoming, are really a good candidate for this kind of breeding work. Heck, even the landrace idea makes sense here, since tomatoes have to be reseeding every year.