Optimum Heat Unit Summation Technique for Harvest Prediction of Fresh-Market and Pickling Cucumbers in North Carolina 

Cucurbit Genetics Cooperative Report 8:9-11 (Article 4) 1985

Perry, Katharine B. and Todd C. Wehner
Department of Horticultural Science, North Carolina State University Raleigh, NC 27695-7609

Heat unit summation techniques have found widespread applicability for predicting stages of development for many horticultural and agronomic crops (3). Boswell (2) first documented the concept of heat summations relative to vegetable crop production in 1929. We were interested in finding a method of predicting harvest maturity of cucumbers, especially one that could be applied to once-over harvest. A number of methods have been used to calculate heat unit summations. The objective of this study was to determine which of those methods predicted harvest with the least variation over years and seasons.

Methods. The data for this comparative study of methods to determine heat unit requirements were taken from 5 years (1980 through 1984) and 2 seasons (spring and summer) of plantings for 2 crops and 2 or 3 crop maturities (early, midseason and late fresh-market cucumbers, and early and mid-season pickling cucumbers). The experiment was run at the Horticultural Crops Research Station near Clinton, N. C. Daily high and low temperatures were recorded from planting through first harvest of each crop and maturity group for a multiple-harvest system. Standard cultural practices were used for all crops.

Number of days from planting through first harvest was used as the standard of comparison for the heat unit summations. The CV (coefficient of variation) over 5 years and 2 seasons for each crop and maturity group was used as the measure of how consistent each method was in predicting time of first harvest. Arnold (1) demonstrated that the CV is the appropriate statistic for identifying the measure with the best prediction ability. Heat unit summations were calculated using 14 different methods, and several different base and ceiling temperatures as follows.

1. The standard Growing Degree Day (GDD) computation was made as follows.
   GDD = Â (Mean-Base) (1) where Base=0, 10, 13, 15.5, or 18°C, and Mean-(daily maximum + daily minimum)/2, and data were summed over all days from planting to first harvest. There was no ceiling temperature.
2. Use daily maximum (Max) instead of Mean as follows.
   GDD = Â (Max-Base) (2) where Base is the same as in 1 above.
3. Use Ceiling to determine Max as follows. GDD = Â (Mean – Base), where Max is Max or Ceiling of 27, 29, 32 or 35 (whichever is lower).
4. Use Ceiling instead of Max as follows. GDD = Â (Max – Base), where Max is Max or Ceiling of 27, 29, 32 or 35 (whichever is lower).
5. If maximum is greater than the given ceiling reset maximum by subtracting the difference between the maximum and ceiling from the ceiling, then use Eqn. 1 for range of ceilings and bases as in Number 3.
6. If maximum is greater than the given ceiling reset maximum by
   subtracting the difference between the maximum and ceiling, then use Eqn. 2 for range of ceilings and bases as in Number 3.
7. If maximum is greater than the given ceiling, subtract the difference
   between the maximum and ceiling from the daily mean used in Eqn. 1 using the range of ceilings and bases from Number 3.
8. Sum Growing Degree Hours (GDH) by using Eqn. 1 for each hourly mean. Use range of ceilings and bases from Number 3.
9. Same as Number 8, but reset maximum as in Number 3.
10. Same as Number 8, but reset maximum as in Number 5.
11. Sum GDH accumulated during day time only.
12. Same as Number 11, but reset maximum as in Number 3.
13. Same as Number 11, but reset maximum as in Number 5.
14. Weight daily GDD accumulation by daylength as follows.
    GDD = Â (Mean-Base) x Daylength)

Results. Comparison of the CV’s indicated that the methods with the most stable ability for harvest prediction were 6 and 14. Both of those methods were better than using days from planting to harvest as the index. Since method 6 required less effort to calculate than method 14, we decided to investigate it further. The lowest CV’s occurred with bases of 15.5 and 18°C and a ceiling of 32°C, so we tried bases of 14.5, 15, 16 and 16.5 and ceilings of 31, 31.5 and 33°C. Those temperatures did not improve the CV over environments for the 5 crops and maturities tested. Thus, the method of choice would be Number 6 with a base of 15.5 and a ceiling of 32°C

 

Table 1. Coefficients of variation (CV) for 14 methods of calculating heat unit summations with varying base and ceiling temperatures (°C) vs. days from planting to harvest.

 

Method	Base	Ceiling	CV	Method	Base	Ceiling	CV	Method	Base	Ceiling	CV
1	0	none	6	4	0	32	7	6	0	32	8
	10	none	11		10	32	4		10	32	4
	13	none	15		13	32	5		13	32	4
	15.5	none	22		15.5	32	6		15.5	32	3
	18	none	32		18	32	8		18	32	3
2	0	none	6	4	0	35	6	7	0	35	6
	10	none	7		10	35	6		10	9	5
	13	none	8		13	35	7		13	35	6
	15.5	none	10		15.5	35	9		15.5	35	7
	18	none	14		18	35	12		18	35	10
3	0	27	7	5	0	27	8	7	0	27	8
	10	27	7		10	27	6		10	27	6
	13	27	10		13	27	7		13	27	7
	15.5	27	16		15.5	27	11		15.5	27	11
	18	27	27		18	27	22		18	27	22
3	0	29	6	5	0	29	7	7	0	29	7
	10	29	8		10	29	5		10	29	5
	13	29	11		13	29	7		13	29	7
	15.5	29	16		15.5	29	11		15.5	29	11
	18	29	25		18	29	18		18	29	18
3	0	32	6	5	0	32	6	7	0	32	6
	10	32	9		10	32	7		10	32	7
	13	32	13		13	32	11		13	32	11
	15.5	32	19		15.5	32	16		15.5	32	16
	18	32	28		18	32	24		18	32	24
3	0	35	6	5	0	35	6	7	0	35	6
	10	35	11		10	35	10		10	35	10
	13	35	15		13	35	14		13	35	14
	15.5	35	21		15.5	35	20		15.5	35	20
	18	35	31		18	35	29		18	35	29
4	0	27	9	6	0	27	13	8	15.5	none	20
	10	27	7		10	27	20	9	15.5	32	19
	13	27	7		13	27	23	10	15.5	32	17
	15.5	27	6		15.5	27	29	11	15.5	none	14
	18	27	5		18	27	39	12	15.5	32	12
4	0	29	8	6	0	29	10	13	15.5	32	10
	10	29	5		10	29	11	14	15.5	32	3
	13	29	5		13	29	11
	15.5	29	4		15.5	29	12	Days (planting
	18	29	5		18	29	14		to harvest)		10

 

Literature Cited

 

1. Arnold, C. Y. 1959. The determination and significance of the base
   temperature in a linear heat unit system. Proc. Amer. Soc. Hort.
   Sci. 74:430-445.
2. Boswell, V. R. 1929. Factors influencing yield and quality of peas.
   Maryland Agric. Exp. Sta. Bull. 306.
3. Wilson, L. T. and W. W. Barnett. 1983. Degree-days: An aid in crop and
   pest management. California Agriculture. Jan.-Feb. pp 4-7.