Article ID: D-25-00004
Lettuce (Lactuca sativa L.) is a major vegetable cultivated worldwide and an essential vegetable consumed throughout the year in Japan. However, the suitable temperatures for open-field and tunnel-covered cultivation have not been sufficiently clarified. The present study aimed to comprehensively and quantitively clarify the suitable temperatures for the cultivation of head lettuce in major production areas across Japan using 1-km resolution daily mean air temperature and statistical data on cropping type, production, and planted area, planting date, and harvesting date on the prefectural and municipal levels. The suitable temperature for open-field cultivation was found to be in the inter-quantile range (IQR) of 13.9-20.5°C. By contrast, the IQR in tunnel-covered cultivation, defined as the air temperature outside the tunnel, was estimated to be 4.3-9.8°C. Furthermore, the growing period tended to be longer at lower temperatures. These findings contribute to our understanding of the cultivability for head lettuce and may inform projections of the effects of climate change on lettuce cultivation.
Lettuce (Lactuca sativa L.) is one of the most widely consumed vegetables in the world. According to statistical data, in 2022, lettuce was cultivated in 101 of 200 countries or regions (FAOSTAT, 2022). Lettuce is also a major vegetable in Japan. According to the Act on Stabilization of Production and Shipment of Vegetables and its enforcement ordinance, lettuce is designated as a high-consumption vegetable (Act on Stabilization of Production and Shipment of Vegetables, 1966). Therefore, because it is cultivated throughout the year, consumers in Japan can obtain lettuce year-round. In other words, the year-round supply of lettuce has been realized (Kikuchi and Matsuno, 1977; Noguchi et al., 1981).
Lettuce is considered to be a plant that thrives in cool climates, which makes temperature a significant factor in terms of its growth. Growth indicators such as dry weight and leaf number increase in higher temperatures (Noguchi et al., 1981). In other words, the growth rate of lettuce stagnates at low temperatures but increases at high temperatures. On one hand, it is clear that some disorders occur under hotter conditions. For example, tip burn is primarily caused by calcium deficiency, but the promotion of an increase in dry weight due to high temperatures also accelerates tip burn (Collier and Tibbitts, 1982; Rubatzky and Yamaguchi, 1997). Flower bud differentiation and subsequent bolting can lead to bitterness and toughness, which are factors that reduce the quality of lettuce (Bisbis et al., 2018). These phenomena also occur as a result of high temperatures (Shibutani and Kinoshita, 1966; Rubatzky and Yamaguchi, 1997). On the other hand, disorders can also occur at colder temperatures. Leaf tip blight occurs on leaf lettuce that encounters freezing temperatures (Takahama and Jishi, 2021). Additionally, freezing causes epidermal peeling and death in head lettuce (Shibutani and Kinoshita, 1970; Takahama and Jishi, 2021). From another perspective, Tudela et al. (2017) indicated that the quality of fresh-cut lettuce is affected by the average temperature during the growth period. Accordingly, lettuce has an suitable temperature range that does not negatively impact yield or quality, including during postharvest (hereafter “suitable temperature”).
Despite its importance, the suitable temperature range for lettuce cultivation in Japan remains unclear. Numerous studies have investigated the relationship between lettuce yield or quality and temperature in open-field cultivation. However, these studies have mostly focused on specific locations, varieties, and years (e.g., Inako and Sakai, 1969; Hoshino et al., 1977; Okada et al., 1997; Takahama and Jishi, 2021). Furthermore, to achieve year-round supply, lettuce is cultivated in winter by covering the ridges with materials such as vinyl in a tunnel shape to elevate the air temperature inside compared with outside the tunnel. However, the suitable outside temperature for tunnel-covered cultivation also remains unclear due to lack of studies. Therefore, it is important to comprehensively and quantitatively evaluate suitable temperatures for both open-field and tunnel-covered cultivation.
The suitable temperature is considered valuable information for predictions in regard to climate change. Increasing temperatures may cause the aforementioned and other disorders for lettuce. In fact, poor head formation, tip burn, bolting, and unexpected earlier growth have already been reported in Japan (Ministry of Agriculture, Forestry and Fisheries of Japan, 2023). As climate change progresses, it cannot be denied that the most suitable areas for open-field cultivation may change (Iizumi, 2019). Conversely, in regions where tunnel-covered cultivation is practiced during winter, increased temperatures may shorten the tunnel-covered period. Clarifying the suitable temperature for lettuce cultivation should make it possible to predict when and where these phenomena will occur.
Given this background, the present study aimed to comprehensively and quantitatively clarify the suitable temperatures for both open-field and tunnel-covered cultivation of head lettuce in Japan based on statistical and meteorological data. The statistical data were based on a prefectural-scale cropping type catalog indicating the specific locations where head lettuce is actually being cultivated, and this represents a pivotal aspect of the present study. Entries in this catalog indicate that lettuce in that location is cultivated with both sufficient yield and quality. Consequently, it is possible to evaluate the suitable temperature for head lettuce cultivation by analyzing the daily mean air temperature at each location.
The Cropping Type Catalog for Each Type of Vegetable, 2009 Edition (National Institute of Vegetable and Tea Science, 2009; hereafter “Cropping Type Catalog”) was used. This catalog records the cropping type, including sowing, planting, beginning and end of harvest, and application of specific materials, for each vegetable in each prefecture. This information is indicated using symbols on a calendar, with a resolution of approximately 5 days. In this study, we read the planting date and beginning of harvest from the calendar for each cropping type. The planting date, rather than the sowing date, was chosen because head lettuce is typically sown in cell trays and then planted to the field after the seedlings have grown; some cropping types also involve temperature control during this period. Similarly, the beginning of harvest was selected because head lettuce is harvested at varying times within its overall harvest period. Therefore, we specifically focused on the period when lettuce is definitively present in the field. For example, the fourth six divisions in September was defined as the 21 September (Fig. 1b). Additionally, if a cropping type applied tunnel-covered, the beginning and end dates of the tunnel installation were used.
Fig. 1. Conceptual diagram of meteorological and statistical data processing. (a) Select municipality and location. (b) Read cultivation date from the Cropping Type Catalog and map from prefecture to municipality. (c) Combine (a) and (b) to obtain the daily mean air temperature for the lettuce growing period at the selected location.
The crop situation and cultivated area included in Crop Statistics Data from 2007 to 2009 (Ministry of Agriculture, Forestry and Fisheries of Japan, 2007-2009) were used for the analysis. The lettuce planted area (ha) and production (t) in each municipality were obtained from the crop situation data. The planted area and production were recorded for each of the three cropping seasons: spring (harvest from April to May), summer-fall (June to October), and winter (November to March). The upland field area (ha), defined as the cultivated land except for paddy fields in each municipality, was obtained from the cultivated area data.
2.3. Meteorological and land use grid dataIn this study, the daily mean air temperatures obtained from the Agro-Meteorological Grid Square Data were adopted. These data were provided in approximately 1 km2 grid resolution based on spatial interpolation of observations by the Automated Meteorological Data Acquisition System (Ohno et al., 2016). We used the 10 years daily mean air temperature for the period from 2000 to 2009. However, when the cropping seasons crossed years, data from 1999 were employed. Land use data for 2009, which had the same resolution as the Agro-Meteorological Grid Square Data, were obtained from the Ministry of Land, Infrastructure, Transport and Tourism of Japan (Ministry of Land, Infrastructure, Transport and Tourism of Japan, 2009). The percentage of “other agricultural land” area greater than 0% in this data set was used as the upland field.
2.4. Merging of the Cropping Type Catalog and Crop Statistics DataThe spatial resolution of the Cropping Type Catalog was in prefectures and the temporal resolution was approximately 5 days (Section 2.1). On the other hand, the spatial and temporal resolutions in the Crop Statistics Data were the municipality and the three cropping seasons, respectively (Section 2.2). The former had relatively high temporal resolution, while the latter had relatively high spatial resolution. These two data sets were merged to produce accurate high spatio-temporal resolution data for assessing the suitable temperature for head lettuce cultivation.
(1) First, we selected municipalities and locations. The total lettuce planted area was divided by the upland field area, and averaged for 3 years (2007-2009). To ensure the reliability of the analysis, 30 municipalities with a value > 0.15 were selected. These municipalities account for 74.7% of the lettuce production in Japan. Next, the upland grid within those municipalities was selected (Fig. 1a). This procedure identified the location of upland cultivation grids.
(2) Next, we focused on the harvest season to exclude inappropriate cropping types. The cropping season in the Crop Statistics Data was used as a reference, and if it did not match the harvest period in the Cropping Type Catalog, the cropping type was excluded. As a result, 55 cropping types were selected. Next, these cropping types were mapped to the municipalities (Fig. 1b), resulting in 183 combinations of municipality and cropping types (hereafter “cropping patterns”). Through this procedure, we determined the planting and other cropping dates in each municipality.
(3) Finally, (1) and (2) were combined. That is, daily mean air temperature was obtained at the location selected by procedure (1) and for the range of the growing period determined by procedure (2) (Fig. 1c).
In this study, the growing period was defined as the time from the planting to the beginning of the harvest. Some crop patterns involved tunnel-covered cultivation for the majority of the growing period, whereas others involved tunnel-covered cultivation for only a small portion of the growing period. To clarify the contrast with open-field cultivation, only cropping patterns involving tunnel-covered cultivation for 80%-100% of the growing period were used.
In total, there were 108 open-field cultivation cropping patterns, which accounted for 59% of the total 183 cropping patterns. The relative frequency distribution for the 10 years daily mean air temperature during the growing period of head lettuce is shown in Fig. 2, and the percentile values are shown in Table 1. In this case, the 25th, 50th, and 75th percentiles were 13.9, 17.4, and 20.5°C, respectively (Table 1). The skewness of the daily mean air temperature was -0.36. In other words, the distribution had a longer tail at low temperatures and a sharper tail at high temperatures. In an experiment involving head lettuce in a phytotron, the ideal temperature for leaf differentiation was approximately 20.0°C, and the highest increase in leaf weight was reported at an average temperature of approximately 15.0°C (Kato, 1972). The former is closest to 19.9°C at the 70th percentile and the latter to 14.7°C at the 30th percentile. Hoshino et al. (1977) estimated that the temperature at which the growth rate of head lettuce, which corresponds to the time change in dry matter weight, was maximized at 17.2°C. This is close to the value of 17.4°C for the 50th percentile. Globally, Maynard and Hochmuth (1997) introduced the minimum, best lower, best upper, and maximum temperatures for lettuce growth as 7°C, 12°C, 21°C, and 24°C, respectively. These temperatures correspond approximately to the 5th, 20th, 80th, and 95th percentiles, respectively. Consequently, the temperature range for open-field cultivation is consistent with previous studies.
Fig. 2. Relative frequency of the 10 years daily mean air temperature between head lettuce transplanting and the beginning of the harvest in open-field and tunnel-covered cultivation. Solid and dashed lines indicate open-field and tunnel-covered cultivation, respectively. Open-field cultivation is calculated using 108 cropping type patterns for 10 years, while tunnel-covered cultivation is calculated using 49 cropping type patterns for 10 years. The skewness of open-field and tunnel-covered cultivation is -0.36 and 0.60, respectively.
Table 1. Percentiles of daily mean air temperature between transplanting and the beginning of the harvest for both open-field and tunnel-covered cultivation.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/T1.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
3.2. Tunnel-covered cultivation
Tunnel-covered cultivation accounted for 27% of the total (49 cropping patterns). In this case, the 25th, 50th, and 75th percentiles of the 10 years daily mean air temperature were 4.3, 6.6, and 9.8°C, respectively (Table 1). The 50th percentile values for tunnel-covered cultivation were 10.8°C lower than those for open-field cultivation. This was as expected and clearly indicated that tunnel-covered cultivation is practiced in lower temperatures than open-field cultivation. The skewness of the distribution was 0.60, which was the opposite trend from open-field cultivation, with a longer tail at high temperatures and a sharper one at low temperatures. This suggests a high priority on avoiding the adverse effects of low temperatures. Inako and Sakai (1969) tested head lettuce yields for various types of heat-retention materials in winter. For plastic tunnels, the daily mean air temperature during the growing period varied from 5 to 13°C, which corresponded to the 30th and 90th percentiles, respectively. However, few studies have been conducted on outside temperatures in tunnel-covered cultivation, so no further comparisons could be made.
From another perspective, studies have been conducted on the temperature difference between inside and outside tunnels. Therefore, we indirectly evaluated the validity of our results regarding this temperature difference by comparing them to the findings from previous studies. According to Sato et al. (2008), the intercept of a linear regression of air temperature at an outside tunnel against that at an inside tunnel was 11.8°C during the daytime (10:00-17:00) and 4.2°C at night (18:00-07:00). In addition, Kitada and Okada (1988) examined the temperature difference between inside and outside tunnels for daikon radish when the tunnel material and ventilation rate were varied. The results showed that in the case of an absence of ventilation, the temperature inside the tunnel was on average 5.5-6.7°C higher than that outside the tunnel. Based on these results, even when the rate increases from 4.2 to 11.8°C were added to the relative frequency distribution for tunnel-covered cultivation, they did not exceed those for open-field cultivation, which suggests the validity of the results regarding the temperature range for tunnel-covered cultivation.
In relation to Fig. 1, head lettuce is generally cultivated in open fields from approximately spring to fall. In spring and fall, it is grown across a relatively wide geographical area. In summer, due to the need for cooler temperatures, head lettuce cultivation shifts to highland regions. In contrast, tunnel-covered cultivation is employed in winter to cope with severe cold. The main production areas for winter head lettuce are consequently located in the warmer western regions of Japan. Although these cultivation patterns are well known, such information is also reflected in Fig. 1. Previous studies have typically focused on specific regions, cultivars, and years. In contrast, this study was able to estimate the temperature range during head lettuce growing period across the entirety of Japan, as shown in Fig. 1 and Table 1.
3.3. Relationship between the growing period and mean air temperatureThe relationship between growing period and mean air temperature is shown in Fig. 3. The shortest and longest growing periods were 30 and 120 days, respectively. The mean air temperatures in open-field and tunnel-covered cultivation ranged from 9.6 to 23.1°C and from 4.0 to 12.8°C, respectively. In both open-field and tunnel-covered cultivation, the growing period tended to be longer as the mean air temperature decreased. Generally, these relationships are explained using growing degree days (GDD). However, GDD did not sufficiently explain the relationship. Specifically, in open-field cultivation, the GDD during the growing period ranged widely, spanning from a minimum of 485°C to a maximum of 1349°C. While differences in earliness among variety might have influenced this temperature variability, we were unable to further investigate this due to the lack of specific data on variety earliness. In tunnel-covered cultivation, the growing period varied from 80 to 120 days, despite a relatively consistent mean air temperature of approximately 6°C. The cropping types in this case were distributed with the planting in late November to late December, with the beginning of harvest generally in late March to late April. This indicates that head lettuce was cultivated during the coldest season. We hypothesize that growing period mean air temperature of around 6°C represents the lower limit for head lettuce in tunnel-covered cultivation. Conversely, this suggests that maintaining an average temperature of approximately 6°C from planting to the beginning of harvest is crucial during severe winters. However, we were unable to find previous research to directly support this explanation.
Fig. 3. Relationship between growing period and mean air temperature. Open circles and crosses represent open-field and tunnel-covered cultivation, respectively. Each mark indicates the 10-year mean for each cropping pattern. Solid and dashed lines represent locally weighted scatterplot smoothing using the R statistical package (R Core Team, 2024) for open-field and tunnel-covered cultivation, respectively.
Dufault et al. (2009) demonstrated that a 1°C increase in the maximum and minimum air temperature shortened the growing period by 5 days in open-field romaine lettuce. The results of the present study in regard to open-field cultivation indicated a value of 5.2 days °C-1, which was comparable to 5 days °C-1 reported by Dufault et al. (2009). Okada et al. (1997) conducted experiments with head lettuce in open fields and tunnels to develop a growth model. According to their results, the mean air temperature and number of days during the growing period in the open field were about 11°C and 60 days, respectively. In the case of tunnels, these values were about 8°C and 120 days, respectively. In both cases, no significant gaps were found compared with the results obtained in the present study.
However, Fig. 3 shows that even for the same growing period, the mean air temperature varied. This is mainly due to the data used in the present study. The first cause is the imprecision of the growing period. As mentioned in Section 2.1, the growing period was evaluated with temporal resolution of approximately 5 days. However, in actual fields, planting and harvesting occur daily. In other words, the temporal resolution of the growing period in this study was coarser than reality. The second cause is the uncertainty of the location. In this study, a non-paddy field grid was selected, as mentioned in Section 2.3. Therefore, the grid includes not only lettuce, but also fruit trees or non-lettuce vegetables. As a result, it is clear that the air temperature in that grid includes non-lettuce fields.
To eliminate these uncertainties, accurate location and detailed cultivation data are needed. It seems that there are mainly two methods to achieve this. The first is to collect location and cultivation records for every single field, which would allow evaluations based on exact locations and growing periods. However, it can be difficult to collect such data on a national scale. The second method is to use remote sensing. Recently, light detection and ranging (LiDAR) has been used for mapping crop types (e.g., Prins and Van Niekerk, 2020; Jayakumari et al., 2021). Once LiDAR can be applied to head lettuce, it will be possible to determine the cultivation location and growing period. Based on the above discussion, it can be said that although there are uncertainties in the methods used in this study, they are the best available at this time to assess the suitable temperature for head lettuce cultivation in Japan.
This study was supported in part by the Environment Research and Technology Development Fund (JPMEERF20S11820 and JPMEERF23S21120) of the Environmental Restoration and Conservation Agency of Japan, and JST grant number JPMJPF2013.
