How the distinction between mature dead cells and aberrant cells can help your breeding program and pollen supply chain optimization

Introduction – About the Author, Silvan Kaufmann

Hi, my name is Silvan. I am an application scientist at Amphasys. In the last couple of years I have visited numerous customers working in the seed industry around the world, and I would like to dedicate this blog post to a topic that I find worth spreading. It’s about tomato pollen. To be more precise: Dead tomato pollen. 

What you will learn from this article

You will learn that dead is not just dead. There are dead cells that have matured before dying, and cells that were aborted in their development. The relative abundance of each subtype on its own reveals important information that can even lead to a refinement of the term pollen viability

Here is why you should continue the read: 

  • For tomato breeders I will present a potential plant stress gauge
  • For the tomato seed producers or a production researchers I will show how a few pollen quality measurements already indicate whether you have a pollen supply chain problem or not
  • And for everyone else, I take you on a hot journey with a mobile laboratory in my backpack. And who knows, maybe the findings are relevant to you too. Many ideas discussed here do not only apply for tomato

Interested? Then grab a coffee and prepare yourself for a 25 minute read. If you don’t drink coffee, go for a chai tea, because our journey begins in India.

First Impressions From a Field Trip to India


A few years back, I was on this exciting trip to South India to collaborate on a Technology Implementation Project for tomato hybrid seed production. In this project we combined Amphasys’ expertise on pollen quality assurance with our customers’ vast experience in seed production, aiming at implementing standardized pollen quality monitoring along the entire supply chain. 

I was sweating that day during the tomato field inspection. It was an open field production, meaning that you are directly exposed to the sun. Early afternoon, the hottest time of the day – and I was just about to start digesting that unbelievably tasty and also incredibly spicy meal we were offered by out hospitable clients. It was that mix between radiation and capsaicin that kept my body gently simmering there, between the bright yellow tomato flowers in this ocean of deep green leaves. 

Sunbathing Tomato Plants

sunbathing tomatoes

I was impressed by all these field workers who were manually pollinating the flowers in the fields. Flower by flower, plant by plant, row by row. But I was also impressed by these plants, being exposed to that intense radiation and high temperature, and still producing green leaves, yellow flowers and red fruits. They must enjoy this daily sunbathing I thought for myself. 

That was not the only occasion where I visited tomato seed production fields. In the following years, I was standing in many greenhouses, net houses and open fields, and they all had one thing in common. No matter whether they were located in Asia, South America or Europe. It was hot. I experienced that tomato is often grown in regions in which daily maximum temperatures easily exceed 30ºC. I am sure that my gametes survived all those field visits without permanent damage, but I am not so sure about the gametophytes of the plants – the pollen. And that’s what this blog post is all about. But let’s go step by step… 

Pollen Quality Matters

In Technology Implementation Projects we have a closer look at the pollen supply chain of customers. The pollen supply chain is a key process in breeding and seed production, and usually consists of various steps of pollen management, such as production, collection, preservation and pollination. 

At the beginning of the projects we usually measure the pollen quality along the whole process, leading to the identification of the bottlenecks and improvement of critical steps. Then we define quality checkpoints. These are specific points along the value chain in which standardized quality controls are implemented. 

Below you see a simplified scheme of a possible tomato pollen supply chain. 

simplified scheme

The Tomato Pollen Supply Chain

The supply chain of tomato pollen is comparably diverse and extensive, and thus usually has a great potential for optimization. The fact that pollen is collected, preserved and transported already opens up a broad field of applications for quality testing, because, you know, when things can go wrong, they will go wrong at some point. And another thing to bear in mind: The pollen quality can only decrease along the supply chain. It never increases. You have to start your process with the best material, otherwise the final quality may be insufficient to obtain a full seed set [1]. 

Considerable Variation in Tomato Pollen Viability

The extent of the tomato pollen supply chain is one aspect making it attractive for optimization. The other aspect is that there is a lot of variation in tomato pollen quality. In the last couple of years, I have measured tomato pollen from flower buds, open flowers, dried stamens, extracted pure pollen, stored pollen, pollen mixed with lycopodium spores or corn starch, pollen collected from a freshly pollinated stigma and many more… and I have encountered everything.

 Pollen qualities ranging from a little over 0% viable cells to almost 100% viable cells are a reality in this business. 

Let me quickly illustrate this for you. 

In the bar plot below I plotted the pollen viabilities of 1,000 tomato pollen samples, collected among all steps of the pollen supply process in many countries. Green bars represent the viable fraction (viability), and red bars the dead fraction, both adding up to 100% for each of the 1,000 measurements. 


On the very left side we have pollen quality of almost 100%, and far to the right the samples are mostly dead. The circles I have added contain typical scatter plots of pollen measurements with the Ampha Z32 Pollen Analyzer. You can identify different populations of points, corresponding to clouds of viable and dead cells. Don’t worry if you are not used to the interpretation of flow cytometry scatter plots, you can find the basics of the technology here

The samples number 900 to 1000 (on the right side of the plot) make 10% of all the measurements, and they all contain less than 10% viable pollen. I guess you can imagine what happens if you use such samples for pollination. The quality checkpoints in a well-organized supply chain monitoring program would not permit further processing or storing of such samples, unless they are used for dilutions. 

The associates risk of loss of seed yield is just too high [1]. 

In-vitro Germination vs. Impedance Flow Cytometry – Comparing Apples and Oranges

The next important insight came from a customer experiment in which he correlated tomato pollen viability with in-vitro germination. 

The lab technician just completed counting germinating and nongerminating cells of the very last sample of the day and came up with an impressive 92%. 


Surprised I was staring at the viability measurement results on my laptop…

viability 65

I had the viability results of the same sample in from of me: 65%. 

92% and 65%. That’s quite a discrepancy. And even more puzzling: Germination higher than viability? Something is rotten…

Together we ran the measurement again. Same result. Then we had another look through the microscope. Indeed, there were lots of long pollen tubes all across the field of view. But also lots of small amorphous particles. According to the germination protocol of our partner such small particles are cells that should be ignored. This immediately caught my attention, as the pollen analyzer can measure virtually any kind of particle above a certain size. I was curious whether those were detected in the previous measurement. 


scatter plot

The scatter plot revealed two distinct clouds of dead cells. A small diffuse cloud of mature dead cells was positioned a bit higher than another population of small dead cells. That’s exactly what we observed with microscopy. Just the way we interpreted the data was different. We were comparing apples with oranges.

When we want to compare the methods, we obviously need to classify those small deformed cells similarly. Let’s call them “aberrant” cells from now on. And let’s make a summary of the different cell populations we typically detect in a measurement:

mature vs aberrant

Quantifying Aberrant Tomato Pollen

 Now the key question is: How do we treat those aberrant cells? The AmphaSoft software (software that comes with the pollen analyzer) includes a useful feature to mark and exclude unwanted populations from further analysis. This exclusion would have led to a perfect match with the results of the germination tests. 

However, in my opinion, ignoring those cells is completely wrong. Here is why. 

First, let’s see how frequently they occur. I plotted again the viability of the 1,000 measurements. Let’s focus now solely on the dead fraction in red. 

dead fraction

Then I divided the dataset of dead cells into mature dead cells (red) and aberrant cells (blue) and sorted based on the aberrant cell fraction. 

mature vs aberrant graph

Now it’s getting interesting. Most samples actually contain some of these aberrant cells. For most it is not a lot, but for 10% of the samples aberrant cells account for more than 40% of the dead cells. And some extemes even consist mostly of aberrant cells. 

Now let’s see the fraction of aberrant cells in the whole sample, not just the dead cells. 

qualityAmong those 1,000 measurements the median aberrant cell fraction was 5.2%. Well, that doesn’t sound like a lot. It is actually not a lot. But there is a significant fraction of samples that are very far from the median. Let’s have a look at the upper end of the spectrum. Here we have samples with an aberrant cell fraction well above 50%. The highest had 87.3%! 

And here we need to be careful. 

The effects of low pollen viabilities can be considerable. This was also one of the main findings of the collaborative research project done with Rijk Zwaan. In a large-scale research project, we found a considerable reduction in seed yield for low pollen qualities. And that was confirmed with all 12 crosses under investigation [1]. 

If aberrant cells are important factors leading to a reduced pollen viability, and a reduced pollen viability is associated with a lower seed yield, then those cells should receive the necessary attention also from an economic point of view. 


The Origin of Aberrant Cells

One of our main goals at Amphasys is to provide a tool that can be used to pinpoint problems in the pollen supply chain. But finding the problem is not the end of story. We also want to provide solutions. 

So the urgent question is: How can we prevent the occurrence of such aberrant pollen? 

When we first collaborated with a customer on this topic, our proven recipes for pollen supply chain optimization did not apply anymore, and we had to have a closer look to the very beginning of the chain. To the breeding and growing part. 


Aberrant cells originate in the Grow part of the supply chain. The reason is linked to the beginning of this blog post, when I was standing in the middle of that tomato field in India. 

Tomato plants like warm temperatures. They can even deal with hot temperatures. Luckily because otherwise that delicious tomato sauce from Southern Italy would not exist. But even if the plant leaves do not mind the sunbathing, the developing pollen does. It absolutely does. In fact, many researchers came to the conclusion that the early pollen development is a highly temperature sensitive process. It’s even considered one of the most susceptible processes in the entire plant sexual reproduction cycle [2,3,4]. And with the global warming predicted for our future it is not getting better. 

 tomatoThat’s why research on heat – and drought – resistant plant lines is so crucial. And I think it goes beyond saying that big attention should be paid to pollen development and pollen quality in all these global efforts to create new robust lines. 

Now, what does that have to do with the aberrant tomato cells? 

There are numerous hints in literature that associate empty, sterile, degenerated, injured, or abnormal pollen grains with plant stress, mostly heat [4,5]. The current hypothesis is that under stress conditions the microspore development can be arrested and the microspore is eventually aborted. As microspores grow in size during maturation, the aborted cell is nonviable and small in size. 

Isn’t that exactly what we observed in our plots? Remember? 

stress scatter plot

I strongly believe that aberrant cells are linked to arrested or aborted pollen development. The abortion can be caused by abiotic factors such as heat or drought. We also made the observation that the fraction of aberrant cells shows some variation over time, but the overall abundance is line-specific. In addition, we saw large amounts of aberrant cells in pollen samples from buds, opening flowers and fully open flowers, indicating that they are not artefacts from pollen processing and storage. 




Preventing Aberrant Cell Formation

After introducing stress as a putative inducer of aberrant cell formation, we can now come back to the question on how we can prevent the occurrence of those cells. 

thermometerWell, basically you need to remove the stressor. In case of heat stress in high-end greenhouses, cooling systems or ventilation could remove excess heat. In tomato hybrid seed production I have seen many simpler net house systems though. Here temperature regulation is usually not feasible. Under such circumstances the production window for a heat-sensitvie line could be narrowed to cooler months of the year. During these months male plants could be grown on a larger surface for this approach, because pollen can be stored easily for months. There are even reports about storage over years. For such long-term storage make sure you optimize pollen dehydration to the right moisture content and choose a suitable storage temperature. 

Still there, dear reader? 

Let me summarize the different results we have come across today, and how they can help you with the optimization of your supply chain or breeding program. 

viability levels

Looking at this, I would even go a step further and recommend reviewing your pollen testing standard operating procedure. Why not adding the quantity of aberrant pollen to your viability data? 


How to make use of the data? 

If you are working in production or R&D, you have a wonderful tool to characterize and optimize your entire pollen management process. Do it! Take your instrument on a trip. Start with analyzing pollen from fresh flowers, and stop after quantifying the viability and pollen number on a manually pollinated stigma. You’ll learn a lot about your process, identify bottlenecks and eventually increase productivity. Or maybe you did that already and still have the impression that the pollen quality could be better? Then maybe the problem is occurring before you collect the pollen, for example during the microspore development under stress conditions… Look at the scatter plots. They contain useful indications. 

If you are working in breeding, I could imagine that heat – and drought – resistance are traits that are on your radar. Do you really pay enough attention to pollen development, the maybe most sensitive step in the whole plant reproduction cycle? How would a male line with impaired pollen maturation perform in hybrid seed production? 

I personally find this an overly exciting topic worth discussing, and that’s why I dedicated my first blog post to it. If you have questions or would like to share your own experience let us know! We are keen on hearing your view. 

Stay tuned



Rijk Zwan Customer Research Project

2 Zinn KE, Tunc-Ozdemir M, Harper JF, Temperature stress and plant sexual reproduction: Uncovering the weakest links. J Exp Bot. 2010;61(7):1959-68

3 Lohani N, Singh MB, Bhalla PL. High temperature susceptibility of sexual reproduction in crop plants. J Exp Bot. 2020;71(2):555-568

4 Giorno F, Wolters-Arts M, Mariani C, Rieu I. Ensuring Reproduction at High Temperatures: The Heat Stress Response during Anther and Pollen Development. Plants (Basel). 2013;2(3):489-506

5 Iwahori S. High temperature injuries in tomato. Journal of the Japanese Society for Horticultural Science. 1965; 34(1):33-41