A CLIMATIC MODEL FOR PREDICTING GEORGIA'S PECAN PRODUCTION¹

Darrell Sparks
Department of Horticulture
University of Georgia
Athens, GA 30602

---

Humid climates are dynamic. Extreme deviations in climate greatly influence crop productivity. One extreme climatic event can affect pecan production for 2 or more years (Hunter, 1963). Long-term effects occur because fruiting in pecan, as in other tree crops, is a function of conditions existing in the year before and during fruiting (Davis and Sparks, 1974; Lockwood and Sparks, 1978). Pecan pistillate flowers are produced from carbohydrates accumulated during the previous year's growing season (Lockwood and Sparks, 1978), with the number of pistillate flowers being fixed by the time of budbreak (Wetzstein and Sparks, 1986). Thus, maximum potential nut production (fruit numbers) for the season has been determined at the time of budbreak and is a direct function of conditions that existed in the previous growing season. The potential nut production realized is dictated by conditions affecting subsequent fruit set and development.

Irregular bearing, a major problem in pecan production (Sparks, 1983), was proposed by Hunter (1963) to be dominated by pronounced climatic deviations.

Irregular bearing is minimal and production tends to be high when no extreme climatic deviations occur over consecutive years, as occurred during most of the 1980s. However, various climatic extremes may induce irregular bearing. Prolonged cloudy conditions and excessive rainfall during a heavy crop year may induce irregular bearing the next year. Severe, season-long drought can likewise induce irregular bearing (Hunter, 1963). Even severe, short-term droughts, common in late August and early September in the southeastern United States, can cause premature defoliation (Alben, 1958; Sparks, 1992c) that suppresses return bloom (Hinrichs, 1962; Moznette, 1934; Worley, 1979), especially on heavily fruiting trees (Sparks, 1983; Sparks and Brack, 1972).

The current year's climatic conditions largely control attainment of the pecan production potential by influencing the development from flower to fruit (Alben, 1958; Sparks, 1989a, 1995a, 1995b; Stein et al., 1989). Nut quality is good with optimum rainfall throughout the growing season. Fruit elongation is maximized in May and June; fruit expansion, from mid-July to mid-August (Sparks, 1995a); and kernel development, in September (Andrews and Sherman, 1980; Sparks, 1992a). Inadequate rainfall during one or more of these stages will negatively affect fruit development. Dry conditions during May, June or July produce small nuts (Sparks, 1989a) that become well filled if soil moisture is adequate during filling in September (Magness, 1955). The nut will be large but poorly filled under condition of adequate soil moisture during fruit elongation and expansion followed by inadequate moisture during kernel development (Andrews and Sherman, 1980; Magness, 1955; Sparks, 1992a). Kernel development and quality are suppressed by excessive rain during the growing season, (Hunter, 1963; Schaller et al., 1968; Sparks et al., 1995). On the other hand, inadequate soil moisture causes fruit abortion (Gammon et al., 1955; Sparks, 1989a).

Potential production is also influenced by scab, the major foliage and fruit disease in pecan. The development of scab is triggered by rains (Latham, 1982; Sparks, 1995b) and the disease can greatly suppress fruit growth and production of susceptible cultivars. Foliage is susceptible during leaf expansion in April and May, but not after leaves are fully expanded (Demaree, 1924; Gottwald, 1985; Valli, 1964). The fruit is susceptible from May to September (Gottwald and Bertrand, 1982). Scab infection on the foliage reduces photosynthesis efficiency (Gottwald and Wood, 1984), but the main effect of severe infection is suppression of leaf size and/or defoliation (Diener and Garrett, 1967). Damage from scab is most severe when the fruit is small in May and June. Damage decreases as fruit development progresses (Gottwald and Bertrand, 1982).

The apparent dominant relationship of nut production to previous and current climatic conditions suggests pecan production might be predicted with acceptable precision from climatic factors. The objective of this study was to develop a model for predicting Georgia's pecan nut production based on variation of climatic conditions. A model would be valuable to predict nut production and to assess the relative importance of individual factors on production.

---

MATERIALS AND METHODS

Production records (Fig. 1) used for modeling were USDA final gathered estimates for Georgia for the 48-year period 1945-92 (Snipes, 1995). The year 1955 was not modeled as most of the crop was destroyed by a late spring freeze (Hunter, 1963). Additionally, the freeze recovery period, 1956 and 1957, (Sparks, 1996) was not included here. The relationship of the USDA final gathered estimates for Georgia was compared with actual nut production from a single 946-acre orchard from 1950-92. The correlation (r2 = 0.725) lends support to the validity of the USDA estimates (r2 =1.0 would be perfect prediction).

Eight variables were used for modeling nut production (Table 1). If these eight variables really do determine pecan production in Georgia, then the model should predict the USDA estimates with acceptance accuracy. If not, the factors do not determine production, the USDA estimates are invalid, or both.

Production year was used to measure change, i.e., trend, in production with time. The relationship of nut production with time was positive (Fig. 1). Increased production with time is assumed to reflect increases in tree size and acres planted, improvements in tree nutrition; advances in pecan weevil, hickory shuckworm, black pecan aphid, and scab control; and some installation of solid set sprinkler and drip irrigation systems beginning in the late 1970's.

Pecan production in a given year is influenced by the production level during the previous year (Sparks, 1983 and Fig. 1). In the model, previous year's production was used as the index of this alternate bearing influence. Because high production tends to suppress production the following year, the relationship between current year and previous year production would be expected to be negative.

Intermittent, variable cloud cover is a normal climatic condition in a humid climate that influences sunlight and temperature in the tree canopy. Heating degree-days were used to estimate the effect of sunlight and temperature and are assumed to be an overall indication of ambient photosynthetic conditions. One heating degree-day was accumulated for each degree that the daily mean ambient temperature was above base 65oF. The mean temperature is the average of the maximum and minimum for the day. A 65o F base was chosen as suitable for pecan growth and development (Sparks, 1989b). Previous year's heating degree days were accumulated daily from 1 Apr. to 31 Oct. The interval from 1 Apr. to 31 Oct. is the period of photosynthetic activity for the pecan tree (Mielke, 1981). Clear skies and warm temperatures maximize photosynthesis and storage of carbohydrates for return bloom. Thus, production would be expected to increase with previous season's heating degree days, that is, the relationship would be positive.

The indicator used for scab severity during the current year was the highest cumulative sum of 2 or more days of consecutive rain occurring in either May, June, or in the first 15 days of July. For example in a given year, if the sum was 10, 2, and 9 days for May, June, and 1-15 July, respectively, 10 days were used. Two or more days of consecutive rain is critical because scab infection follows rain (Latham, 1982; Sparks, 1995b) with maximum infection following 2 days of continuous leaf wetness over a wide temperature range (Gottwald, 1985; Valli, 1964). May, June, and early July were selected because leaf infection is usually highest in May (Latham, 1982), and scab infections on the fruit in May, June, and early July can result in total fruit loss or unmarketable kernels (Gottwald and Bertrand, 1982). Scab infection occurring after July 15 is not considered to appreciably affect production. The assumption was made that scab damage is proportional to the most intense rain period. Thus, highest cumulative sum by month and not the total for the 3 months was used. Because of the destructive nature of scab, production would be expected to decrease with number of consecutive rainy days, or the relationship would be negative.

Current rainfall for April-August and for 1-15 Sept. were included as variables. April-August encompasses leaf expansion (Davis and Sparks, 1974) and fruit enlargement (Sparks, 1986), which is governed by soil moisture (Finch and Van Horn, 1936; Sparks, 1989a, 1995a). Current season's rainfall in September influences production, as this month is the major period of kernel development (Davis and Sparks, 1974). Nut weight is about 50% kernel and kernel growth is water dependent (Alben, 1958; Andrew and Sherman, 1980; Stein et al., 1989). The 1-15 Sept. interval was used because of being critical to kernel development (Sparks, 1992a). The relationship of production to rainfall in April-August could be negative or positive depending on whether rainfall is insufficient or excessive. The relationship of production to rainfall in the first two weeks of September should be positive because September is the second driest month of the growing season.

The intervals for previous year's rainfall were May-July and 1-15 Sept. Deficient (Stein et al., 1989) or excessive soil moisture (Hunter, 1963) in the previous year reduces production the following year. Observations indicate that a critical rainfall interval is May-July. The rationale for using 1-15 Sept. as a critical rainfall interval was that water stress during this time, simultaneously coupled with the stress of kernel development, would be expressed as reduced return bloom during the following year. The relationship of production to rainfall in May-July could be positive or negative depending on if rainfall is inadequate or excessive. As indicated previously, the relationship of production to rainfall in September should be positive due to the dry nature of September.

Heating degree-days and rainfall data were averages of conditions in Albany, Ga. and Cordele, Ga. These two locations are located within the major pecan production area of Georgia and their climates appear to be indicative of the pecan growing region (Sparks, 1989b).

The model, based on the relationship of production and the eight variables, to predict Georgia's production was developed by using established procedures (Ezekiel and Fox, 1959). The model delineates the effect of each individual variable on production and allows calculation of the combined effects of the eight variables on production.

Once the model was developed, the graphical relationship of nut production to each variable was calculated. Also, from the model, the relative importance of the individual variables on nut production was evaluated by calculating their individual percentage contribution to production. The graphical relationships illustrate the degree nut production can be affected by the variables in a given year, whereas, the percentage contribution shows the importance of the individual variables over years.

---

RESULTS AND DISCUSSION

The equation for modeling nut production is:

Production = - 454980 + 3260X1 - 0.5497X2 + 62X3 - 3676X4 - 3798(1/X5 ) + 10108X6 - 218X62 + 57388 X7 - 425X72 - 2828(1/X8)

Where:

X1 = production trend or year, 45-92 (yr)
X2 = previous year's nut production (lbs-millions)
X3 = previous year's heating, April - October (degree-days)
X4 = current year's rain, 2 or more consecutive rainy days, highest cumulative
sum occurring in May, June, or 1-15 July (days)
X5 = current year's cumulative rain, 1-15 Sept. (inches)
X6 = current year's cumulative rain, April - August (inches)
X7 = previous year's cumulative rain, May - July (inches)
X8 = previous year's cumulative rain, 1-15 Sept. (inches)

At first glance, the model appears complex. However, the expression for each variable (X1 - X8) is preceded by either a plus or a minus sign. Thus, the model is simply a summation of the positive and negative factors influencing production. The net summation is the predicted production for a given year. Nut production predicted from the model compared favorably with USDA estimates of Georgia pecan production (Fig. 2). The R2 is 0.92 (R2 = 1.0 would be perfect prediction). The close relationship indicates that the factors used to construct the model do determine Georgia's production. The close relationship also suggests the USDA estimates generally have good creditability.

During the interval, 1988-95, Georgia's pecan production fluctuated wildly (Table 2). In order for the model to have acceptability, estimates of production made by the model and estimates made by the USDA would have to have reasonable agreement. The agreement between the USDA and the model estimates is good. The close agreement confirms the creditability of the model. The one exception is the year 1995 for which the model predicted a production 20 million pounds higher than the USDA estimate. However, the model estimate for 1995 should be taken with reservation because production in 1995 followed a massive flood during the previous year. The model does not have a variable for measuring flood damage.

Beginning in 1988, the model consistently over predicted the USDA estimates. Furthermore, the amount of over prediction increased with time. Consistent and increasing over prediction of production from 1988 indicated that production had at least leveled off. Consequently, the model estimates of production (Table 2) were made with trend adjusted to 87 or the year 1987.

Relationship of nut production to production trend, previous year's production, previous year's heating degree-days, and days of consecutive rain in May, June, or 1-15 July is linear as illustrated by Figs. 1, 3 - 5. Relationship of nut production to previous and current season rainfall is curvilinear (Figs. 6 and 7). Obviously, extremes in previous year's production (Fig. 3), previous year's heating degree-days (Fig. 4), consecutive rainy days (Fig. 5), and rainfall during the previous and current year (Figs. 6 and 7) can have a pronounced influence on nut production in a given year.

The relative importance of the individual variables varied greatly (Table 3). By far, production trend had a greater effect on pecan production than any other factor. Production over the study period increased four fold (Fig. 1). Extrapolation of the diagonal line in Figure 1 to the year 2020 indicates that average production in Georgia would be 200 million pounds. Unless extensive plantings are begun in the very near future, a production of 200 million pounds appears to be unrealistic. Consequently, Georgia's production has apparently leveled off as previously indicated.

Previous year's nut production, the alternate bearing index, was the fourth most important factor. The less than predominant role of alternate bearing is contrary to general opinion, but is not surprising in that production can increase or decrease for 2 or more years without alternate bearing (Fig.1). In fact, classical alternate bearing occurred only for one prolonged period, 1960 -1976. Following this period, production increased for 2 years with record production occurring in 1978. Based strictly on alternate bearing, 1978 should have been an "off" year. Variation in production is not necessarily alternate as production is influenced by factors in addition to previous year's fruiting intensity. Therefore, irregular bearing is more descriptive of cyclic pecan production than is alternate bearing.

Previous year's rain for May - July was the most important climatic factor influencing pecan production (Table 3) and had an effect 6.5 times greater than current year's rain in April - August. Soil moisture in the previous year affects the number of nuts produced (Stein et al., 1989), whereas current year's moisture affects nut size (Finch and Van Horn, 1936) and kernel quality (Sparks et al., 1995). Current year's soil moisture has less effect on nut production because pecan production is dominated by nuts per tree and not by nut size (Sparks, 1992b). The curvilinear relationship of nut production to rain in the previous and current year (Fig. 6) shows that pecan nut production is sensitive to deficit as well as excessive soil moisture as proposed by Hunter (1963). Excessive soil moisture induces prolonged reduction in photosynthesis (Smith and Ager, 1988) and suppresses leaf growth (Smith and Bourne, 1989), consequently suppressing flower formation by limited carbohydrates (Lockwood and Sparks, 1978). Waterlogging has less effect on current year production than the dramatic suppression of next year production. The primary effect of waterlogging during the current year is due to inducing shuck decline (Sparks et al., 1995) and associated poor kernel development. Shuck decline usually occurs late in fruit development (Schaller et al., 1968) and production is reduced mainly because of suppressed kernel development. At the time shuck decline occurs, shell (which is about 50% of the nut weight) development is near completion. For this reason, suppressed kernel growth late in development does not result in a striking reduced production (Fig. 6). However, the effect on edible or marketable kernels can be pronounced (Sparks et al., 1995).

Within recent years, excessive rains have been a major problem in Georgia. The year 1991 was especially severe. Rains from April through August were record breaking. Scab was severe. The detrimental effect of scab on nut quality was further accentuated by excessive rains as such. Shuck decline was rampant. Kernel quality was very poor. Besides 1991, rainfall was excessive in 1989 and 1994. Following each of these wet years, return production was suppressed (Table 2). The 1992 crop was a 26-year low.

Consecutive rainy days in May, June, or 1-15 July, the index of scab infection, was the second most important climatic factor (Table 3). This is not surprising as the scab program recommended in the past (Ellis et al., 1991) did not adequately control scab in seasons of high scab pressure (Latham and Campbell, 1991; Sparks, 1995b). The decrease in production with consecutive rainy days (Fig. 5) supports the contention that scab occurs following rains that result in extended leaf wetness (Latham, 1982; Sparks, 1995b).

Previous year's degree-day accumulation was the third most important climatic factor (Table 3) and ranked with current year's rain, April-August in importance. Years with high production within the 1945-92 study period were associated with above average heating degree days. The relationship of production to heating degree-days supports the contention that low sunlight reduces next year's nut production (Hunter, 1963). Current season's heating degree-days were originally included in the model, but were not a determining factor in production. Because pistillate flower production is fixed at budbreak (Wetzstein and Sparks, 1986), any effect of current season heating would be on nut size and kernel development. As indicated earlier, nut size is a relatively minor component of total production (Sparks, 1992b). Nut size and kernel development are dominated by soil moisture (Finch and Van Horn, 1936; Sparks, 1989a), not heat variations.

Rain during the first 15 days of September of the current year was the fourth most dominant climatic factor (Table 3). Current seasons' rain had about 1.3 times the effect on nut production as the previous year's rainfall for 1-15 Sept. Rainfall the first 2 weeks of September can have a dramatic effect of production (Fig. 7). The two-week period is critical because it coincides with the beginning of rapid kernel development (Davis and Sparks, 1974) and because kernel development at this time is crucial dependent on soil water (Sparks, 1992a). The effect of soil moisture on kernel development is expected as September is historically the second driest month of the year and approximately one half of the total nut weight accumulates during this month. Effects of drought during 1-15 Sept. of the previous year are presumed due to suppressed carbohydrates accumulation from reduced leaf efficiency (Loustalot, 1945; Mielke, 1981; Rieger and Daniell, 1988) or, in extreme cases, to drought-induced defoliation.

The summation of the contribution (Table 3) for previous year's factors is 35.4. The sum for the current year's factors is 16.3. Previous year's factors have about 2.2 times more influence on current season production than do current season's factors. If the sums include only climatic factors, the contribution for previous season's factors is 28.8 and the current year's factors is 16.3 showing that previous year's climate has more effect on production than does current year's climate.

Although production trend was the dominant factor affecting production over time, it is relatively minor in a given year. The effect of trend is a constant within a year whereas the effect of the other factors affecting production has the potential to vary greatly (Figs. 3-7). Thus, production for a given year mainly depends on previous year's production and the climate of the previous and current year. The sum of the contribution (Table 3) for climatic factors (45.1) is about 7.0 times greater than the contribution for previous year's production (6.6), showing the predominant dependence of pecan production in a humid environment on variation of climatic factors. The climatic factors are mainly rainfall related. Sum of contribution for rainfall factors, including the scab index, is about 9.0 times greater than the contribution for heating degree-days.

Once the current growing season is completed (October 31), the model can be used to predict production for the following year under assumed conditions. Such a prediction would be valuable in making marketing decisions. For example, if the prospects are poor for a good return bloom, nuts from the current year can be placed in storage for next year's market. If prospects are high for a good return bloom, storing nuts would likely be an unprofitable decision. For instance, the prospects for a good return bloom in 1997 are high because of the favorable climate existing in 1996. If the climatic factors for 1997 are average (1945 to 1992), production in 1997 is predicted to be 103 million pounds. Thus, storing part of the 1996 crop would have been a risky decision. However, the final decision to store should be based on the expected total crop for the United States and Mexico.

The model can be used to predict production assuming adverse climatic conditions. If September is dry in 1997, predicted production is reduced from 103 to 67 million pounds. If the dry September is preceded by high scab pressure as occurred in 1991, predicted production is further reduced to 50 million pounds. On the other hand, production can increase from the predicted 103 million pounds if current year's climate is more favorable than the average climate for the 1945-92 period. Consequently, any production prediction based on the model presented here requires revision after completion of each of the three indices of current year's climate (Table 1).

Climatic factors influencing nut production manageable by the grower are scab from frequent rains, drought, and, indirectly, sunlight. The historical scab control program (Ellis et al., 1991) gives poor control during seasons of high scab pressure (Latham and Campbell, 1991; Sparks, 1995b). A new approach to scab control has been proposed (Sparks, 1995b) and, so far, the program appears to be working well. Whatever the preventive program employed, scab control can be greatly improved by minimizing mutual shading among trees.

Disastrous droughts occur mainly in late August through September. The 2.5 inches of rain per acre needed during 1-15 Sept. for maximum kernel development (Sparks, 1992a) is often deficit. During the 1945-92 study period, less than 2.5 inches of rain occurred during the critical period in about 60% of the years. Since the importance of water in September was reported (Sparks, 1992a), many growers have improved irrigation during early September with increased kernel quality and return bloom. Early September is an extremely critical period for kernel development. If only one irrigation was possible per year, it should be applied during 1-15 Sept.

The detrimental effect of excessive moisture from prolonged rains can be minimized by selecting pecan sites with adequate surface drainage. In this regard, return bloom following the excessive rains in 1994 was much better in orchards situated on slopes or ridges than in orchards on flat sites (Sparks, 1995c). The effect of surface drainage was especially apparent in the Albany vs. Fort Valley-Perry areas in 1995. Production was good in Fort Valley-Perry but poor in Albany. Although the excessive rains of 1994 flood began in the Fort Valley-Perry area, the water quickly drained off the ridges. Quick surface drainage was not the case in the Albany area.

Many Georgia pecan orchards are over crowded with a major reduction in photosynthesis from mutual shading (Andersen and Brodbeck, 1995; Mielke, 1981) and greatly reduced tree productivity (Hinrichs, 1961). The use of available sunlight and tree productivity can be improved by thinning crowded trees (Alben, 1958; Alben and Sitton, 1950; Crane et al., 1934; Romberg et al., 1959).

The climatic model predicts pecan nut production with acceptable precision. The major limitation of the model is that acreage in pecan production is assumed to be presently stable. If pecan acreage deviates substantially from stability, the model must be adjusted accordingly. Furthermore, the model will not adequately predict production following a major freeze. The same may be true following a major flood. However, these are not considered major flaws in the model, as the 1994 flood was an event last happening 100 years ago and the 1955 freeze 60 years ago. The model clearly defines factors influencing Georgia's pecan production and indicates areas for improved management. Additionally, the model can be used as a tool for planning marketing strategies.

Reliability of this model as well as USDA estimates are dependent upon knowing the commercial acreage of pecans in production. Pecan tree counts on a state wide basis should be updated every 5 years. In addition, a composite average of long term grower production records is needed to compare with the USDA estimates. Historical USDA estimates could be adjusted, where needed, to variations in actual production. The model could then be reevaluated based on adjusted USDA estimates. The net result should be an improved model.

---

¹Modification of a model published in the Journal of the American Society for Horticultural Science 121:908-914.

---

Literature Cited

Alben, A. O. 1958. Results of an irrigation experiment on Stuart pecan trees in east Texas in 1956. Proc. Southeastern Pecan Growers Assn. 51:61-68.

Alben, A. O. and B. G. Sitton. 1950. Fourteen years' results of thinning the stand of trees in a pecan orchard on alluvial soil. Proc. Amer. Soc. Hort. Sci. 56:98-102.

Andersen, P. C. and B. V. Brodbeck. 1995. Light preconditioning and fluctuating irradiance levels influence gas exchange of pecan leaves. p. 148-159. In: M. W. Smith, W. Reid, and B. W. Wood. (eds.). Sustaining pecan productivity into the 21st century. USDA, ARS 1955-3.

Andrews, C. P. and W. B. Sherman. 1980. Reduction in percent kernel as influenced by prolonged autumn drought. Proc. Southeastern Pecan Growers Assn. 73:159-161.

Bertrand, P. F. and T. R. Gottwald. 1984. The effect of deletion of late season fungicide sprays on pecan yield and quality. Proc. Southeastern Pecan Growers Assn. 77:51-54.

Crane, H. L., M. B. Hardy, N. H. Loomis, and F. N. Dodge. 1934. Growth and yield of pecan trees, as affected by thinning the stand of trees and other orchard practices. Proc. Amer. Soc. Hort. Sci. 32:33-37.

Davis, J. T. and D. Sparks. 1974. Assimilation and translocation patterns of carbon-14 in the shoot of fruiting pecan trees, Carya illinoensis Koch. J. Amer. Soc. Hort. Sci. 99:468-480.

Demaree, J. B. 1924. Pecan scab with special reference to sources of early spring infections. J. Agr. Res. 38:321-330.

Diener, U. L. and F. E. Garrett. 1967. Summary of three years results for pecan disease control in Alabama. Proc. Southeast Pecan Growers Assn. 60:36-43.

Ellis, H. C., J. Hadden, T. C. Crocker, and S. Brown. 1991. Georgia pecan spray guide. Univ. Georgia Coop. Ext. Ser. Bul. 841.

Ezekiel, M. and K. A. Fox. 1959. Methods of correlation and regression analysis. Wiley, NY.

Finch, A. H. and C. W. Van Horn. 1936. The physiology and control of pecan nut filling and maturity. Arizona Agr. Sta. Tech. Bul. 62.

Gammon, N., R. H. Sharpe, and R. G. Leighty. 1955. Relationship between depth to heavy-textured subsoil and drought injury to pecans. Soil Sci. Soc. Florida Proc. 15:31-34.

Gottwald, T. R. 1985. Influence of temperature, leaf wetness period, leaf age, and spore concentration on infection of pecan leaves by conidia of Cladosporium caryigenum. Phytopathology 75:190-194.

Gottwald, T. R. and P. F. Bertrand. 1982. Effect of time of inoculation with Cladosporium caryigenum on pecan scab development and nut quality. Phytopathology 73:714-718.

Gottwald, T. R. and B. W. Wood. 1984. The effects of scab and powdery mildew on photosynthesis and respiration of pecan fruit and foliage. Proc. Southeastern Pecan Growers Assn. 77:139-143.

Hinrichs, H. A. 1961. The relationship of native pecan tree spacing to yield. Oklahoma State Univ. Agr. Expt. Sta. Bul. B-574.

Hinrichs, H. A. 1962. Pecan industry and some research developments in Oklahoma. Annu. Rpt. Northern Nut Growers Assn. 53:80-85.

Hunter, J. H. 1963. Pecan-production cycles as related to weather in Georgia. Proc. Southeastern Pecan Growers Assn. 56:10-12.

Latham, A. J. 1982. Effects of some weather factors and Fusicladium effusum conidium dispersal on pecan scab occurrence. Phytopathology 72:1339-1345.

Latham, A. J. and H. L. Campbell. 1991. Evaluation of selected fungicides for control of scab and anthracnose of pecans. Proc. Southeastern Pecan Growers Assn. 84:78-83.

Latham, A. J. and F. E. Garrett. 1973. Control of pecan scab in Alabama during 1972. Proc. Southeastern Pecan Growers Assn. 66:149-152.

Lockwood, D. W. and D. Sparks. 1978. Translocation of 14C in 'Stuart' pecan in the spring following assimilation of 14CO2 during the previous growing season. J. Amer. Soc. Hort. Sci. 103:38-45.

Loustalot, A. J. 1945. Influence of soil moisture conditions on apparent photosynthesis and transpiration of pecan leaves. J. Agr. Res. 71:519-532.

Magness, J. H. 1955. Practicability of irrigating pecan orchards in the southeastern states. Proc. Southeastern Pecan Growers Assn. 48:23-29.

Mielke, E. A. 1981. Effect of stress on pecan photosynthesis. Western Pecan Conf. Proc. 15:47-69.

Moznette, G. F. 1934. Experiments in control of pecan black aphids under orchard conditions. Proc. Southeastern Pecan Growers Assn. 28:55-61.

Pena, J. G. 1995. Pecan/tree nut marketing season: an overview of production, prices and market collapse. p. 88-109. In: M. W. Smith, W. Reid, and B. W. Wood. (eds.). Sustaining pecan productivity into the 21st century. USDA, ARS 1955-3.

Reid, W. and R. D. Eikenbary. 1990. Developing low-input management strategies for native pecan orchards, p. 69-76. In: B. W. Wood and J. A. Payne (eds.). Pecan husbandry: challenges and opportunities. USDA, ARS - 96.

Rieger, M. and J. W. Daniell. 1988. Leaf water relations, soil-to-leaf resistance, and drought stress in pecan seedlings. J. Amer. Soc. Hort. Sci. 113:789-793.

Romberg, L. D., C. L. Smith, and H. L. Crane. 1959. Effect of irrigation and tree re-spacing (thinning) on pecan tree growth and nut production. Proc. Texas Pecan Growers Assn. 38:60-75.

Schaller, C. C., F. N. Dodge and G. E. KenKnight. 1968. Increased occurrence of shuck disease of pecan (Carya pecan). Plant Dis. Rpt. 52:189-190.

Smith, M. W. and P. L. Ager. 1988. Effects of soil flooding on leaf gas exchange of seedling pecan trees. HortScience 23:370-372.

Smith, M. W. and R. D. Bourne. 1989. Seasonal effects of flooding on greenhouse-grown seedling pecan trees. HortScience 24:81-83.

Snipes, L. E. 1995. Georgia agricultural facts. Georgia Agr. Stat. Serv., Athens

Sparks, D. 1980. Energy requirement for pecan, Carya illinoensis (Wang.) K. Koch, p. 327-333. In: D. Pimentel (ed.). CRC handbook of energy utilization in agriculture. CRC Press, Inc., Boca Raton, Fla.

Sparks, D. 1983. Alternate fruit bearing in nut trees. Annu. Rpt. Northern Nut Growers Assn. 74:197-230.

Sparks, D. 1986. Pecan, p. 323-339. In: S. P. Monselise (ed.). CRC handbook of fruit set and development. CRC Press Inc., Boca Raton. Fla.

Sparks, D. 1989a. Drought induces fruit abortion in pecan. HortScience 24:78-79.

Sparks, D. 1989b. Predicting nut maturity of the pecan from heat units. HortScience 24:454-455.

Sparks, D. 1992a. Pecan cultivars - the orchard's foundation. Pecan Production Innovations, Watkinsville, GA.

Sparks, D. 1992b. In pursuit of a better pecan cultivar. Fruit Var. J. 46:174-182.

Sparks, D. 1992c. Stress factors affecting the Georgia pecan crop in 1991 and fruit set in 1992. Annu. Rpt. Northern Nut Growers Assn. 83:75-62.

Sparks, D. 1995a. Nut sizing period in pecan and soil water. Annu. Rpt. Northern Nut Growers Assn. 86: in press.

Sparks, D. 1995b. A climatic approach to pecan scab control. HortTechnology 5:225- 230.

Sparks, D. 1995c. Pecan fruit set following the wet 1994 growing season. The Pecan Grower 5(1): 38.

Sparks, D. 1996. A climatic model for pecan production under humid conditions. J. Amer. Soc. Hort. Sci. 121:808-914.

Sparks, D. 1997. Pecan. In: T. J. Michailides (ed.). Compendium of temperate nut crop diseases. Amer. Path. Soc. (In press).

Sparks, D. and C. E. Brack. 1972. Return bloom and fruit set of pecan from leaf and fruit removal. HortScience. 7:131-132.

Sparks, D., W. Reid, I. E. Yates, M. W. Smith, and T. G. Stevenson. 1995. Fruiting stress induces shuck decline and premature germination in pecan. J. Amer. Soc. Hort. Sci. 120:43-53.

Stein, L. A., G. R. McEachern, and J. B. Storey. 1989. Summer and fall moisture stress and irrigation scheduling influence pecan growth and production. HortScience 24:607-611.

Valli, V. J. 1964. Weather conditions as related to pecan scab infection. Proc. Southeastern Pecan Growers Assn. 57:85-87.

Wetzstein, H. Y. and D. Sparks. 1986. Flowering in pecan, p. 217-255. In: J. Janick (ed.). Hort. Rev. AVI, Westport, Conn.

Worley, R. E. 1979. Fall defoliation date and seasonal carbohydrate concentration of pecan wood tissue. J. Amer. Soc. Hort. Sci. 104:195-199.

 

---

Table 1. Variables used for modeling pecan nut production.

 

Table 2. USDA vs. model estimates of Georgia pecan production (lbs-millions).

ªProduction without an adjustment for pecan nut casebearer damage. Damage from this insect was atypically a problem in Georgia in 1996. Loss from this insect in 1996 is estimated to have been 10 to 15%. The model estimate includes nuts left in orchards because of depressing prices.

Table 3. Importance of variables on pecan production.

---

Fig. 1. Pecan nut production in Georgia, 1945-92. The diagonal line depicts trend.

 

Fig. 2. Relationship of predicted vs. Georgia's nut production, 1945-92. R2=0.92.

 

Fig. 3. Relationship of current year pecan production to the previous year's production.

 

Fig. 4. Relationship of pecan production to previous year's heating degree days.

 

Fig. 5. Pecan production as a function of the highest sum of days with two or more consecutive rainy days in May, June, or 1-15 July.

 

Fig. 6. Graphical relationship of pecan nut production to previous year's rainfall in May - July and current year's rainfall in April - August. Rainfall is cumulative in each interval.

 

Fig. 7. Graphical relationship of pecan nut production to current year's and previous year's cumulative rainfall from 1-15 Sept.

---

Pecan Page