|
|
|
 |
F. M. Roka and R. E. Rouse Reprinted from: Proc. Fla. State Hort. Soc. 110:82-86. 1997 Abstract As a result of not having a more current source of information, citrus managers often base production expectations on the Savage yield tables developed 40 years ago. This paper presents the first analysis of citrus production data collected in southwest Florida for eight scion/rootstock combinations. The results reflect trees planted since 1984 in high density tree spacings. Hamlin production reached over 550 boxes and 3,000 pound solids per acre by year eight. 'Valencia' production also peaked during the eighth year with over 400 boxes and almost 2,800 pound solids per acre. Comparisons between Carrizo and Swingle rootstocks did not indicate significant yield differences. Introduction Forty years ago, Zach Savage (1960), an agricultural economist at the University of Florida, developed yield tables for seven categories of citrus-early, midseason, and late oranges, seedy grapefruit, seedless grapefruit, tangerines and Temple oranges. The Savage tables reflected the deep sand conditions of central Florida and tree densities of between 48 to 70 trees per acre. For citrus growers in central Florida, the Savage tables proved to be valuable benchmarks of production performance. Freeze events of the 1980s forced a significant relocation of citrus operations to southwest Florida (Charlotte, Collier, Glades, Hendry, and Lee Counties). Between 1984 and 1996, citrus acreage in southwest Florida increased 175%, from 65,000 acres in 1984 to over 179,000 acres in 1996 (FASS, 1984; 1996). Much of this increased acreage was planted on poorly drained soils that had previously supported cattle pastures and seasonal vegetable production. In addition, tree densities of the new groves in southwest Florida were planted at an average density of 150 trees per acre. Realizing that the Savage tables were not relevant for southwest Florida, growers requested through the Citrus Advisory Council at the UF/ IFAS Southwest Florida Research and Education Center that a new study be initiated to develop a set of yield tables to reflect typical growing conditions in southwest Florida. The objective of this study was to develop representative citrus yields by scion variety, rootstock and tree age that reflect high density planting and the growing conditions in southwest Florida. This paper reports the first analysis of citrus yield data from southwest Florida. Results include trees with up to ten harvest seasons. Materials and Methods In a manner similar to how Dr. Savage conducted his study, this report utilized yield record information from cooperating growers. Citrus growers provided annual production data from designated blocks in their groves. Blocks were selected by the following criteria:
Eight different scion variety/rootstock combinations were being tracked in this study. The combinations were 'Hamlin', 'Valencia', 'Rohde Red Valencia', and grapefruit scion varieties on Swingle citrumelo (Citrus paridisi x Poncirus trifoliata) and Carrizo citrage (C.sinensis x P. trifoliata) rootstocks. Table I provides a breakdown of acreage and block numbers representing each scion variety/rootstock combination. All acreage in the study was located either in Hendry or Collier Counties. Cooperators were asked to provide two sets of information. Initial data were collected when each block entered the study. This data summarized basic information about the block, including scion variety, rootstock, planting date, tree density, spacing and net planted acres. In addition, information about the predominate soil type, bed configuration, type of irrigation, and general fertilizer practices were provided. The second set of information detailed block production since planting date. This record provided total boxes harvested from the block and mean pound solids per box. Pound solids data were not available for every block during every production season. These data were available only if at least part of the block's fruit was sold for processing. If a crop was sold to the fresh market, pound solids information was not available. In addition to production information, tree mortality, number of resets, and hedging/topping practices were recorded annually. A spreadsheet was created for each block, compiling all the information provided by the respective cooperator and coded for anonymity. The yield information forms were provided to the cooperators for their respective production data. Data were grouped by tree age. Consequently, a mean yield from three-year-old trees included trees that were planted in any year between 1985 and 1992. In effect, the yield-age profile in this study averaged across seasonal climatic variations. Tree age was determined relative to a reference date. Setting the reference date was a topic of discussion among citrus horticulturists at the IFAS Centers in Immokalee and Lake Alfred. For this paper, the reference date for tree age was I May. Any trees planted between 1 May and 30 April of the following year were considered to be the same production age. Block information and production data were merged into a single spreadsheet where data could be sorted by variety and rootstock and analyzed by tree age. The current scope of the study includes 59 blocks, representing eight cooperators and almost 3,700 net tree acres of citrus. Mean tree density was 150 trees per acre. Yield results reported in this paper end with the 1996-97 production season. Production means by tree age were computed for boxes per acre, boxes per tree, pound solids per box, and pound solids per acre. For each block, boxes per acre were calculated by dividing the reported production by net tree acres. Dividing boxes per acre by tree density provided an estimate of boxes per tree. When pound solids information was available, similar calculations were made to determine pound solids per net tree acre. Insufficient information was available to report grapefruit yields. The thirteen blocks and 638 acres of grapefruit reflect a mixture of red and white seedless varieties (Table 1). It is important to reiterate that yield values reported in the tables represent long term means. The mean yield by tree age incorporates several production seasons. The more blocks used in the calculation of a mean, the more confidence one can have that the calculated number represents "typical" conditions in southwest Florida. Therefore, means were reported only if at least four blocks had records for a given tree age. Tables 2-5 report orange yield information by tree age. Table 2 presents yield means for 'Hamlin' and 'Valencia' oranges where yield information was pooled across rootstocks. 'Hamlin' production reached 550 boxes per acre by year eight. Yield projections for 'Hamlin' beyond year eight are as yet unclear. Average number of boxes declined in year nine but increased dramatically in year ten. 'Valencia' yields peaked during year seven at 416 boxes per acre. Preliminary data suggest that 'Valencia' yields tend to plateau around 400 boxes per acre. As more study blocks mature, yield trends will become more apparent. Pound solids per acre increased for both 'Hamlin' and 'Valencia' through year eight (Table 2). As was seen with harvested boxes, average pound solids per acre of 'Hamlin' dropped in the ninth year, but rebounded to over 3500 in the tenth year. Pound solids per acre of 'Valencia' peaked during the eighth year at 2,790 and declined to 2,409 during the tenth year. A difference-in-means test was performed to identify if tree-age yields were statistically different between 'Hamlin' and 'Valencia' varieties. The null hypothesis stated that no difference existed between 'Hamlin' and 'Valencia' yields. Mathematically, XiH - XiV =0, where XiH and Xiv, are the mean yields of 'Hamlin' and 'Valencia', respectively, during the ith year. The columns labeled "Difference" in Table 2 report the magnitude of difference between mean yields at each tree age and T-statistics are reported in parentheses. At the 95% confidence level, the critical value for a one-tailed test is around 1.8, implying that if the T-statistic is greater, mean yields between varieties are not the same. After the third year, 'Hamlin' produced significantly more boxes per acre than 'Valencia'. Comparing pounds solids per acre, 'Hamlin' yields were significantly more than 'Valencia' through year pear to produce an equal number of boxes per acre. seven. While pound solids from 'Hamlin' in the eighth and However, there was some evidence that during years five ninth year remained numerically greater than 'Valencia', the though seven Rohde's produced more pound solids per acre. difference was no longer statistically significant (i.e. failed to reject the null hypothesis). Rootstock effects were tested by calculating separate yield means for 'Hamlin' and 'Valencia' on both Carrizo and Swingle rootstocks. Table 3 presents mean boxes per acre by tree age for each scion variety/rootstock combination. Except for six and eight-year-old 'Valencia', Carrizo rootstocks yielded more boxes per acre than Swingle rootstock. However, the null hypothesis that there are no yield differences between Carrizo and Swingle rootstocks could not be statistically rejected. In other words, at the 95% confidence level, yields differences between Carrizo and Swingle rootstocks were not significantly different from zero. Table 4 presents similar information with respect to pound solids per acre. For the years where there were sufficient data, pounds solids per acre were generally more from Carrizo rootstock as compared to Swingle rootstocks. However, yield differences were not statistically different from zero. The one exception occurred on six year-old 'Valencia's where Swingle produced more pound solids than Carrizo. However, there could be unexplained anomalies because on five and seven-year-old 'Valencia', Carrizo outperformed Swingle. Limited yield information was available for 'Rohde Red Valencia'. Table 5 compared production from Rohde to standard 'Valencia'. Rohde and standard 'Valencia' varieties appear to produce an equal number of boxes per acre. However, there was some evidence that during years five through seven Rohde's produced more pound solids per acre. Future Work Currently, all study blocks are located in either Hendry or Collier Counties. As the study continues, it is planned to increase the number of study blocks to include citrus acreage in Charlotte, Glades, and perhaps Lee Counties. Annual collection of production data from the study blocks will continue. As the study blocks "mature", the yieldage profile will be extended. During the next five years, it is hoped that tree-age peak yields and typical production profiles can be characterized for each scion/rootstock combination. In particular, it is hoped that hedging practices be incorporated into the analysis to examine their effect on the tree-age yield profile. Finally, if the study is allowed to continue indefinitely into the future, it may be possible to place parameters on the expected life span of a citrus grove in southwest Florida. Literature Cited FASS. 1984 and 1996. Commercial Citrus Inventory. Florida Department of Agricultural and Consumer Services, Florida Agricultural Statistics Service, Orlando, Florida. Savage, Zach. 1960. Citrus Yields per Tree by Age. Economic Series 60-8, Food and Resource Economics Dept., University of Florida, Gainesville, FL.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
 |
|
|
|
||
