Boards > Plant Breeding Projects

Promiscuously Pollinating Tomatoes

(1/3) > >>


I am working on a project to convert the tomatoes in my garden from a population that is primarily self-pollinating, to a population that is primarily cross-pollinating. That means that the flowers need to be more attractive to pollinators. In my garden that is probably bees. It seems to me that if two heirloom tomato varieties are crossed to make hybrid tomatoes, that the hybrid offspring (averaged over many crossing combinations) produce about 50% more fruit than the most productive parent. I think that explains why hybrid tomatoes are overwhelmingly preferred over heirlooms by people to whom productivity is very important. More natural cross pollination would mean that I am growing more naturally occurring hybrids and could thus expect higher productivity.

I live in a climate that is near to the ecological limits of tomatoes... That means that about 90% of the varieties I trial fail to produce ripe fruit in the available growing season. If I grow heirlooms, or commercial hybrids, I am growing varieties that were selected for different conditions in far away places with different soils, and sunlight, and weather, and bugs, and microbes, and farmer's habits. Heirlooms have been inbred for as many as 50 or 60 generations. They tend to do poorly in my garden. I attribute their poor performance to a combination of inbreeding depression, and not being localized to my growing conditions. Commercial hybrids are expensive because I would have to purchase the seed every year. I could make my own hybrids, using varieties as parents that are known to thrive on my farm. That'd require meticulous care to make sure I maintain the parent lines, and that I make the crosses correctly, and keep the paperwork straight. Not really my way of doing things. I prefer as much as possible to allow the natural systems around me to take care of cross pollination.

A few years ago, I was doing a trail for frost/cold tolerance in tomatoes... I noticed two varieties in the trail that had bumblebees on them pretty much any time I walked past the plants. The bees visited other tomato plants on scouting expeditions, but didn't hang around for long periods of time collecting pollen.  Those two plants also happened to be the most productive plants in my garden that growing season. (One of them has become my main early season production tomato.)  The plants were highly attractive to bumblebees because they drop clouds of pollen when shaken.

That got me to thinking about how glorious it would be if I could return my tomatoes to closer to their ancestral state of being mandatory out-crossers. That would allow lots of new genetic combinations to be generated without requiring manual pollination. And with the genetic roulette wheel spinning faster, I can expect local adaptation to occur much quicker. So I started paying attention to the flowers on my tomatoes to see if I was currently growing anything that might be more susceptible to cross pollination.

The tomato that drops clouds of pollen has flowers that look like this. I call this an industrialized type flower. It's all closed up to prevent cross pollination. Makes the purists happy. At least it drops clouds of pollen, so it's not as industrialized as it could be.

Industrialized flower (complete with pike phalanx)

The nearest thing to an heirloom tomato for my valley was developed for the Campbell's soup company about 45 years ago. It grows great here, although technically, it was developed for the next valley to the west which is at lower elevation and thus has a longer growing season. So it requires just a bit longer growing season than my garden provides. But look at those wide open flowers! Pollen can get all over that stigma. The anther cone is open. Pollen can easily get in and out of the flower.

Open flower.

So I made a manual cross pollination between these two varieties...
I am hoping to find offspring among the grandchildren that combine the trait of open flower structure with the trait of dropping clouds of pollen and being highly attractive to bumblebees (or other pollinators). I grew the F1 hybrid in the basement overwinter, and currently have second generation plants growing in the greenhouse. They are ready to be planted into the field soon. It would also be nice to combine the super early productivity of one of the parents with the larger fruit size of the other... Here's what they looked like a few minutes ago.

F2 Hybrid Tomatoes: DX52-12 X Ot'Jagodka

I also found other varieties with loose or open flower structures and tried to make crosses. Here's what one of the flowers from one of those varieties looked like. I mostly flubbed the cross pollinations. (I should have put pollen from the long season varieties onto a short season tomato. I did the cross the opposite direction, so most of the fruits didn't mature before getting frozen.) But I was able to grow a few of these during the winter and have some F2 plants currently growing. One of the plants where the cross succeeded was to Hillbilly which is an orange/red tomato. I'm excited about that because I prefer the taste of orange or yellow tomatoes. Eventually I may convert all of my tomatoes to orange.

Beelover's Tomato: Croatian Brandywine

So this year, I'm expecting to be looking at lots of tomato flowers to find flowers that are wide open to pollination, and that the bees adore. I'm not fixated on using bumblebees, I'd select for flowers that are attractive to any species of pollinator. Bumblebees were the pioneer species that got this project started, so they are likely to play a big role in selection.

Longer term, I am  looking for genetics that can be incorporated from wild tomatoes to require crossing by using self-incompatibility genes.

Eventually, I want to get the promiscuous pollination trait sufficiently established so that I can send the tomatoes into areas for trial that are really bothered by dampness diseases such as late blight... Instead of throwing a few hundreds of heirlooms and commercial hybrids at the problem, we can throw tens of thousands of plants at the problem which are able to easily swap pollen with each other and self-select for the perfect combination of genes to handle changes in the environment and in the microbiome. If I find a particular species of hole nesting bee that really takes to the tomatoes, then it might be possible to share the bee at the same time as I share the tomatoes.

Further Reading:
Landrace Gardening: Promiscuously Pollinated Tomatoes

Original Post 2013-12-31:

A few years ago I was conducting a cold/frost tolerant tomato trial. The primary purpose of the trial didn't have anything to do with cross-pollination or bees. That arose by accident. It seemed like every time I walked past the tomato patch there were a few bumblebees in it. So I started paying more attention. The bumblebees visited many tomato plants, but they were what I have come to call fly-bys. They'd stop at a flower for less than a second, and then be zipping away. When they got to two particular plants, they would stop, and spend up to ten seconds per flower, then fly to the next flower on the plant and spend another ten seconds, etc...

The two cultivars that were highly attractive to the bumblebees were also the two most productive tomatoes in my garden last year. I attribute that partially to the heavy presence of bumblebees. The flowers got pollinated better and thus set heavier loads of fruit that filled out better. One variety started flowering weeks before these plants, but most of the flowers dropped off without forming fruits. I didn't ever see a bumblebee on that plant. I'm wondering if the flowers simply didn't get pollinated?

The two varieties were both Russian varieties 'Jagodka', and 'Nevskiy Red'.

I primarily grow landrace crops. Landraces are genetically diverse crops which have been selected by natural and human selection to thrive for a particular farmer in a particular location. I would like to convert my tomatoes into a promiscuously pollinating landrace population so that I can more easily do mass-selection plant breeding. Because bumblebees were the only pollinator that I noticed that was interested in working the tomato flowers this summer, it seems like I should cater to their needs.

Feedback I received online  recommended the following varieties.

Attractive to Bumblebees:

* Hillbilly
* Virginia Sweets
* Blondkopfchen
* Tommy Toe
* Stupice
* Croatian Brandywine
Produced lots of pollen:

* Indian Stripe
* Danko
* Black Early
* Sungold
* PI 120256 (Not highly attractive to bumblebees in my garden in 2013. Exposed Stigma.)
I decided that I will work on a new project for the next few years: creating a population of tomatoes that is promiscuously cross pollinating. With that goal in mind I started looking at flowers more carefully. I used a 20X magnifying glass to examine the flowers. Yes, I am nerdy enough to carry a magnifying glass with me just about everywhere I go. Here are traits that I noticed that I think might contribute to a successful project.

Simple Flowers
The first thing that was obvious even without magnification is that some tomato flowers have huge numbers of petals. And those petals can be arranged in a manner that prevents bees from getting to the pollen. Here's what I mean by that:

Open Anther Cone
Writers generally talk about an 'anther cone' in relation to tomatoes because the anthers are fused together making a cone around the style. Around 10% of the tomato cultivars that I examined had anthers that were not (fully) fused together. They provided a more open architecture to the flower. It seems like it would be easier for foreign pollen to reach the stigma. Here's what that looked like:

Loose Anther Cone
There were differences regarding how tightly the anther cone wrapped around the style. In some cases the diameter of the opening in the anther cone was much wider than the style. In other cases the anther cone was so tight that it looked completely blocked off. It seems like a wide opening would allow more pollen to fall out of the flower to feed bees, and be more likely to be cross pollinated. I didn't document this trait as carefully as I should have. Here's what that looked like:

Extended Stigma
The trait that everyone talks about in relation to tomato flowers and cross pollination. In some cases the stigma was entirely inside the anther cone, in other cases the two were about of equal length, and in some cases the stigma protruded slightly from the anther cone. Cultivars with an exposed stigma were about 15% of the patch. It seems that a more exposed stigma would be more susceptible to cross pollination (If pollinators were attracted to the flower for other reasons). Here's what that looked like:

Free Flowing Pollen
Two varieties had pollen that poured out of the flower when vibrated. These were the two that were highly attractive to bumblebees. These represented 5% of the cultivars in the patch. They were my most productive plants. On a different variety that I dissected the pollen looked like jelly. I noticed this free flowing pollen trait in my new potato varieties when I switched from growing sterile plants to growing abundantly fruiting plants. (Bumblebees love my current potatoes.) I'm wondering if a similar trait is at work in tomatoes. Here's what that looked like:

Larger Flowers
Perhaps larger petals, bigger flowers, or stronger petioles would make flowers more attractive to pollinators: Especially to some of the larger species of bumblebees.

Self Incompatibility
Some species of wild tomatoes are self-incompatible. Long term, I would like to incorporate that trait into my tomatoes. Tomatillos are an example of a closely related self-incompatible species.

UV Markers
Some flowers may have markings on them that are visible only in the UV range that make them more attractive to pollinators.

When I originally did the evaluation of the tomato flowers, I did not know what to look for, and my vocabulary was a bit loose, so my records are not as easy to understand as I would like. Before I do another evaluation I will create a template, so I can score each of the traits mentioned here. I think that it would be really useful to me to attempt to combine some of these traits into the same plants. An exposed stigma, a loose or open anther cone, and a cloud of pollen in the same flower sounds really exciting. That kind of challenge is what keeps me going as a grower and plant breeder.

I think that I currently have enough germplasm to do this project... More is always better. The two varieties that were highly attractive to bumblebees were planted about 6 feet apart this year. I'm intending to plant large numbers of plants and screen them for any natural hybrids that might show up. One of the varieties that I have grown for years has an exerted stigma. A cross between these could be interesting. I'm thinking that highly attractive to bumblebees or other pollinators (probably implies a loose or open anther cone and free-flowing pollen) and extended stigmas would be a great combination.

Some years ago I saved the seeds from Sungold, and have been replanting them ever since. I call them "orange cherry tomatoes". A tomato grower that I highly respect, Darrel Jones, speculates that one of the ancestors of Sungold is Solanum habrochaites which is a wild tomato with large flowers and exerted stigmas.

Yesterday when I was examining the tomato flowers I noticed that one of the  plants among the orange cherry tomatoes had huge flower petals. Others had exerted stigmas. These are exactly the traits that I am looking for in my promiscuous pollination project!!!

And I like the flavor of orange tomatoes better than red tomatoes. So it looks like I am well on my way to achieving my goals of promiscuous pollination and orange tomatoes.

Some populations of S. habrochaites are self-incompatible, meaning that they require another plant for pollination. I would love to incorporate that trait into my tomatoes. I'm intending to screen the orange cherry tomatoes for self incompatibility. I only grew 6 plants. If I had known that this variety had so many of the traits I am looking for, I would have planted dozens.

Yesterday I planted seeds from LA1777 (Solanum habrochaites) and from a cross between LA1777 and Columbianium which is a wild-type cherry tomato. It's late in my growing season, so I expect to grow them in pots so that I can bring them indoors when the frosty weather starts.

My hopes for this cross are to bring the self-incompatibility trait from LA1777 into domestic tomatoes. There are conflicting reports about whether LA1777 is self-incompatible, but I can screen it and make my own determination. In any case, the large flowers and exerted stigmas are a good first step.

I found these flowers yesterday. They are descended from Sungold. We speculate that Sungold has a wild ancestor (Solanum habrochaites) with large flowers and extra long styles. This plant called attention to itself by having extra large flower petals. When I looked closer I was thrilled that the stigma is outside the anther cone which aught to really increase the rate of cross pollination. I grew out about a hundred of these this spring, but only kept 6 plants for myself. Some of them have more industrialized-type flowers. Boo Hoo. If I had known, I wouldn't have been sharing....

Tomato flower with exerted stigma and extra large petals.

Today I installed a bee nesting log into the tomato patch.

I put many different sized holes:

In the barn I installed a more formal block with like sized holes grouped together. Perhaps making it easier to match up species visiting the tomatoes with what's using the nesting block. I read that more bees are ground nesting than wood nesting, but gotta start somewhere.

Micro-bees were visiting the tomatillos today, but I wasn't able to capture a good image.


[0] Message Index

[#] Next page

Go to full version