Modeling the Relationship between Apple Quality Indices and Air Temperature

Abstract

The future impacts of climate change on the yields of staple crops have been widely investigated. However, owing to insufficient data compared to that available for staple crops and the complexity of the quality determination process, the relationship between quality changes in horticultural crops and climate change has not been quantified, and potential future changes in fruit quality are not well understood. We conducted temperature treatment experiments to quantify the sensitivity of apple quality to air temperature and collected quality indices records through field observations to propose a model for estimating apple quality indices from the air temperature. In the temperature treatment experiment, ‘Fuji’ apple trees were placed in glass chambers set at a constant temperature of 17.3–25.6°C from 110 days after full bloom (DAFB). The fruits were harvested at 170 DAFB to measure the quality indices. The results indicated that the acidity and ratings for blush, peel ground color, starch disappearance, and watercore were all significantly lower at higher air temperatures. The relationship between these quality indices and air temperature could be linearly approximated. Sugar content and fruit firmness were not clearly affected by the air temperature. In addition, data from field observations conducted at experimental orchards in Aomori and Nagano Prefecture over 50 years (1970–2019) were analyzed. The relationship between fruit quality indices at 170 DAFB and the mean temperature in the 60 days from 110 DAFB in Aomori and Nagano was similar to that observed for the chamber experiment results, and no significant difference was observed in the slope of the linear regression equation between the chambers and orchards. A model was developed to estimate the fruit quality indices based on air temperature by accumulating daily amounts of change in quality indices calculated from daily mean temperatures using the results of experimental and field observations. The model could be used to assess the impact of future long-term temperature increases on apple quality indices.

Introduction

Global surface temperature is expected to increase throughout the 21st century unless greenhouse gas emissions are substantially reduced in the coming decades (IPCC, 2021). The future impacts of climate change on crops have been widely investigated; however, most studies have focused on the yields of staple crops such as maize, wheat, and rice (Hong et al., 2020). The impacts that climate change will have on the quality of horticultural crops remain unknown, owing to insufficient data compared to that available for staple crops and the complexity of the quality determination process.

A long-term observation regarding changes in apple quality (Sugiura et al., 2013) has suggested a possible influence of warming. However, the causes of quality index changes have not been identified, as fruit quality is affected not only by changes in environmental conditions, but also by changes in human factors, such as cultivation methods and criteria for determining harvest dates (Kingston, 1992; Musacchi and Serra, 2018).

The link between phenological changes in perennial crops and spring climate change is well known (Aono, 2015; Chmielewski et al., 2011; Fujisawa and Kobayashi, 2010; Grab and Craparo, 2011; Menzel et al., 2020; Vitasse et al., 2022). Phenological changes may affect fruit quality because they change the number of days from flowering to harvest, but the effects on fruit traits are unclear. Predicting future fruit quality indices is essential to guide orchardists when selecting cultivars because fruit tree replanting takes a long time (Palmer et al., 2003). However, the relationship between quality indices change and climate change has not been quantified, so future changes in fruit quality are not well understood.

In this study, physiological experiments were conducted over a range of air temperature conditions to quantify the sensitivity of apple quality indices to air temperature. The aim of the study was to develop a model for estimating quality indices from the air temperature using these results and the quality indices records of ‘Fuji’ apples over 50 years (1970–2019) at Aomori and Nagano.

Materials and Methods

Chamber experiment

Temperature treatment experiments for ‘Fuji’ apple trees (Malus domestica Borkh.) were done at the Institute of Fruit Tree and Tea Science, NARO, Tsukuba, Ibaraki Prefecture, Japan (36.0°N, 140.1°E; elevation 23 m) to investigate the effects of air temperature on fruit quality indices.

During the fruit maturation period in three seasons (2005, 2010, and 2014), trees (8–14 years old and 2.0–2.3 m tall) planted in 25-L pots containing horticultural soil were subjected to temperature treatments in glass chambers (3 m tall, 15 m²) exposed to natural light at a constant temperature of 17.3–25.6°C. The trees were placed in the glass chambers 110 days after full bloom (DAFB), and the fruits were harvested at 170 DAFB to measure the quality indices. The numbers of temperature-treated trees and fruits for which quality indices were measured are shown in Table 1.

Table 1

Plant materials and experimental plots in the chamber experiment.

Long-term field observations

The records (Sugiura et al., 2013) collected from experimental orchards at Aomori Prefectural Industrial and Technology Research Center, Kuroishi, Aomori Prefecture, Japan (40.6°N, 140.6°E; elevation 70 m; ando soil) and Nagano Fruit Tree Experiment Station, Suzaka, Nagano Prefecture, Japan (36.6°N, 138.3°E; elevation 360 m; brown lowland soil) for 50 years (1970–2019) were used as the field observation data for the quality indices and full bloom date of ‘Fuji’ apples.

Apple trees were grown in the same orchard at each prefectural institute during the observation period. In each experimental orchard, the apple trees were consistently managed according to the management system recommended by Aomori (Aomori Prefecture Apple Production Guideline Editing Group, 2012) and Nagano (Nagano Prefecture, 2006) Prefectures, including factors such as soil management, pruning, hand-pollination, and fruit thinning.

The full bloom date was observed visually and was defined as the date when ≥ 70% of all the king flowers bloomed. Approximately 20 fruits were randomly picked from each orchard every 5–10 days for 40–50 days during the maturation period to measure the acidity, sugar content, fruit firmness, and ratings for watercore, peel blush, peel ground color (only in Nagano), and starch disappearance (only in Aomori). Ground color and starch disappearance are often used as harvest indices in the respective regions.

Quality indices values at 170 DAFB were determined by interpolating the values measured every 5–10 days between the start and end dates of the annual observations with a cubic function.

Measurement of fruit quality indices

The fruit quality indices in the chamber experiment and field observations were measured according to the evaluation methods in a national trial of selections and standard varieties of Japan (NARO Institute of Fruit Tree Science, 2007). The peel blush (0 = none and 1 = delicate red to 6 = dark red) and ground color (1 = green to 8 = deep yellow) of each picked fruit were assessed visually by comparing apples with a Fuji standardized color chart (Yamazaki and Suzuki, 1980). Fruit firmness was assessed via the Magness-Taylor’s pressure test using a manual penetrometer (Kanamaru et al., 2021) with an 11-mm plunger after removing the skin on two equator positions on the opposite sides of each fruit.

Soluble-solid concentration (sugar content, °Brix) was measured from the flesh, which was cut from two different equator positions using a hand-held refractometer. The titratable acid concentration (acidity) was measured in approximately 2 mL of juice from the flesh of each fruit through titration with 0.1 N NaOH and expressed as a rate of malic acid. The extent of watercore (0 = none to 4 = severe) was visually determined using a standardized chart (Aomori Prefecture Apple Production Guideline Editing Group, 2012) after the fruits had been cut transversely. For starch level (SL), a transverse fruit section was dipped into a solution of 1% iodine and 5% potassium iodide, and the color was compared with that on a stained starch pattern chart (0 = completely white to 5 = completely black) (Aomori Prefecture Apple Production Guideline Editing Group, 2012; NARO Institute of Fruit Tree Science, 2007). The starch disappearance rating was defined as 5 − SL.

Climate data

We used the daily mean temperatures in the Agro-Meteorological Grid Square Data (Ohno et al., 2016), a climate dataset with 1-km resolution (each grid cell measures 45” longitude × 30” latitude) based on records at observatories from 1981 for the daily mean temperatures in the Aomori and Nagano experimental orchards from 1981–2019.

The daily mean temperatures before 1981 were recorded at the observatory (Statistics from Japan Meteorological Agency, https://www.data.jma.go.jp/obd/stats/etrn/ indices. php, December 9, 2021) in Aomori (until 1975, 40.8°N, 140.8°E, 24 km northeast from the experimental orchards in Aomori) or Kuroishi (after 1975, 40.7°N, 140.6°E, 4 km northwest from the experimental orchards in Aomori) and Nagano (36.7°N, 138.2°E, 11 km west of the experimental orchards in Nagano). They were corrected using the daily difference in the 20-year average of the daily mean temperatures between the observatory data and the Agro-Meteorological Grid Square Data for 1981–2000.

Statistical analysis

The relationships between quality indices and air temperatures were evaluated using Pearson’s correlation analysis. P values of the correlation coefficients (r) were obtained using a two-tailed t-test. Slopes between air temperatures and fruit quality indices for the chamber, Aomori, and Nagano were compared using analysis of covariance (ANCOVA).

Results and Discussion

Air temperature effects on quality indices

In the chamber experiment, acidity and ratings for blush, ground color, starch disappearance, and watercore were all significantly lower at higher air temperatures for 60 days (Fig. 1 and r and P in Table 2). The relationship between these quality indices and air temperature could be linearly approximated. However, air temperature did not clearly affect sugar content or fruit firmness.

Fig. 1

Response of quality indices to air temperature during the maturation period. Relationship between mean temperature (110–169 days after full bloom) and (a) blush, (b) ground color, (c) starch disappearance, and (d) watercore ratings, as well as (e) acidity, (f) sugar content, and (g) fruit firmness at 170 days after full bloom in the chamber experiments and field observations from Aomori and Nagano orchards. Approximation lines derived from equations (1) or (2) are shown.

Table 2

Relationship between air temperature and fruit quality indices as well as quality estimation model parameters.

The measurements were compared with long-term field observation records for ‘Fuji’ apples at the experimental orchards in Aomori and Nagano Prefectures, two of Japan’s most important apple-producing regions. The relationship between quality indices at 170 DAFB and the mean temperature during 60 days from 110 DAFB in the orchards was similar to that obtained from the chamber experiment results (Fig. 1). The relationships with air temperature were significant for acidity, blush, ground color, starch disappearance, and watercore, but not for sugar content or fruit firmness (r and P in Table 2). Although the air temperature in the chamber experiment was higher than the mean temperature range in both prefectures (11.9–22.3°C), no significant difference was observed in the slope of the linear regression equation in the chambers and the Aomori and Nagano orchards (P* in Table 2).

Previous chamber studies that used several of the other widely grown apple cultivars showed that the values of acidity (Tromp, 1997; Yamada et al., 1994), watercore (Yamada et al., 1994), blush (Blankenship, 1987; Honda et al., 2014), ground color (Olsen and Martin, 1980), and starch disappearance (Yamada et al., 1994) decreased with higher air temperatures, whereas sugar content and firmness did not change (Blankenship, 1987; Tromp, 1997). Therefore, these temperature responses are common in many cultivars.

A model for the relationship between quality indices and air temperature

A model was developed to estimate the quality indices based on air temperature using the results of chamber experiments and field observations. The linear regression of the relationship between the quality index values and air temperature in the chamber (Fig. 1) was as follows:

  

$$Q_{c\mathit{170}}={\mathrm{A}}_{\mathrm{c}}·T_c+{\mathrm{B}}_{\mathrm{c}}$$(1)

where Q_c170 is the quality index value in the chamber experiment at 170 DAFB; T_c (°C) is the treatment temperature (the mean temperature for 110–169 DAFB in the chamber); A_c is the slope; and B_c is the intercept.

The A_c values are shown in Table 2. All watercore values above 22°C were 0 (Fig. 1d); therefore, data from the chamber experiment ≤ 22°C were used for calculations. The A_c values for sugar content and fruit firmness were set to 0 as they did not relate significantly to the air temperature.

The slopes of the regression lines for quality indices at 170 DAFB and mean temperature for 110–169 DAFB among the chamber and field observations were not significantly different (P* in Table 2). Therefore, the annual quality indices at 170 DAFB in the field observation (Q_r170) could be approximated using the following equation with the A_c obtained from the chamber experiment:

  

$$Q_{r\mathit{170}}={\mathrm{A}}_{\mathrm{c}}·T_m+{\mathrm{B}}_{\mathrm{r}}$$(2)

  

$$T_m=\sum _{d=110}^{169}{T}_d/60$$(3)

where T_m (°C) is the mean temperature for 110–169 DAFB in each field and T_d (°C) is the daily mean temperature of the d th day after full bloom. The model assumes that the quality indices on a given day are determined by the air temperature up to the previous day. B_r is the intercept of the regression line when the slope is Ac, which varies by region. Each B_r value is shown in Table 2. The approximation line obtained from equation (2) is shown in Figure 1.

From equation (2) and (3), the amount of change in quality indices in the fields between 110 and 170 DAFB is as follows:

  

$$Q_{r\mathit{170}}–{\mathrm{Q}}_{\mathrm{r}110}={\mathrm{A}}_{\mathrm{c}}·T_m+{\mathrm{B}}_{\mathrm{r}}–{\mathrm{Q}}_{\mathrm{r}110}$$(4)

  

$$=\sum _{d=110}^{169}\left({\mathrm{A}}_{\mathrm{c}}·T_d+{\mathrm{B}}_{\mathrm{r}}-{\mathrm{ }\mathrm{Q}}_{\mathrm{r}110}\right)/60$$(5)

where Q_r110 is the quality index at 110 DAFB in each field and was set as a constant for each region. This equation can be expressed using the daily change from 110–169 DAFB, that is, the quality index change rate (QCR), as follows:

  

$$\mathit{QCR}\left(T_d\right)=({\mathrm{A}}_{\mathrm{c}}·T_d+{\mathrm{B}}_{\mathrm{r}}-{\mathrm{Q}}_{\mathrm{r}110})/60$$(6)

  

$$Q_{r\mathit{170}}={\mathrm{Q}}_{\mathrm{r}110}+\sum _{d=110}^{169}QCR\left(T_d\right)$$(7)

Although QCR is the value obtained from the quality indices measured at 170 DAFB, we assumed that the quality indices change in a similar manner after 170 DAFB, and equation (7) was extended as follows:

  

$$Q_{rD}={\mathrm{Q}}_{\mathrm{r}110}+\sum _{d=110}^{D-1} QCR\left(T_d\right)$$(8)

where Q_rD is the annual fruit quality index at any given D DAFB.

Q_r110 was not observed. However, because the ratings for watercore, blush, and starch disappearance before maturation are defined as 0, and the ground color rating is defined as 1, the Q_r110 values were assumed to be 0 and 1, respectively.

Q_r110 can be obtained by substituting a certain quality index observation (Q_rD) and the measured temperature value (T_d) into equations (6) and (8). Therefore, each year’s Q_r110 of acidity, sugar content, and firmness were obtained from the Q_rD and T_d identified in the first observation of each year in Aomori and Nagano and then averaged over the observation years to produce a constant for each region. If the first observation was not the maximum value for acidity, the maximum recorded value was used for the Q_r110 calculation. The Q_r110 values are shown in Table 2.

By using the observed T_d and constants of A_c, B_r, and Q₁₁₀ shown in Table 2 in equations (6) and (8), Q_rD could be estimated. The relationship between observed and estimated values on all observation dates in Aomori and Nagano is shown with the RMSE in Figure 2. This result suggests that there was no apparent difference between observed and estimated values.

Fig. 2

Relationship between observed and estimated quality indices in Aomori and Nagano. Relationship between observed and estimated values of (a) blush, (b) ground color, (c) starch disappearance, and (d) watercore ratings, as well as (e) acidity, (f) sugar content, and (g) fruit firmness. Each 1:1 line and root mean square error (RMSE) are shown.

In conclusion, our results indicate that equations (6) and (8) can be used as a model to estimate fruit quality indices from air temperatures during the maturation period. The mean temperature for 110–169 DAFB (horizontal axis in Fig. 1) in the chamber experiment (17.3–25.6°C) was higher than the mean temperature in field observations from the Aomori and Nagano orchards (11.9–22.3°C). Therefore, the model would be useful for assessing the impact of future long-term temperature increases on apple quality indices.

However, these results do not exclude the possible development of more accurate models in the future. For example, the relation between each quality index and air temperature after 110 DAFB was assumed to be constant in the present study, but the air temperature response may have changed up to the harvest date. Ground color tended to be overestimated at observed values below 2 and underestimated at values above 4 (Fig. 2b). This suggests a change in the temperature response of ground color. The temperature treatment period in this chamber experiment was 110–169 DAFB; however, more detailed models could be developed by experiments using various treatment periods. Chamber experiments over 26°C may also be necessary. Moreover, quality indices vary not only with air temperature during the maturation period, but also with other environmental conditions, such as light (Arakawa et al., 1985) and precipitation (Qu and Zhou, 2016). Acidity produced larger errors between observed and estimated values than other quality indices (Fig. 2e), but it may be possible to develop a more accurate model by considering these factors.

Acknowledgements

We thank all the researchers who have been conducting these field observations since 1970.

Literature Cited

• Aomori Prefecture Apple Production Guideline Editing Group. 2012. Ringoseisansidoyoko 2012–2013 (In Japanese). Aomoriken-ringokyokai, Hirosaki.
•  Aono,  Y. 2015. Cherry blossom phenological data since the seventeenth century for Edo (Tokyo), Japan, and their application to estimation of march temperatures. Int. J. Biometeorol. 59: 427–434.
•  Arakawa,  O.,  Y.  Hori and  R.  Ogata. 1985. Relative effectiveness and interaction of ultraviolet-B, red and blue light in anthocyanin synthesis of apple fruit. Physiol. Plant. 64: 323–327.
•  Blankenship,  S. M. 1987. Night-temperature effects on rate of apple fruit maturation and fruit quality. Sci. Hortic. 33: 205–212.
•  Chmielewski,  F.,  K.  Blümel,  Y.  Henniges,  M.  Blanke,  R. W. S.  Weber and  M.  Zoth. 2011. Phenological models for the beginning of apple blossom in Germany. Meteorol. Z. 20: 487–496.
•  Fujisawa,  M. and  K.  Kobayashi. 2010. Apple (Malus pumila var. domestica) phenology is advancing due to rising air temperature in northern Japan. Glob. Change Biol. 16: 2651–2660.
•  Grab,  S. and  A.  Craparo. 2011. Advance of apple and pear tree full bloom dates in response to climate change in the southwestern Cape, South Africa: 1973–2009. Agric. For. Meteorol. 151: 406–413.
•  Honda,  C.,  H.  Bessho,  M.  Murai,  H.  Iwanami,  S.  Moriya,  K.  Abe,  M.  Wada,  Y.  Moriya-Tanaka,  H.  Hayama and  M.  Tatsuki. 2014. Effect of temperature on anthocyanin synthesis and ethylene production in the fruit of early- and medium-maturing apple cultivars during ripening stages. HortScience 49: 1510–1517.
•  Hong,  C.,  N. D.  Mueller,  J. A.  Burney,  Y.  Zhang,  A.  AghaKouchak,  F. C.  Moore,  Y.  Qin,  D.  Tong and  S. J.  Davis. 2020. Impacts of ozone and climate change on yields of perennial crops in California. Nat. Food 1: 166–172.
• IPCC. 2021. Summary for Policymakers. p. 3–32. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge and New York.
•  Kanamaru,  K.,  K.  Kohyama,  K.  Konisho and  K.  Izumi. 2021. Instrumental evaluation of flesh texture in apple cultivars. Hort. J. 90: 270–279.
•  Kingston,  C. M. 1992. Maturity indices for apple and pear. Hort. Res. 13: 407–432.
•  Menzel,  A.,  Y.  Yuan,  M.  Matiu,  T.  Sparks,  H.  Scheifinger,  R.  Gehrig and  N.  Estrella. 2020. Climate change fingerprints in recent European plant phenology. Glob. Chang. Biol. 26: 2599–2612.
•  Musacchi,  S. and  S.  Serra. 2018. Apple fruit quality: Overview on pre-harvest factors. Sci. Hortic. 234: 409–430.
• Nagano Prefecture. 2006. Kajushidoshishin (In Japanese). Nagano Prefecture, Nagano.
• NARO Institute of Fruit Tree Science. 2007. Ikuseikeitotekioseikenteishiken-tokuseikenteisiken-chosaho (In Japanese). NARO, Tsukuba.
•  Ohno,  H.,  K.  Sasaki,  G.  Ohara and  K.  Nakazono. 2016. Development of grid square air temperature and precipitation data compiled from observed, forecasted, and climatic normal data. Climate in Biosphere. 16: 71–79 (In Japanese with English abstract).
•  Olsen,  K. L. and  G. C.  Martin. 1980. Influence of apple bloom date on maturity and storage quality of ‘Starking Delicious’ apples. J. Amer. Soc. Hort. Sci. 105: 183–186.
• Palmer, J. W., J. P. Prive and D. S. Tustin. 2003. Temperature. p. 216–237. In: D. C. Ferree and I. J. Warrington (eds.). Apples: Botany, Production and Uses. CABI, Wallingford.
•  Qu,  Z. and  G.  Zhou. 2016. Possible impact of climate change on the quality of apples from the major producing areas of China. Atmosphere 7: 113.
•  Sugiura,  T.,  H.  Ogawa,  N.  Fukuda and  T.  Moriguchi. 2013. Changes in the taste and textural attributes of apples in response to climate change. Sci. Rep. 3: 2418.
•  Tromp,  J. 1997. Maturity of apple cv. Elstar as affected by temperature during a six-week period following bloom. J. Hort. Sci. 72: 811–819.
•  Vitasse,  Y.,  F.  Baumgarten,  C. M.  Zohner,  T.  Rutishauser,  B.  Pietragalla,  R.  Gehrig,  J.  Dai,  H.  Wang,  Y.  Aono and  T. H.  Sparks. 2022. The great acceleration of plant phenological shifts. Nat. Clim. Chang 12: 300–302.
•  Yamada,  H.,  H.  Ohmura,  C.  Arai and  M.  Terui. 1994. Effect of preharvest fruit temperature on ripening, sugars, and watercore occurrence in apples. J. Amer. Soc. Hort. Sci. 119: 1208–1214.
•  Yamazaki,  T. and  K.  Suzuki. 1980. Color charts: Useful guide to evaluate the fruit maturation. 1: Colorimetric specifications of color charts for Japanese pear, apple, peach, grape, kaki and citrus fruits. Bull. Fruit Tree Res. Stn. A7: 19–44 (In Japanese with English abstract).