.97. Rootstocksand Mineral Nutrition of Citrus H. K. Wutscher Introduction Mineral nutrition of plants is a topic which has beendi~ussed s'ince the beginningof agriculture. Aristotte wrote treatises on it, and until the 15th and 16th century the consensuswas that whateverplants needto grow, it al camefrom the soil. Al- though Nicholasde Cusaand Van Helmont, after his famousexperiment with a willow cutting, had ideasthat the sourceof the materialsthat makeup a plant was not quite assimple, it wasthe work of Priesdy,Ingenhousz, and De Sausa.re,and the discoveryof photosynthesisin the 18th and early 19th century that put mineral nutrition in its proper persp..:tiveas only one facet in the metabolismof plants. The grandold men of plant physiology,Sachs and Pfeffer, worked out the basisof mineral nutrition aswe know it today in the secondhalf of the last century and gaveus an understandingof the constituentsof plant tissuesand the essentialityof someelements. In a more practi:al vein, their contemporaries,Boussingault, Liebig, Gilbert, and Lawes,showed the possibility of increagngcrop yields by applicationof mineral fertilizers. Furtherwork, mostly with solution culture, showedthe essentialityof a seriesof elementswhich plants needin only very small anounts, the so-calledmicro-elements. A greatamount of effort went into trying to find the combination of variouselements for optinum plant growth and resultedin many publicationsof little practicalvalue for 2 reasons.First, becausemuch of the wort<was donewithout the statisticaltools to separatereal from apparentdifferences, and secondb..:ause mineral uptake is a function of a mazeof subtlecross connections of environment,development, genetics, and other variouselements involved. Mineral nutrition ceasedto be a glamourtopic of research,although there was a brief revivalwhen isotopes became generally available. Mineral nutrition studiesof citrus and other tree crops havealmost becomesynonymous with leaf analysis. Following the developmentof instrumentspermitting the rapid analysisof largenumbers of samplesand the pioneeringwork of Lundegardh, a large body of knowledgehas beenbuilt up empirically correlatingthe levelsof nutrient elementsin the leavesto tree perform- ance,providing a more sensitiveand lessambiguous method than deficiency or toxicity symptomsto diagnosethe nutrient sta- tus of trees. The rangesgiven in tablesof nutrient standardsare usuallyfairly wide (2). Nevertheless,leaf levelslisted as "opti- mum" often cannot be maintainedbecause of local peculiarities,like irrigation water high in salt and variationsin soil and climate. It is easyto overestimatethe effect of mineral nutrition; miraclesare often expectedfrom fertiizer application. But the plant doesn'tnecessarily take up everythingthat's applied andas Smith (31) hasshown in a nitrogenfertilization trial, the difference in yield betweentrees starved for N and thosereceiving high levelsis often no more than 200/0. Rootstock Eff~ts Becausethe root systemis the part of the plant which absorbsmineral elements(with the exceptionof nutrients appliedas foliar sprays)it is only logical that rootstocksshould havesome influence on the compositionof the scion. Substituting a gen- etically more or le$ distinct root systemis boundto havean effect on the scion and many reports bearthis out (seeLiterature Cited); however,the influence is by no meansonesided. The scion also influencesthe sizeand composition of the root system (17). Basicallythe scion and the rootstock, becauseof their different geneticmake-ups, remain separate entities, but one can influencethe behaviorof the other within certain narrow limits. The bud union is not a major factor in nutrient differences(36). Hodgson(18, 19) and Shannonand Zaphrir (28) investigatedthese relationships using reciprocally grafted rough lemon and tri- foliate orangeplants, and plantswith 2 root systemsof the sameor both species.The scionsseemed to havea greaterinfluence on determiningplant sizethan the rootstock. Two root systemsgave no advantagein mineral uptake over one, but the rootstock specieshad distinct but different effects on the levelsof K, Ca and Fe in the leaves.Trifoliate orangeleaves were hi~er in K and lower in Cathan rough lemon leaves,regardless of rootstock,which seemsto indicate that the scion influencewas dominant in this case. Whenused as rootstock for rough lemon,trifoliate orangeimposed the pattern of lower Caand higher K on the scion. Rootstockand scion seemedto be equally effective in influencingthe Fe concentrationin the leaves,but the Fe concentrations reportedare excessivelyhigh, which castssome doubt on the m:curacyof the analyses.Two componentsdetermine the amount of an elementin the leaf; uptake by the roots and trandocation. The root only passeson materialsto the scion after its own re- quirementsare met. Analysesof plantswith deficiencies,particularly micro-elementdeficiencies, often show that while the above- groundparts arelow in someelement the roots still contain adequateor evensurprisingly high levelsof it. The trunk, of course, -98- is the site of ttanslocationand the effect of insertingan interstock of sufficient length should givesome indication on the relativeimportance of root uptake and translocation. Effectsof interstockson leaf composition of deciduousfruit trees havebeen reported (35,39). Table 1 shlPNsr~ative effectsof rootstocksand 45-cm long interstockson the leaf lev~s of 7 elementsin young grapefruittrees. The treeshad beengrown in containersfor 2 yearsbefore being planted in the field. Leaf Sllnpleswere taken from 4 two-tree plots of eachrootstock/interstock treatment 2 yearsafter planting. Analysisfor 12 ~ementsshowed no significantdifferences in P, Fe, In, Cu and Na with rootstock or interstock. In only a few instances wasit possibleto overridethe root influencewith an interstock. Treeswith Citrus macrophyllaroots and sour orangeand 'Cleopatra'mandarin interstocks had lower N levelsthan treeson C. macrophyJ/awithout an interstock. They bm aved much like treeson sour orangeand 'Cleopatra'roots. Treeswith EremocitrusgJauca hybrid interstocksand C. macrophylla roots accumulatedmore chloridesthan treesdirectly on C. macrophylla. Interstocks,with C. macrol!hylla as the common rootstock, affectedthe N concentrationin the leaves.With trifoliate interstock it washigher than with sour orange,'Cleo- patra', E~mocitrus glaucahybrid, and satsumainterstock. E. glaucahybrid interstock resultedin higher K lev~s than 'Changsha'mandarin interstock. Mn washigher with 'Savage'citrange and 'Changsha'mandarin than with E. glauca hybrid interstocks. Chlorideswere lower with sour orangeand 'Cleopatra'interstocks than with E. glaucahybrid interstock and lower with sour orangethan with 'Troyer' citrange. 'Cleopatra'mandarin interstock lowered B comparedto 'O.ri' Sat- sumointerstock. This is contrary to its behavioras a rootstock, where 'Cleopatra'is chloride-tolerantand B-sensitive(5). In spite of theseeffects of interstocks,root uptake and not tran~ocation appearsto be the dominant factor in determining leaf nutrient lev~s. There are severalreasons why eff~ts of rootstockson miner~ nutrition are important. They haveto be taken into ~. count when interpretingleaf analysisdata. Without a knowledgeof the nutritional idiosyncrasiesof a particular rootstock it is easyto misiudgethe nutritional statusof trees. The excessiveuptake of one elementcan set in motion one or more nutritional imbalancereactions, such as depressionof N by excessiveamounts of Ca (38). ExcessK depressesMg. High levelsof heavymetals can induce Fe deficiencysymptoms. At leastpart of the mechanismof rootstock influenceson fruit quality (30) is probably nutritional. If the rootstock is one of the sp~ies in the subtribal group Citrus there is some,but not too much, variation in leaf nutrient levelsbetween rootstocks, but \IIA1engratt-eompatible citrus relativesare used greater differencescan be expected. The ealtier mentionedinterstock datashowed that the interstockcausing most differencesin nutrient levels w. an E. glaucahybrid. The data in Table 2 showthatSeverinia can cause a rangeof unusualleaf nutrient patterns,among which accumulationof very high Mn levelsis the most striking feature. Chlorosisremains as a little understoodmineral nutrition pro~em, althou~ Smith et al. (33) haveshown that levelsof Fe areconsistently lower in chlorotic than greenleaves. But often only part of the leavesof a tree arechlorotic, or they are chlorotic only at certain times of the year. In Texaswe haveobselVed that asthe treesget older th~ seemto be lesschlorosis- prone. Nevertheless,rootstocks clearly influencethe tendencyof treesto showchlorosis, and this is often strikingly demon- stratedin grovescontaining trees on more than one rootstock. In the rootstock trial describedin Table2 the sanplescontained both greenand chlorotic leavesand the correlation Fe content-chlorosisis not very good. The hi~ chlorosisresistance of 'Cleopatra'mandarin rootstock is noteworthy becausethis rootstock is often thou~t of aschlorosis-prone.With the advent of Fechelates, chlorosis is no longer the seriousproblem it oncewas, but chelatesare expensiveand if other considerations permit, the selectionof a chlorosis.resistantrootstock may be the most reasonablesolution to the problem. s~t T 01erance One of the critical aspectsof differencesin mineraluptake with rootstocksis salt tolerance. Citrus is often grown in arid areaswhere the irrigation water containshigh levelsof salts. Strictly spe*ing "s~t tperar.:e" refersto only sulfatesand