HORTSCIENCE 47(8):

Transcription

1 HORTSCIENCE 47(8): Horticultural and Economic Impact of Rootstocks on Fresh-market Marsh Grapefruit William S. Castle 1 Universit of Florida, IFAS, Horticultural Sciences Department, Citrus Research and Education Center, 700 Eperiment Station Road, Lake Alfred, FL Additional inde ords Poncirus trifoliata Milam hbrid, C-35 citrange, Citrus paradisi, crop value, fruit sie distribution Abstract. A rootstock trial ith a Marsh grapefruit (Citrus paradisi Macf.) scion as established in the flatoods of the Florida east coast Indian River region in The trees ere planted in an Alfisol of the Pineda series. The trial consisted of trees on 16 rootstocks, primaril citranges [C. sinensis (L.) Osb. Poncirus trifoliata (L.) Raf.], citrumelos (C. paradisi P. trifoliata), mandarins (C. reticulata Blanco) and various hbrids, in three or si replicates of three-tree plots in a randomied complete block design. Tree groth and survival, ield, and juice qualit ere measured annuall or periodicall for 10 ears. In three seasons, hole-tree crops ere sied in the field. Using the fruit sie distribution data, crop value or income/tree as estimated. Tree height after 10 ears ranged from 1.7 to 4.1 m and survival as greater than 90% ith a fe eceptions. Mean cumulative ield as 1202 kg/tree and ranged from 649 (Hamlin + Fling Dragon trifoliate orange) to 1615 kg/tree for a hbrid of trifoliate orange Milam. At tree age 9 ears, mean soluble solids production as 3594 kgha L1 ith a 240% difference beteen the loest and highest value. There ere differences in the distribution of four commercial-sied categories based on analsis of individual sies and using PROC FREQ to eamine and compare hole-tree distributions. When the ield and fruit sie data ere combined for 3 ears and converted to income/tree using commercial Florida Freight On Board prices in November and March, the trees on a trifoliate orange Milam hbrid (1584) had the highest estimated income ($354 U.S., March data) folloed b Calamandarin ($321) and Norton citrange ($292). The loest income/tree as $112 (Hamlin + Fling Dragon trifoliate orange). When all data ere considered, the best matches to current groer interest in smaller sied trees, and high ield and fruit qualit, ere the hbrid 1584 and C-35 citrange rootstocks. The Florida grapefruit industr built its reputation largel on fresh fruit gron in the Indian River region along the east coast (Stover and Castle, 2002). The particular combination of soils, climate, and the use of sour orange (C. aurantium L.) rootstock resulted in the production of fruit ith high eternal and internal qualit. Trees on sour orange rootstock are ell knon for producing ecellent qualit fruit and for thriving ell in man challenging tpes of sites related to soils and drainage that eist in the Indian River region. Hoever, in recent ears, the importance of sour orange declined because of continuing problems ith citrus tristea virus (CTV), hich as eacerbated ith the arrival of the efficient insect vector, the bron citrus aphid, Tooptera citricida (Kirkald), in Groers need information about alternative rootstocks to replace sour orange. Attempts to discover other rootstocks ith qualities equal or sufficientl similar to those of sour orange have not been Received for publication 23 Apr Accepted for publication 11 June To hom reprint requests should be addressed; bcastle@ufl.edu. ver successful (Castle et al., 1992). Moreover, the long-term implication from introducing ne rootstocks must be carefull evaluated. For eample, Single citrumelo became a general purpose rootstock for the Florida industr leading to its use as a rootstock for grapefruit throughout the state (Castle et al., 1988; Division Plant Industr, Bureau Citrus Budood Registration, 2010; Stover and Castle, 2002). Hoever, it as eventuall learned that it as not ell suited to certain Alfisols and related site conditions here an argillic or cla horion as ithin 45 cm of the surface (Bauer et al., 2005, 2007; Castle and Bauer, 2005). Grapefruit remains a popular crop in Florida, but the industr has shifted from groing mainl Marsh, a hite cultivar, to red and deep red cultivars such as Ra Rub, Redblush, and Rio Red. Furthermore, the proportion of the crop used as fresh fruit has declined. Currentl there are 31,000 bearing ha of grapefruit in the United States of hich 19,500 (63%) are in Florida (Fla. Agric. Stat. Serv., 2010). Of the Florida-gron grapefruit, 73% ere processed in the season. Of the fresh-market grapefruit raised in Florida, 53% in the seasons ere eported to Japan. It is ell knon among Florida grapefruit groers and packers that Japanese buers impose high standards on both internal and eternal (fruit sie) qualit. Thus, there is a continuing need to identif and evaluate suitable rootstocks that are tolerant of CTV, thrive under a range of soil conditions, and induce good fruit qualit, especiall high soluble solids concentration and fruit sies. The objectives of this stud ere to evaluate a series of rootstocks for tree groth, fruit internal qualit, ield, hole-tree fruit sie distribution, and to conduct an economic interpretation of those combined data from a freshmarket perspective. Materials and Methods Plant material and propagation. The rootstocks ere mostl citranges plus a fe commercial rootstocks and somatic and other hbrids (Table 1). Sour orange as not included because of the risk of CTV. Nurser trees ere produced in a commercial nurser using rootstock seeds provided from certified local sources. Seedlings ere budded ith Marsh grapefruit scion, clone F57-4-X-E, for hich the budood source had been tested for viruses and viroids and no CTV, loporosis (X), or eocortis (E) as detected. Field trials. The commercial field site (lat #13.46$; long #58.83$; elevation, 7.3 m) as located est of Fort Pierce, FL, and had been planted ith citrus that as removed about 1 ear before planting this trial. The soil series at the site as mapped as Pineda, an Alfisol in the famil of Arenic Glossaqualfs. The surface 15 cm is dark bron sand underlain b 30 to 50 cm of a bronish ello argillic horion. At the south end of the trial area as Riviera, a closel related series, but ith the cla or argillic laer usuall closer to the surface than in soils of the Pineda series. The south end of the site as not as ell drained and as generall etter, hich negativel influenced tree performance. The site as tpical of the Indian River region in that drainage as provided b dredging perimeter ditches. Shallo furros ere ecavated in the planted area and the soil used to form double-ro beds on hich the trees ere planted. Such beds are tpicall 15 m ide and 0.5 to 1 m above the natural landscape. Trees on 16 rootstocks ere planted 4.6 m in the ro and 6.7 m beteen ros in Sept in a randomied complete block design ith si replicates of 14 rootstocks and three replicates of four rootstocks in three-tree plots in the tree ro. The different number of replicates as the result of space limitations and tree availabilit. The trees on a fe rootstocks ere not planted until Mar. or Sept The trial as irrigated ith a microsprinkler sstem, hich also provided cold protection. The trees ere cultured ith local standards for fresh fruit including regular fertiliation ith dr materials at the annual rate of 180 kgha 1 (Obrea and Morgan, 2008; Rogers et al., HORTSCIENCE VOL. 47(8) AUGUST

2 2012). The trees ere topped mechanicall ever 2 to 3 ears at 4 to 5 m height. At the time of this trial, citrus canker as just becoming a problem but did not affect the conduct of the trial and Huanglongbing had not et been discovered in Florida. Data collection and analsis. Tree height as measured in 1994, 1997, and 2001 along ith tree survival. Annual samples of 50 fruit ere collected from each plot near the time of harvest. Juice as etracted and analed ith standard commercial equipment at the Citrus Research and Education Center, Lake Alfred (Wardoski et al., 1995). Fruit ield as measured annuall ith a commercial harvesting cre b placing fruit in a commercial field container holding 385 kg and recording volume in boes, the standard commercial unit of ield hich hen filled ith fruit eighs 38.5 kg. Yield in the (tree age 8 ears), (9 ears), and (12 ears) seasons as measured b harvesting the entire crop on each tree and placing it on a portable packing line (Autoline ProSort II sier; Autoline, Inc., Reedl, CA) that measured the sie and calculated the eight of each fruit. It has the same cameras, fruit cups, eight sensors, softare, and logic boards as commercial packing lines. Fruit from each tree ere placed in a hopper and transferred in a single line to the camera for measurement. The softare accumulated the number of fruit in each of the defined fruit sies based on parameters set up b the user. Fruit sie is the maimum diameter of the fruit. Fruit sies ere separated based on 3-mm increments. The packing line and softare ere set to divide the fruit into a maimum of 14 sie categories ith some overlap in all the fruit sies. The fruit ere separated into the various sies to closel correspond to the midpoint of the overlap beteen official Florida Department of Citrus fruit sies. The sies for our stud ere based on the number of fruit packed into a standard commercial cardboard carton that holds 20 kg of fruit: small (48 or more fruit/carton), medium (40 and 48), large (32 and 36), and etra large (27 or feer fruit/ carton). For Florida eport fresh grapefruit, the desired sies are 32 or 36 (Japan) and 40 or 48 (Canada). Data analses ere according to the randomied complete block design using PROC GLM ith mean separation b the least significant difference test. Some variables ere eamined b simple linear correlation analsis at *P # 0.05*, **0.01, or *** Three ears of fruit sie data ( , , and seasons) ere analed for differences among rootstocks for each individual packing sie and the overall distributions. PROC GLM as used to compare differences among rootstocks and for each fruit sie class and ear using arcsine transformed percentages. Fruit sie distributions/ tree ere eamined using PROC FREQ in each of the 3 ears (Louada et al., 2008). Rootstock as a significant factor in each ear, but rather than eamining all possible pairs of distributions in each ear, Table 1. Height (n = 6) of Marsh grapefruit trees gron in the Florida Indian River region on various rootstocks and tree survival at age 10 ears. Tree ht (m) at age Rootstock 3 rs 6 rs 10 rs Survival (%) Calamandarin Citrus madurensis (Lour.) C. reticulata Blanco Cleopatra mandarin C. reticulata Ridge Pineapple seet orange trifoliate orange ( ), C. sinensis (L.) Osb. Poncirus trifoliata (L.) Raf. Murcott C. reticulata C. sinensis Minneola tangelo trifoliate orange [C. paradisi (Macf.) Duncan C. reticulata Danc ] P. trifoliata Sun Chu Sha mandarin,v C. reticulata Single citrumelo C. paradisi P. trifoliata Norton citrange Carrio citrange Trifoliate orange Milam (1584) P. trifoliata C. jambhiri Milam Benton citrange,v Morton citrange C-35 citrange Trifoliate orange Milam (1572) P. trifoliata C. jambhiri Milam Ridge Pineapple trifoliate orange ( ), Hamlin + Fling Dragon trifoliate orange Hamlin + S. disticha No data 0 C. sinensis Severinia disticha (Blanco) Sing. Trifoliate orange Ridge Pineapple ( ) No data 0 P. trifoliata C. sinensis Mean Least significant difference Trees ere planted Sept at m. Rootstocks are sorted b tree height at age 10 ears. The classification of this rootstock is unclear. See this eb site for additional information: < Planted Sept Onl nine trees in three replicates. v Planted Mar Table 2. Seasonal Marsh grapefruit Freight on Board prices (U.S. dollars) for fruit harvested in the Florida Indian River region. Week No. fruit/20-kg shipping carton Weighted ending Season avg 19 Nov No data No data No data Mean Mar No data No data No data Mean Data from the Florida Citrus Administrative Committee. < org/>. a representative season ( ) as selected. A fe comparisons illustrating differences among rootstocks ere conducted after a visual inspection of the data. The economic impact of fruit sie distribution as assessed b calculating crop value or 1008 HORTSCIENCE VOL. 47(8) AUGUST 2012

3 total income/tree using Freight On Board (FOB) price data collected b the Florida Citrus Administrative Committee (< for Indian River-gron hite grapefruit packed in commercial fruit sies in 20-kg cartons. Data from the first eek of March ere recommended as representative in the five seasons of through , but crop value as also calculated for November, a time in each season hen fruit prices for the larger sies ere higher (Table 2). The number of fruit of each sie of each rootstock and ear as converted to carton equivalents and multiplied b the five-season mean FOB price for a given fruit sie. Because the trial fruit sie data ere placed into sie groups rather than individual sies, some interpolation of the price data as necessar. The income/tree or crop value as the amount summed across all fruit sie groups and the three seasons. Income/tree data ere subjected to an analsis of variance ith mean separation b the least significant difference procedure. Results and Discussion Tree height and survival. After 3 ears, mean tree height as 2.2 m and ranged from 1.8 to 2.5 m (Table 1). The relationship among rootstocks remained virtuall the same as the trees aged so that after 10 ears, the tallest trees ere those on such rootstocks as Cleopatra mandarin and Murcott at 4.1 m. The smallest trees included those on C-35 citrange (2.9 m) and a Hamlin seet orange + Fling Dragon trifoliate orange somatic hbrid (2.2 m). Tree survival after 10 ears as 90% to 100% among the trees on most rootstocks ecept for the , , , and 1572 hbrids in hich man trees succumbed to Phtophthora foot rot. Those hbrids ere produced b Harr Ford (retired, Citrus Research and Education Center) for their tolerance to burroing and citrus nematodes. Similar problems ith Phtophthora ere encountered in other trials (Castle and Baldin, 2008). The trees on Hamlin + S. disticha gre off eakl and ere unthrift in appearance. Fruit ield. The trees began to bear a commercial crop in their second ear (Table 3). At that time, the mean ield as 32 kg/ tree ith the trees on C-35 and Morton citranges bearing greater than 50 kg/tree. No adjustments ere made for differences in tree age; thus, ield in the first ears as relativel lo among the trees on Benton citrange and Sun Chu Sha mandarin because the ere planted 6 months later. Hoever, their ields lagged behind for onl 1 ear among the trees on Benton citrange and for a longer period on Sun Chu Sha. The trees on and ere planted 12 months after the trial as established, but their ields also onl lagged no more than 1 ear. Annual ields increased steadil until the trees ere 7 to 8 ears old and then the trees on some rootstocks displaed a more evident tendenc toard alternate bearing. Peak ield Table 3. Yield (kg/tree; n = 6) of Marsh grapefruit trees on various rootstocks planted in Sept.1990 in the Florida Indian River region at m. Tree age (rs) Cumulative Rootstock ield Trifoliate orange Milam (1584) C-35 citrange Calamandarin Ridge Pineapple Milam ( ) Murcott Norton citrange Single citrumelo Cleopatra mandarin Morton citrange Benton citrange Minneola Trifoliate orange Carrio citrange Trifoliate orange Milam (1572) Sun Chu Sha mandarin Ridge Pineapple trifoliate orange ( ) Hamlin + Fling Dragon trifoliate orange Hamlin + S. disticha Trifoliate orange Ridge Pineapple ( ) Mean Least significant difference No data ere collected at tree age 10 ears. Data are sorted according to the 10-ear cumulative ield. N = 3; planted in Sept N = 3; planted in Mar Data collection as discontinued because of tree loss and generall poor groth. Data ere not included in the statistical analsis. Yield is the value for all surviving trees and as determined b converting ield measured b volume to eight using 38.5 kg/bo as the conversion factor. for the trees on most rootstocks ranged from 100 to greater than 250 kg/tree. In man ears, statisticall significant differences occurred onl among the highest and loest ielding trees. The trees ith highest 10-ear cumulative ield ere those on the trifoliate orange Milam hbrid, 1584 (1615 kg/tree), and C-35 citrange (1456 kg/tree). At the lo end of the range ere trees on such rootstocks as Carrio citrange (1088 kg/tree). The differences in cumulative ield among rootstocks ere on the order of 50% to 60%. Differences in ield ere not full eplained b differences in tree height because their simple linear correlation coefficient as onl r.05 = 0.54*. The strength of the relationship as disrupted b the higher ields of smaller trees like those on C-35 citrange. Fruit qualit. Florida grapefruit are used for the fresh market and processing. Standard qualit measurements determine the value of the processed juice as ell as maturit for harvesting fresh fruit (Fellers, 1990; Wardoski et al., 1995). Mean juice contents ere 63.6% and 50.6%, soluble solids concentrations 10.2 and 10.1, acid contents 1.12 and 1.16, and soluble solids acid ratios 9.2 and 8.8 for hen the trees ere oung (5 ears old) or older (9 ears), respectivel (Table 4). The mean data are tpical for Florida grapefruit and the ranges of each variable ere small ith no eceptional values. Hoever, hen those data ere combined ith ield to calculate soluble solids in kgha 1, the highest values (greater than 4000 kg) hen the trees ere 9 ears old ere generall 1.5- to threefold larger than those of the trees ith the loer values. Fruit sie. Fruit sie distribution data ere collected hen the trees ere 8, 9, and 12 ears old. In those ears, the mean number of fruit/tree as 386, 385, and 223, respectivel, a 58% difference beteen trees at ages 9 and 12 ears (Table 5). Those 3 ears ere the ear hen annual increases in ield stopped (8 ears) and 2 ears thereafter in hich the trees on some rootstocks displaed a tendenc to alternate bear. Trees on the highest ielding rootstocks produced greater than 450 fruit/tree (Calamandarin, Murcott, and 1584). The trees on 1584 ere highl productive in each ear, hereas those on man other rootstocks including Murcott and Calamandarin had substantiall loer numbers of fruit/tree at age 12 ears. Rootstock and ear affected the portion of fruit in each sie class but ithout an significant interactions in the 3 ears. Fruit measuring beteen 92 and 113 mm represented 60% of the total number of fruit/ tree (Table 5). Those sies are equivalent to Florida commercial sies 48 to 32 (number of fruit packed in a standard cardboard carton), respectivel. For the trees on most rootstocks at each age, the portion of fruit sie 48 and smaller as 10% to 20% ecept at tree age 8 ears hen the portion generall eceeded 30%. Crop sie as not consistentl related to the portion of fruit sie 48 or smaller. The trees on some rootstocks like Hamlin + Fling HORTSCIENCE VOL. 47(8) AUGUST

4 Table 4. Juice qualit (n = 6) of Marsh grapefruit from trees on various rootstocks planted in Sept in the Florida Indian River region. Dragon trifoliate orange had a relativel high percentage of small fruit ever ear despite variabilit in fruit number/tree. The trees on Benton citrange had essentiall the same number of fruit/tree hen the ere 8 and 9 ears old, but 34% and 9.4% sie 48 fruit in those 2 ears, respectivel. Those and other similar outcomes contradict the concept that fruit sie distribution shifts to smaller fruit as crop sie increases (Davis et al., 2004). Whole-tree fruit sie distributions rather than individual fruit sies ere compared using PROC FREQ to determine if the trees on certain rootstocks produced crops that tended toard higher proportions of larger or smaller fruit and the approimate magnitude of the difference needed to make that distinction. The outcomes indicated that there ere highl significant (P < 0.01) differences each ear among rootstocks. The Soluble solids conc. (%) Acid (%) Soluble solids (kgha 1 ) Rootstock Tree age (rs) Juice content (%) SSC/A ratio Calamandarin Murcott Trifoliate orange Milam (1584) C-35 citrange Minneola Trifoliate orange Norton citrange Sun Chu Sha mandarin Benton citrange Single citrumelo Cleopatra mandarin Carrio citrange Ridge Pineapple Milam ( ) Morton citrange Trifoliate orange Milam (1572) Ridge Pineapple Trifoliate orange ( ) Hamlin + Fling Dragon trifoliate orange Hamlin + S. disticha v Trifoliate orange Ridge pineapple ( ) v Mean Least significant difference Mean Least significant difference Fruit samples ere harvested in mid-feb and mid-mar hen the trees ere 5 and 9 ears old, respectivel. Rootstocks are ranked b soluble solids at tree age 9 ears. N = 3; planted Mar N = 3; planted Sept v No fruit ere sampled at tree age 9 ears. SSC = soluble solids concentration; A = acid. mean number of fruit/tree across the entire eperiment varied from ear to ear but appeared not to be a direct factor determining fruit sie distribution. The simple correlation coefficients of total number of fruit/ tree compared ith the portions of individual fruit sies ere not significant. Therefore, the data of the season ere used for selected pairise comparisons among rootstocks. The trees on Hamlin + Fling Dragon trifoliate orange usuall had one of the highest percentages of smaller fruit and that as most evident in the season (tree age 12 ears) hen the portion of the most desirable 32 and 36 sies as 20% (Fig. 1). The fruit distribution of the trees on the somatic hbrid, Hamlin + Fling Dragon trifoliate orange, as skeed toard smaller sies, hereas those of the trees on Cleopatra and Sun Chu Sha mandarins, Single citrumelo, Carrio citrange, and others tended toard larger fruit sies ith less than 20% small fruit. The fruit distribution of the trees on Hamlin + Fling Dragon trifoliate orange rootstocks as significantl different from that of each of the latter rootstocks in the season. There as no difference beteen the trees on Cleopatra and Sun Chu Sha mandarins, but each as different from those on Single citrumelo and C-35 citrange. Fruit distributions and economic value. Mean crop value or estimated income/tree summed over three seasons hen the trees ere 8, 9, and 12 ears old as $305 U.S. earl in the season (November) hen FOB prices ere higher and $249 in March hen the prices ere more tpical of the entire season (Table 6). Regardless of the time of ear, the crop on the trees on the 1584 hbrid rootstock had the highest value and those on the somatic Hamlin + Fling Dragon trifoliate orange ith their large portion of smaller fruit sies and loer ield had the loest crop value. Certainl the sustained high fruit number per tree and relativel high proportion of larger sied fruit contributed to the economic performance of the 1584 hbrid. Highl significant simple correlations of number of fruit/tree and income in each of the three seasons, r = 0.80 to 0.98, suggested that ield more than fruit sie distribution increased income. The Florida FOB data sho that the prices received per carton increase regularl and substantiall as sie advances from 56 per carton up to sie 32 at hich point the increment in value for further increases in fruit sie is smaller (Table 2). That relationship suggests that trees ielding high proportions of fruit sie 36 or larger ould be ideal. Hoever, large percentages of desirable fruit sies do not necessaril equate to large incomes hen ield is comparativel lo. For eample, the mean percentage of sie fruit as 52% for the trees on Carrio citrange and 38% for those on 1584 rootstock, but the trees on 1584 had 76% more fruit (Table 5). Those differences are reflected in the mean March crop value of $354 U.S. for the trees on 1584 vs. $210 U.S. for those on Carrio citrange (Table 6). Trees on other rootstocks like Norton and C-35 citranges had loer percentages of the sie 27 and sies but higher ields leading to higher incomes (Table 5). Another aspect of fruit sie distribution, rootstocks, and prices is the possible impact of seasonal price variabilit and fruit maturit. There are to questions: first, is there a time in a season hen the price of fruit sie 27 or larger is so great compared ith smaller and medium-sied fruit to favor rootstocks that produce relativel high proportions of larger fruit but ith relativel lo ield, thus leading to higher crop value? The Florida fresh-market grapefruit season runs from approimatel November through April. FOB prices var seasonall ith values being highest at the beginning of the season and 1010 HORTSCIENCE VOL. 47(8) AUGUST 2012

5 Table 5. Marsh grapefruit sie distribution (n = 6) among trees on selected rootstocks planted in Sept and groing in the Florida Indian River region. Fruit sie (mm) Less than 91 (48) ( ) ( ) Greater than 114 (27) No. fruit Percent No. fruit Percent No. fruit Percent No. fruit Percent Rootstock Tree age (rs) Total no. fruit/tree Benton citrange C-35 citrange Calamandarin Carrio citrange Cleopatra mandarin Hamlin + Fling Dragon trif Minneola trif Morton citrange Murcott Norton citrange P. trif. Milam (1584) Sun Chu Sha mandarin Single citrumelo Mean Least significant difference Mean Least significant difference Mean Least significant difference Measurement dates ere mid-mar. 1999, 2000, and mid-feb Rootstocks are sorted b name. The fruit sies, 91 mm or less, mm; mm and 114 mm or greater are equivalent to commercial sies of 48 or more, , and feer than 27 fruit, respectivel, packed in the Florida industr standard carton of 0.8 bushels b volume (28 L), hich holds 20 kg of fruit. N = 3; planted Mar primaril for the largest sied fruit, hich comprise onl a small portion of the industr crop (Table 1). In the comparison of crop values (estimated income/tree) earl in the season (November) hen prices are higher than more tpical prices (March), incomes for the trees on all rootstocks ere higher in November b 15% (Table 6). Therefore, as an eample, crop value increased for the trees on Carrio citrange, but its relative position among the rootstocks as virtuall unchanged indicating earl-season higher prices for large fruit did not compensate for its loer ield. The second question is, does rootstock influence fruit maturit and, if so, should the crop value of each rootstock be determined at its optimum harvest time? Rootstock affected fruit maturit based on soluble solids acid ratio (Table 4). Grapefruit soluble solids:acid ratio increases approimatel half a point/ eek so a 1-or 2-point difference beteen rootstocks equates to as much as 2 to 4 eeks difference in maturit. Hoever, fruit maturit, thus, harvesting, is determined b to regulated factors: minimum requirements for soluble solids and juice content according to fruit sie (Wardoski et al., 1995). Thereafter, fruit sie becomes a major factor. Fruit of sie 36 or larger are usuall spot-picked at the beginning of the season, then in Januar, the remaining fruit are harvested. Thus, earl maturit ould not provide an meaningful financial advantage. Fruit prices are high onl for the first fe eeks of an season and do not change markedl afterard ithin a sie class. Thus, the use of the more tpical March data for comparing rootstock crop values is reasonable. Rootstock assessments. The desired attributes of orchard sstems in Florida have graduall been shifting from idel spaced, large trees to smaller sied, more closel spaced trees (Castle et al., 2010b). Rootstock options have an important role in that shift as ell as contributing to other long-standing favorable horticultural traits such as larger fruit sie and higher juice qualit. Using those traits combined ith the estimated incomes to evaluate rootstock performance in this trial, the 1584 hbrid and Norton and C-35 citranges ould provide the highest incomes at closer spacing (800 to 1000 trees/ha). The trees on those rootstocks had among the highest income over the 3 ears presented, tree height as relativel small, survival rates ere high, the consistentl produced good fruit sies and juice qualit, HORTSCIENCE VOL. 47(8) AUGUST

6 Table 6. Estimated cumulative crop value (U.S. dollars) per tree (n = 6) of Marsh grapefruit gron in the Florida Indian River region on selected rootstocks hen the trees ere 8 ( season), 9 ( ), or 12 ( ) ears old. 3-r cumulative income Rootstock November March P. trif Milam (1584) Calamandarin Norton citrange Murcott Cleopatra mandarin Single citrumelo C-35 citrange Benton citrange Sun Chu Sha mandarin Morton citrange Minneola trif Carrio citrange Hamlin + Fling Dragon trif. Mean income Least significant difference Incomes ere estimated using Freight On Board prices from the Florida Citrus Administrative Committee and for the eeks ending approimatel 19 Nov. and 4 Mar. of the same season. The prices are the mean of the 5 ears, N = 3; planted Mar Fig. 1. Whole-tree distribution of Marsh grapefruit in four standard commercial fruit sie groups (number of fruit packed in a 20-kg carton) according to rootstock and ear: season (upper panel); (middle panel); season (loer panel). For those three seasons, the mean fruit count/tree as 386, 385, and 223 fruit/tree and tree ages ere 8, 9, and 12 ears, respectivel. The percentage distributions are based on measuring the entire crop on each of nine (three replicates) or 18 trees (si replicates)/rootstock (see Table 1 footnotes). and soluble solids production as relativel high. In addition, fruit shape, although not measured, tended to be more oblate (desired) ith thin peel for those rootstocks as observed at harvest time. Among those rootstocks, 1584 as hbridied b Harr Ford (CREC, retired) and has et to attract commercial attention (O Bannon and Ford, 1977). It is a poor seed producer. Also, juice qualit as not above average in this trial, but as adequate to meet commercial standards. Thus, ield as the critical factor along ith fruit sie and shape. Norton citrange has been tested in a number of field trials (Castle, 1987; Wutchser, 1979) but also has not achieved commercial status perhaps because of mied evaluation results. C-35 citrange as released in 1986 (Cameron and Soost, 1986) and has steadil gron in commercial interest including in Florida because of reliable performance and its tolerance to nematodes and Phtophthora (Castle et al., 2000, 2010a; McCo et al., 2004). Juice qualit among the trees on C-35 as better than for those on 1584, but ield as loer. The trees on Calamandarin and Murcott also had high estimated incomes and shared man of the same attributes ith those on other rootstocks. Their high incomes can be attributed to good ield in the middle ears of the trial, but their productivit appeared to be declining after the reached 9 ears old. That decline ma have resulted from the inabilit of vigorous trees on the particular soil of the trial site to survive as long as those on less vigorous rootstocks. Also, the trees on Calamandarin and Murcott ere larger and, thus, ould be more suited to more idel spaced orchard designs (250 to 450 trees/ha). Calamandarin as introduced b W.P. Bitters (UC, Riverside; deceased) and has been tested elsehere in Florida here its traits ere as reported herein (Castle et al., 2010a). Murcott is another rootstock that has been tested in field trials partl because of potential benefits from its presumed seet orange mandarin parentage and its production of ver uniform seedling populations (Castle, 1987; Castle et al., 2010a; Wutchser 1979). The remaining commercial rootstocks include Single citrumelo, the most popular commercial rootstock in Florida, one often used for grapefruit, along ith Cleopatra and Sun Chu Sha mandarins and Carrio citrange. The intermediate crop value of the trees on Single citrumelo as accompanied b a generall intermediate ranking of their overall horticultural performance. Those results coincide ith the general groer and field trial eperience ith trees on Single citrumelo in the Indian River area recogniing that there are limitations to its use in Florida under certain site conditions (Bauer et al., 2005; Castle et al., 1988, 2000, 2010a, 2010b; Youtse and Lee, 1995). The trees on the mandarin rootstocks, although having a reputation for producing smaller fruit (Castle, 1987; Castle et al., 1993, 2006; McCollum et al., 2002), had intermediate ields ith high percentages of larger sied fruit. Those characteristics alone might justif the decades-long commercial use of Cleopatra mandarin as a rootstock for a modest proportion of Florida citrus 1012 HORTSCIENCE VOL. 47(8) AUGUST 2012

7 plantings. For grapefruit, Cleopatra mandarin ould not be a suitable choice for a modern orchard design primaril because of the large tree sie; hoever, it does have a special purpose use because fruit from trees on Cleopatra mandarin usuall have high soluble solids concentration and juice acidit leading to later maturit (Castle, 1987; Wutchser, 1979). Mandarin rootstocks are used commerciall to complement trees on other rootstocks to lengthen the harvest season. In contrast to man of the rootstocks described, the trees on the somatic hbrid, Hamlin + Fling Dragon trifoliate orange, seemed to be the most promising for higher densit orchards because of ver small tree sie, ecellent juice qualit, and reasonable crops for the tree sie. Hoever, the consistentl produced mostl small fruit and relativel lo ield leading to a lo crop value. Absent from this discussion is a direct comparison of the results to those from trees on sour orange rootstock. Hoever, in another Florida trial ith Marsh grapefruit, the trees on several rootstocks including C-35 citrange had higher juice qualit than those on sour orange and comparable fruit sie based on the mean eight of a sample (Castle et al., 2011). Conclusions There ere significant differences among various standard horticultural traits of the trees and fruit of Marsh grapefruit trees gron on various rootstocks in an Alfisol of the coastal Florida flatoods. When the horticultural outcomes ere combined to estimate the crop value, or income/tree, of the fruit sold in the fresh market, crop value summed for three seasons ranged from $354 U.S. to $112 U.S./tree based on March FOB prices. That range in estimated income as primaril the result of differences in ield, not differences in fruit sie distribution. Some rootstocks tended to produce crops skeed toard smaller or larger fruit sies, but those differences ere not apparentl sufficient to compensate for differences in ield. Based on the current groer interest in smaller sied trees, high ield, and ecellent fruit qualit, the most promising, highest income rootstocks ere those on a ne hbrid of trifoliate orange and Milam (1584) and C-35 citrange. Literature Cited Bauer, M., W.S. Castle, B.J. Boman, and T.A. Obrea Economic longevit of citrus trees on Single citrumelo rootstock and their suitabilit for soils of the Indian River region. Proc. Fla. State Hort. Soc. 118: Bauer, M., W.S. Castle, B.J. Boman, and T.A. Obrea Field guide to soil identification for Florida s citrus-groing regions. Univ. Fla. Coop. Etension Publ. SP-362. Cameron, J.W. and R.K. Soost C35 and C32 : Citrange rootstocks for citrus. Hort- Science 21: Castle, W.S Citrus rootstocks, p In: Rom, R.C. and R.F. Carlson (eds.). Rootstocks for fruit crops. Wile, Hoboken, NJ. Castle, W.S. and J.C. Baldin Rootstocks affect the 17-ear survival and performance of Valencia trees gron in Immokalee. Proc. Fla. State Hort. Soc. 121: Castle, W.S., J.C. Baldin, and J.W. Grosser Performance of Washington navel orange trees in rootstock trials located in Lake and St. Lucie counties. Proc. Fla. State Hort. Soc. 113: Castle, W.S., J.C. Baldin, and R.P. Muraro. 2010a. Rootstocks and the performance and economic returns of Hamlin seet orange trees. HortScience 45: Castle, W.S., J.C. Baldin, R.P. Muraro, and R. Littell. 2010b. Performance of Valencia seet orange trees on 12 rootstocks at to locations and an economic interpretation as a basis for rootstock selection. HortScience 45: Castle, W.S. and M.G. Bauer The 10-ear performance and survival of Marsh grapefruit trees on Sun Chu Sha mandarin and various citrumelo rootstocks on Riviera sand, depressional, an Alfisol. Proc. Fla. State Hort. Soc. 118: Castle, W.S., K.D. Boman, J.C. Baldin, J.W. Grosser, and F.G. Gmitter, Jr Rootstocks affect tree groth, ield, and juice qualit of Marsh grapefruit. HortScience 46:1 8. Castle, W.S., K.D. Boman, J.H. Graham, Jr., and D.P.H. Tucker Florida citrus rootstock selection guide. Univ. Fla. Coop. Etension Publ. SP-248. Castle, W.S., R.R. Pelosi, C.O. Youtse, F.G. Gmitter, Jr., R.F. Lee, C.A. Poell, and X. Hu Rootstocks similar to sour orange for Florida citrus trees. Proc. Fla. State Hort. Soc. 105: Castle, W.S., D.P.H. Tucker, A.H. Kredorn, and C.O. Youtse Rootstocks for Florida. 2nd Ed. Univ. Fla. Coop. Et. Publ. SP-42. Castle, W.S., H.K. Wutscher, C.O. Youtse, and R.R. Pelosi Citrumelos as rootstocks for Florida citrus. Proc. Fla. State Hort. Soc. 101: Davis, K., E. Stover, and F. Wirth Economics of fruit thinning: A revie focusing on apple and citrus. HortTechnolog 14: Division Plant Industr, Bureau Citrus Budood Registration Annual report. Fla. Dept. Agr. Consumer Serv., Tallahassee, FL. Fellers, P.F Florida s juice standards for grades and their differences from United States standards for grades and United States Food and Drug Administration standards of identit. Proc. Fla. State Hort. Soc. 103: Fla. Agr. Stat. Serv Florida citrus statistics Fla. Dept. Agric. Consumer Serv. and U.S.D.A. Agr. Marketing Serv. Louada, E.S., H.S. del Rio, M. Setamou, J.W. Watson, and D.M. Sietlik Evaluation of citrus rootstocks for the high ph, calcareous soils of South Teas. Euphtica 164: McCollum, T.G., K.D. Boman, and W.S. Castle Effects of rootstock on fruit qualit and postharvest behavior of Marsh grapefruit. Proc. Fla. State Hort. Soc. 115: McCo, C.W., W.S. Castle, J.H. Graham, J.P. Svertsen, and R.J. Stuart Pesticide suppression of Diaprepes abbreviatus (L.) (Coleoptera: Curculionidae) promoted differential groth and survival of Hamlin orange trees budded to five rootstocks in a Phtophthora infested grove. Proc. Fla. State Hort. Soc. 117: O Bannon, J.H. and H.W. Ford Resistance in citrus rootstocks to Radopholus similis and Tlenchulus semipenetrans (Nematoda). Proc. Int l Soc. Citricult. 2: Obrea, T.A. and K.T. Morgan (eds.) Nutrition of Florida citrus trees. 2 nd Ed. Univ. Fla. Coop. Et. Publ. SL-253. Rogers, M.E., M.M. Dedne, and T.M. Spann (eds.) Florida citrus pest management guide. Univ. Fla. Cop. Et. Publ. SP-43. Stover, E. and W. Castle Citrus rootstock usage, characteristics, and selection in the Florida Indian River region. HortTechnolog 12: Wardoski, W., J. Whigham, W. Grierson and J. Soule Qualit tests for Florida citrus. Univ. Fl. Coop. Etension Bull. SP-99. Wutchser, H.K Citrus rootstocks. Hort. Rev. 1: Youtse, C.O. and O. Lee A 6-ear comparison beteen 16 rootstocks budded ith Hamlin seet orange. Proc. Fla. State Hort. Soc. 108: HORTSCIENCE VOL. 47(8) AUGUST