Evaluation of Upland Rice Suitability for the Northeast Through Testing SRI Principles and Measuring Yield Across Varieties

Takeaways Across Trials

We were able to effectively implement our experiment as planned despite some setbacks due to weather, disease and general learning curve. A cold, damp spring that kept the soil from drying out meant we were forced to delay constructing the plastic beds and in turn, our planting date--we'd anticipated planting on May 19th and instead planted on May 26th. This slightly shifted the age of the seedlings we planted out, but overall the experiment still demonstrated a diversity of seedling ages at transplant date and their impact on overall yield, seedling and crop vigor, survival rate, and ultimately, final yield. We have selected 22 days as our optimum outplanting seedling age, and are using this outcome to shift our planting date for production rice earlier, from late May or early June to the second week of May. The hope is that these transplants will be able to withstand slightly cooler (but not frost) spring temperatures, and this early outplanting will allow plants to experience more growing degree days as soon as possible after sowing. Ideally, this extension of the growing season would ensure a more robust production window before the first frost of fall.

Not all of our varieties produced substantial amounts of grain, both due to disease incidence and their adaptation to longer growing seasons than we experience in the Northeast. Detailed observations are below in the Results of Variety Trial section.

Results of Variety Trial

Yield

Across both the variety trial and the SRI trial, Loto was the heaviest yielder, even when accounting for winnowed weight (weight after the removal of chaff and unfilled spikelets). The comparison of yield from the "center 5" plants across varieties is shown in Figure 1 in the document attached. We used measurements of yield per plant, per 5 plants sampled, and per lot of seed to calculate potential yield per acre; those results are summarized in Table 3 below. The yields of the next-highest yielding varieties was significantly lower, with Estrella and Delmati generating 5.14 and 3.32 lbs post-harvest respectively. We used the total plot data in our calculation of yield per acre, assuming that each plot was a more accurate estimation of yield at scale compared to the center 5 plants' yield, which were likely protected from wind, bird pressure, etc. The data in Figure 1 are presented using the center 5 plots because it was the most precise data we were able to collect across varieties. 

One of the most significant factors impacting our ability to maximize yield in the variety trial seemed to be the days to maturity per cultivar. Disease pressure aside, several of the cultivars produced many spikelets that did not end up filling out with mature grain because the temperature dropped, day length decreased, and first frost came before the grains were able to fill. Jupiter and IAC 600 demonstrated this most apparently, with Jupiter not producing a usable amount of grain across the trial plots. The loss of significant Sutarsar and Kumamoto crop was a result of partial or entire plant death due to rice blast, but the plants appeared otherwise promising. The days until harvest from transplant date is summarized in Table 4 below.

Table 3. Summary of Yields Across Varieties. Results are given on a per plant, per 5 plants sampled, per plot, and per acre basis. Yield per acre was calculated using the assumption of 39150 plants per acre given the spacing used in our trial.

Table 4. Days to Harvest for varieties included in the variety trial.

Lodging

Lodging incidence was strongly variable across cultivars. Estrella, Kumamoto, and Kyzyl Shala experienced the most lodging, as they were the tallest, most top-heavy varieties at maturity. The timing of the lodging was mostly after the grain had developed, and was in response to a few particularly strong storms that may have weakened the soil structure in combination with wind damage. This did not significantly impact yield at our scale. However, the incidence of lodging, up to 50% in Estrella plots, was significant enough that at scale it would be an important consideration. Many rice combines have a setting that allows farmers to pick up lodged rice so it would not be likely to severely impact yield.

Disease Incidence

The cultivar trial experienced two pest pressures of note: flea beetles and rice blast. The impact of each on the affected varieties is summarized in Table 5 below. The flea beetle damage tended to appear across cultivars as small "peepholes" in emerging leaves. These leaves did tend to senesce more rapidly than unaffected leaves, but it did not appear to have an impact on plant growth or yield.

In comparison, our experience with rice blast was of greater concern. We believe the rice blast may have originated from other members of the Poaceae (such as millet) that had been used as cover crop in the past and had remained in the soil as resistant spores. Blast infection and spread tends to appear concurrently with floral emergence, and water management is a key factor in preventing its transmission between plants: heavy watering consistently from the time of first observed flowering is thought to help manage the spread of blast. We seem to have unfortunately missed a critical window and began our intensive watering protocol a few days late, allowing the blast to spread rapidly from its entry point in a plot of Kyzyl Shala. The lessons we learned and will apply to future trials and production are expanded upon below ("Issues Encountered"). We are preparing a protocol for our next trial to include rigorous attention to inflorescence onset for timing of watering.

Table 5. Summary of disease and pest pressures encountered in the variety trial.

Results of SRI Trial

Across the three varieties, 22 days of age at time of transplant produced robust, resilient transplants. Whereas the yield of 22 day-old plants was not significantly higher than 16- or 28- day plants, there was an observable trend across varieties, with plants older than 22 days old producing the same amount or less of rough yield (pre-winnowing) than 22-day old plants. This may be attributed to 28-day plants having exhausted more of their nutrient stores while in trays, becoming rootbound, and/or having less plasticity to become established in the field compared to younger plants. These plants did survive transplant, but my have diverted less energy to seed production overall. The average yield for seedling age at transplant across the three varieties trialed is summarized in Table 6 below.

Younger plants that did survive were able to become vigorous and productive; however, we observed that across varieties, 16-day old plants were less likely to survive transplant. The most significant driver of plant loss of 16-day-old transplants was the small size of these seedlings at transplant time. There were several cases in which the plants were too short to peek up over the plastic, partially due to plant height and partially due to the inconsistent level of ground beneath the plastic row cover. In addition, some of these 16-day-old seedlings did not have large, relatively established root systems and may have lacked the resilience to survive transplant shock, compared to 22-day transplants. In order to account for both of these variable factors, we would not recommend transplanting seedlings less than 5-6 cm tall. 

See Figure 2 for a simple visualization of the effect of transplant age in the SRI trial on yield across the three varieties trialed. Although the 22-day plants were not significantly higher yielders than 28 day old plants, we can see that the yield begins to trend downward after 22 days, with Yukihikari experiencing this downward trend most noticeably. 

We observed negligible disease pressure in the SRI trial, likely (and luckily) because it was set up diametrically opposed within our rice field to the origin of the infection in the Kyzyl Shala plots. 

Flowering, Harvest Date, and Yield

Across all the transplant ages included in the SRI trial, Loto had the latest heading date, that is, Loto plants were older when they reached 50% flowering within the Loto plots compared to the Estrella and Yukihikari plots (see Figure 3). Yukihikari was one of the first varieties to flower (across both SRI and variety trials). Across all varieties, seedlings transplanted later (28, 25 day seedlings) flowered later (Figure 4) but reached maturity earlier (Table 7). As also seen in Figure 4, varieties did differ in their sensitivity to transplant date: Estrella seedlings responded most strongly to transplant date, with its younger seedlings reaching 50% flower emergence much sooner than older seedlings, whereas this relationship was less pronounced in Loto seedlings, which simply flowered later in the season overall. Further, while Yukihikari plants flowered mostly densely, with more flowers per plant (Figure 5A), its flowering density was less sensitive to seedling age at transplant compared to Estrella and Loto seedlings across transplant dates (Figure 5B). This shows us that Yukihikari is most resilient to transplant date with regards to number of heads/tillers produced and the date of those heads' emergence, making it a good choice for farmers needing flexibility in transplant dates. Meanwhile, Estrella and Loto remain good choices for farmers who may need to transplant later due to risks of late frost -- perhaps those in colder microclimates or higher elevations-- without impacting flowering density (Figure 5B).

Table 6. The average yield for seedling age at transplant across varieties included in the SRI trial, separated by transplant age.

Table 7. Days to harvest amongst SRI trial plots. Across all varieties, the trend seemed to be that plants which were older at time of transplant matured more quickly, despite flowering later in the season. 

Understanding the relationship between seedling transplant age and any possible benefits to plant success is central to trying out SRI in the Northeast, where the season is relatively short compared to other rice-growing regions. Therefore, we are interested in finding out what the drivers of yield might be within this system. To that end, we compared the yield of the varieties used in the SRI trial on the basis of their flowering density at Week 9 of the trial. At that point of sampling, it does not appear that flowering density predicts yield; in fact the trend was the opposite, showing our highest yielder (Loto) as having the lowest flowering density at mid-to-late season (Figure 5B). We will repeat this measurement over a longer span of time in our 2026 trials to account for what we eventually observed to be a much later timing of peak flowering in Loto, which may more accurately predict yield. See Figure 5 in the document attached for a breakdown of yield at each transplant age as it relates to 50% flowering date (heading date). 

While plants that were younger (16-day) at transplant flowered earlier in the season, with their heading date occurring at 85 days or less (with the exception of one outlier), this earlier flowering did not significantly correspond to greater yields (Figure 6). There is a slight trend wherein plants transplanted at later age yielded slightly more across all varieties; though this trend was not statistically significant, the trend appears to reflect a greater sensitivity of younger plants overall. Transplants that were older at time of transplant did flower later in the season, by almost 2 weeks on average, yet this only seemed to result in slightly greater yields. Those that did flower sooner amongst their cohort yielded slightly less grain in total, underscoring a need to understand determinants of flowering onset in future studies. While not showing statistically significant trends, these data are still informative: this could encourage future growers with exceptionally short growing seasons, even further north than our farm, to transplant as close to last frost as possible, as yield did not seem to be impacted by the earlier heading date of younger transplants (see yield data in Table 6). Further, waiting longer to transplant did not necessarily lead to higher yields, so growers in the northeast can be confident that they can transplant sooner, as long as the seedlings are of sufficient height and stature, without risking an impact on yield. Growers should keep in mind, though, that younger plants did take longer for grain to mature (see Table 7), so a balance must be struck between these variables to avoid running into risk of first frost; this is another reason why a "middle ground" at around 22 days seemed to be a measured recommendation.

Reliability of Yield at Different Transplant Ages

Growers may be interested in the performance reliability of these earlier-aged transplants in addition to total yield. Figure 7 visualizes that information by providing a curve for each yield, with a narrower bell curve indicating that more of the plots sampled had yields relatively similar to each other, whereas broader curves indicate greater inconsistencies in yield between samples. Although Yukihikari consistently yielded the least, 22-day-old seedlings demonstrated relatively consistent production, whereas 16-day-old and 35-day-old seedlings demonstrated a broader distribution. Although data were missing for Estrella 22-day-old seedlings, the variety seemed to be exhibiting a similar trend to Yukihikari, with the yield broadening as transplant age increased. Predictability is an important consideration for farmers attempting to decide on the distribution of their crop planning each year, so planning to transplant out around 22 days seems to confer greater confidence in those projections for Yukihikari and Estrella, whereas growers of Loto might wait another week to transplant for higher and more consistent yields (Figure 7). Within our 2026 trials, seedlings were between 21 and 29 days old at time of transplant, so we will analyze these data again to find out whether this trend holds up.

Issues Encountered

Dehulling Inconsistency

Many dehullers are designed to process larger quantities of rice than we yielded from many of our experimental plots. In addition, the so our ability to dehull rice in a uniform manner was inconsistent between yield quantities. One small–quantity dehuller we had access to was designed for short-grain rice, and therefore it was able to sufficiently dehull our Yukihikari yields, but caused breakage amongst the longer-grained varieties, including Estrella and Delmati; we decided to halt dehulling in order to preserve more of our grain yield. Therefore, we decided to proceed with data analysis using yield data from our lots of rice before dehulling. 

Disease Outbreak: Rice Blast

Another factor impacting the yield was disease outbreak. Blast, a common fungal disease affecting rice, showed up in one variety, Kyzyl Shala and impacted some of the surrounding plots in one area of the field. We ended up rogueing the worst of the diseased plots as to try and stem the spread, though from what we learned about Blast, if it's in one area of the field it's likely infected most of the field. We surmise that Kyzyl Shala may not have had any resistance to blast that might have been present in any more modern varieties; as a landrace from Uzbekistan, it may not have encountered this or other North American pest pressures. This is something we plan to monitor for in the future-- though we would like to incorporate landrace growing into our system for the preservation of culture and resilience, we will need to anticipate interesting trends and responses to modern pressures. Additionally, we learned how important it is to anticipate rice developmental stage with regards to irrigation planning. Blast tends to show up right as plants are about to head out, and it affect plants more that are not in wet soil. Since we are upland rice producers, we were prepared to water every day once flowering begins, however, the level of observation as to how quickly some varieties flower means we need to 1) anticipate the day of flowering and 2) begin soaking the soil before that time. We are going to add monitoring for the "boot" stage-- when an inflorescence begins to form, but has not yet emerged -- and beginning to soak the soil as close to that time as observations allow-- to our rice management and irrigation protocol.

Although it may have been present at our Clermont, NY lease in dormant or dessication-resistant spores in the soil, it is possible that the blast may have been seed borne. This showed us the importance of seed treatment before planting. Some other rice farmers a few towns over taught us their method for pre-seeding water treatment - keeping the seed at a consistent 140F for 7 minutes is supposed to eliminate Blast and a number of other seed borne pathogens from the seed.  We completed this type of seed treatment, in addition to an oxidate soak, prior to our 2026 planting and will be carefully monitoring for blast and other disease pressures once again. 

We did notice, interestingly, that the blast spread throughout the buffer row of rice (variety Estrella) adjacent to the SRI and variety trials, but did not spread to the SRI trial plots to any observable extent. The spread gradually tapered off in severity as it spread, but it seems that closely spaced plants are likely to be able to spread blast more effectively. We believe the continuity of this planting (same age, same cultivar, one continuous bed) may have facilitated this spread, which underscores the importance of monitoring for blast in large-scale upland plantings.

Public Engagement and Interest

We also learned that there is a significant amount of interest in rice in our region. We had over 80 people come to our SARE field day, and later hosted a Mid-Hudson CRAFT agricultural education event to further discuss seed saving, with our rice at the center of the conversation.

Finally, we were able to host, along with a local farm-to-table chef, a community-oriented rice tasting and a ticketed dinner event centering rice dishes made with rice harvested from the trial. We worked closely with chef Pruitt at Eliza, in Kingston, NY, who is drawing from her Thai background and her knowledge of seasonal availability of produce as inspirations for her menus. She put together several dishes that were clearly Thai at heart but had interesting "northeastern farming" twists, especially making use of our Yukihikari polished rice. Photos from the event, including the menu, are attached to this final report as the final figures included. Please get in touch if you would like to see higher resolution images!

Coming-Home-Seeds-SARE-Report-2025-Figures