Physiological and Morphological Changes Over the Past 50 Years in Yield Components in Tomato
Tadahisa Higashide and Ep Heuvelink
Horticultural Supply Chains Group, Wageningen University, Marijkeweg 22, 6709 PG Wageningen, The Netherlands
Citation
Higashide, T., & Heuvelink, E. (2009).Physiological and Morphological Changes Over the Past 50 Years in Yield Components in Tomato . Journal of the American Society for Horticultural Science, 134(4), 460-465.
Background
Greenhouse tomato yield in The Netherlands has more than doubled since the 1980s. This increase is caused by environmental effects such as greenhouse and controlled environment production practices and improved cultivation techniques. In addition to production environment improvements there are genetic effects that positively influence performance that are attributable to breeding efforts.
The aim of this research was to investigate whether tomato cultivars that were released between 1950 and 2000 show an increasing trend in the trait “yield” which is an aggregate of many traits that are influenced by plant morphology and plant physiology.
Experimental Design
Eight Dutch tomato cultivars [Moneymaker (release in 1950), Premier (1960), Extase (1960), Sonatine (1975), Calypso (1982), Liberto (1988), Gourmet (1991), and Encore (2002)] and one Japanese cultivar [Momotaro Fight (2001)] were tested in a randomized complete block design with each cultivar (genetic treatment) occuring in each block randomly. Two blocks were tested under the same environmental conditions to account for spatial variation. All cultivars were indeterminate type and had medium–large round fruit. Plants were measured destructively and non-destructively for various morphological and physiological traits that are considered to be components of yield.
Results
An increase in tomato yield because of breeding efforts was not caused by an improvement in resource partitioning to the fruit but by an improvement in resource partitioning to vegetative characteristics that resulted in higher dry matter. This increase in vegetative dry matter production was caused by higher light use efficiency and is influenced by tomato morphology and architecture. This result is consistent with previous studies in maize (Hay, 1995).
The leaf photosynthetic rate of the modern cultivars increased proportionally to light use efficiency indicating that there is a positive relationship between light use efficiency and leaf photosynthetic rate. Light use efficiency and leaf photosynthetic physiological traits were indirectly selected for over the course of 50 years as a product of selecting and releasing cultivars that had the highest yield. Yield an aggregate trait of many physiological and morphological characteristics. A more detailed study of leaf photosynthetic rate would need to be done to clarify the cause of its increase over the 50 years of variety release.
Yield of the Japanese cultivar was significantly lower than the other Dutch cultivars. This is likely because breeding objectives of Japanese cultivars are geared more toward quality instead of yield. Soluble solids in the Japanese cultivar was significantly higher than the dutch cultivars.
We can use this knowledge to inform future breeding efforts
Yield as we measure it is an aggregate of many components (Figure 2). From the results of this study we know that all components of yield do not contribute to yield improvement equally. We can optimize future breeding efforts to improve morphological and physiological characteristics that directly improve performance of a variety measured as yield. Focusing on yield components that are directly proportional to variety performance measured as yield could improve our ability to map regions of the genome that are associated with specific morphological and physiological traits to determine the genetic basis of yield. To better implement morphological and physiological traits resulting in measurable yield increase into a breeding program seed companies would need to develop high throughput phenotyping techniques as many of these measurements are not cost effective in larger population sizes.
Literature cited
Hay, R.K.M. 1995. Harvest index: A review of its use in plant breeding and crop physiology. Ann. Appl. Biol. 126:197–216.
Can you think of a cost effective way to measure morphological and physiological traits in large population sizes. Phenotyping is a bottleneck in the breeding process. I think that plant architecture could be measured and analyzed efficiently using imaging technology but what about the physiological traits? The LI-COR is too slow for large population sizes, is there something better out there?.
General levels of growth could be imaged well using drones or other vantage points in the field. Physiological measurements could be taken by hand, but that takes time and limits the amount of information that can be gathered in a given period of time. Using a LI-COR or similar device takes more time than might be acceptable.
As it stands now, there is no method out there that can examine the complex physiology of a stand of plants without losing specificity of certain plants. The best methods currently use random sampling to determine the overall effects, but you do have to account for error in those models.
Currently, controlled environments are used in order to measure physiological performance under known conditions. In this way, a stand of plants (or individual plants) can have gas exchange measured and give a wealth of information that can estimate the plant performance. There are limitations to this, however, as you do not replicate true field conditions. But, it does allow for a higher resolution when examining plant physiology and performance. This is an approach that is gaining more interest in recent years, and likely will gain more attention due to its potential.
High throughput imaging could be very effective, especially developing analysis techniques in open source software such as ImageJ. I think it would be especially useful if there was there was an imaging technique that was highly correlated with physiological measurements.
I also agree that high-throughput measurement techniques using mounted sensors/imagery devices for physiological traits like photosynthetic capacity, transpiration rates, etc are already already available (and at reasonable prices for GH-based systems), but more research is needed to develop the analysis platforms to handle large amounts of data. Cost-effectiveness may come later after the physiological traits are linked to plant performance traits or genes which allow for indirect selection or simulation studies.
If the size of each population is large enough, there is a method to measure photosynthesis and transpiration using aerodynamic approach. This includes measurements of vertical profile of CO2 and water vapor concentrations within the boundary layer. Since phenotyping does not have such room to create a large canopy, this approach has limitations in applying in such mixed canopies.
Do you think that the method that Körner et al. (2009) used in “Maximizing crop photosynthesis across the entire canopy requires the optimization of many environmental factors” would be an effective and quick method to estimate the photosysthetic rate of large populations?
Yes. But, it might not be applicable in a field scenario. The model that Körner et al. (2009) used worked so well because it was in a controlled environment. But, more factors come into play in the field (mostly environmental).
I think that modelling field crops would be immensely more challenging and the model may not even reflect the true nature of the crop. You cannot gather gas exchange data reliably on a large scale in the field, which would be one of the best measures of physiological performance and behavior. Therefore, I am not sure that model would even be applicable without considerable concessions and proper experimental design that can account for the majority of factors without confounding them.
I think you could use a controlled environment to acquire a relative model for field production. As long as light intensity, temperature, and CO2 concentrations remain in a field average range, the plant response should be comparable. Soil conditions can be mimicked or the structure could be over bare soil. It wouldn’t be easy, as mentioned in the previous comment, and several assumptions would need to be made. However, for a relative population response, I think the method used in Korner et al. (2009) could be promising.
Controlled environment to replicate field conditions for future trait introgression would be a very cost effective method of screening lines, as long as the trait was field stable. The trouble is when the trait is highly influenced by environment.
If you are screening and selecting individuals for breeding purposes you would need to measure each individual.
It would be interesting to do a mirrored study with all Japanese cultivars compared to one Dutch cultivar. How do you think the results would be different? Do you think this would shed some light on other components to select on instead of simply soluble solids when breeding for quality?
Sumida et al. (2008) reported that breeding objectives of Japanese cultivars Momotaro and its sister cultivars such as Momotaro Fight were sweetness, hardness and succulence of fruit, and short internodes of plants.
Sumida, A., Kaya, T., & Hatanaka, M. (2008). Breeding and promotion of tomato [Lycopersicon esculentum] cultivar’Momotaro’resulted from innovation of its shipping and taste. Horticultural Research (Japan).
Interesting. So based on the original question, it would be interesting to do a similar type of analyses with focus on factors affecting taste and quality attributes of tomato. For brix, we know that water content is a significant factor and we know that ratio between transpiration and photosynthesis correlates well with brix.
Yield is one of the most significant traits in almost all production systems and tomato is not the exception. In class you discuss a shift in the market for having better nutritional qualities or more sugars (Brix) depending on the ending product. From your own experience, is breeding for high sugars or lycopene is something the industry is willing to pay for? Is there room in the current market for focusing in more than yield and getting a revenue for it?
In CEA industry, they are more focusing on taste and flavor now, compared with 20 years ago (for example). The huge competition with Mexican tomatoes made the greenhouse industry look for something outstanding in the market (value) rather than price.