CROP PRODUCTION

HORTSCIENCE 29(9):1008–1015. 1994.

dohb

hytc

li-l,e.l’s

ire-odlyfee

,o-

sdasyer-

e

s

-h,”

,

Microclimate and Columnar AppleTree Performance within Insect-exclusionary CagesD.S. LawsonDepartment of Entomology, New York State Agricultural Experiment StaGeneva, NY 14456-0462

S.K. BrownDepartment of Horticultural Science, New York State Agricultural ExperimStation, Geneva, NY 14456-0462

J.P. Nyrop and W.H. ReissigDepartment of Entomology, New York State Agricultural Experiment StaGeneva, NY 14456-0462

Additional index words. microenvironment, Malus domestica, protected cultivation, meshcages, arthropods

Abstract. A barrier system for pest control consisting of insect-exclusionary cages coverewith three types of mesh material was placed over columnar apple (Malus domesticaBorkh.) trees. This system has been shown to provide arthropod control equivalent tinsecticides. Light intensity, evaporation, and air and soil temperature were reduceinside the cages. Shoot elongation of columnar apple trees grown inside insect-exclusiary cages was significantly greater than that of trees grown outside the cages. However, tincreased shoot growth was not due to etiolation. Tree performance was unaffected insect-exclusionary cages. Fruit set and fruit soluble solids concentration were not reduceby the cages; however, fruit color intensity was reduced as the degree of shading from tmesh increased. These findings, in conjunction with high levels of arthropod control binsect-exclusionary cages, may allow insect-exclusionary cages to be used for evaluaintegrated pest management thresholds, predator–prey relationships, and apple prodution without insecticides.

.

otda

e

)is

se-

taicerrptl

rieehhee

This study reports on microenvironmenchanges and tree performance in a physbarrier system (insect-exclusionary cagplaced around apple trees for pest contUntil recently, the use of insect-exclusionacages to reduce arthropod colonization of aptrees has not been feasible due to large size. However, the availability of a new apptree form provided the impetus to test a barpest control system. Columnar-form apple trhaving compact growth habits characterizby reduced internode length, limited brancing, numerous fruit spurs, and stiff upriggrowth (Fig. 1) were used in this study (Fish1970; Looney and Lane, 1984). These tr

1008

d.es

dsh

h

d-

,

Received for publication 20 Aug. 1993. Acceptefor publication 20 Apr. 1994. We thank David Terryand Stephen Valerio for their technical assistancIan Merwin for his critical review of this manu-script, John Barnard for his assistance in data anasis, and Alan Lakso for his insightful suggestionabout methods for measuring changes in microcmate and tree performance inside insect-exclusioary cages. The use of trade names in this publicatdoes not imply endorsement by Cornell Univ. of thproducts named, nor criticism of similar ones nomentioned. Partial funding for this research is frothe U.S. Dept. of Agriculture regional integratemanagement of apple pests in the northeast. The cof publishing this paper was defrayed in part by thpayment of page charges. Under postal regulatiothis paper therefore must be hereby marked adver-tisement solely to indicate this fact.

allowed insect-exclusionary cages to be dveloped on a much smaller scale than threquired for dwarf or standard trees. Inseexclusionary cages made from metal framcovered with various mesh materials and placaround columnar trees have significantly rduced damage from direct and indirect apppests (Lawson et al., 1994).

A pest-control system with low-chemicainput is desirable for apple production anresearch. Apples are one of the most heavsprayed agricultural commodities producedthe United States because more than 500 scies of arthropods feed on apple (Croft, 197Reissig et al., 1984). Any system that careduce pesticide use from apples while smaintaining fruit quality will provide newinsight into low-chemical input productionFrom a research standpoint, the feasibility studying naturally occurring predators anparasitoids of phytophagous apple pests inorchard setting has usually been hamperedthe use of insecticides to control direct fruifeeding insects (Collyer, 1958; Croft, 1978Herbert and Sanford, 1969; Sanford anHerbert, 1967). A system that uses nonchemimeans to exclude all direct fruit-feeding insects, without excluding foliage-feeding opredatory arthropods, could foster biologiccontrol and allow the production of applewithout the use of insecticides. As agriculturproduction moves away from exclusive rel

lal

s)ol.yle

reeeeresd-tr,es

d

e,

ly-sli-n-

ionet

mdoste

ns,

H

tion,

ent

tion,

d

o

n-isyde

ing-

e-at

ct-esede-le

ldily

inpe-8;n

till

.ofd an byt-;d

cal-r

alsali-

ance on pesticides toward an increasing reance on beneficial organisms for pest controsystems that offer opportunities to examinpredator–prey relationships will be valuable

Columnar trees with resistance to severadiseases are being developed at Cornell Univ.New York State Agricultural Experiment Sta-tion in Geneva. Because these trees requminimal pruning, are resistant to several diseases, and can be protected from arthropdamage by insect-exclusionary cages, a trulow-chemical input and sustainable method oapple production could be developed for usby amateur orchardists and in small-scalnonchemical fruit production systems.

The use of insect-exclusionary cageswhether for research purposes or apple prduction, will require adequate tree performancewithin the enclosure. Insect-exclusionary cagethat adversely affect tree performance woulnot be acceptable. Our research objective wto determine the effect of insect-exclusionarcages on the microclimate surrounding thtrees and the resultant effect on tree perfomance.

Materials and Methods

Two insect-exclusionary cage designs werevaluated. Large cages with arched tops wereconstructed of 9.5-mm-diameter steel rodand were 4.5 m long × 0.91 m wide at the base× 1.8 m high (Fig. 2). There were six treatments. Three consisted of the following mestypes placed over a cage: 1) fine (“No-seeumwhite polyester material, pore size 1.0 × 0.13mm; The Baslon Hercules Group, ProvidenceR.I.); 2) medium (“Lumite” style 18 × 14amber, pore size 1.0 × 1.5 mm; Chicopee,Gainesville, Ga.); 3) large (“DUR-30”, whitepolyester, pore size 3.0 × 3.0 mm; Apex MillsCorp., Inwood, N.Y.), fully enclosing the treesThe fourth treatment consisted of the mediummesh with the top open (medium-O). In theremaining two treatments, the trees were ncovered, one group receiving insecticides anthe other none. Cages were arranged in randomized complete-block design with tworeplicates of each treatment in each of threblocks.

‘Starkspur Compact Mac’/MM.106 appletrees (Stark Bros. Nursery, Louisiana, Miss.that were not disease resistant were used in thstudy because adequate quantities of disearesistant columnar trees could not be obtaineThese trees were planted 0.91 m apart in thnorth–south direction in staggered double rowin each of the three blocks on 18 Apr. 1991.The cages and mesh coverings were installeover the trees before budbreak, and the mecoverings were removed after leaf fall during1991 and 1992.

In Fall 1990, the field for this experimentwas plowed, prepared, and fumigated witmethylisothiocyanate (Vorlex; Nor-AmChemical Co., Wilmington, Del.) (258 kg a.i./ha). Soil samples were taken in Mar. 1991 anrequired nutrient applications were incorporated by rototilling. Before planting, the treeswere dipped in a 2.5% solution of petroleumoil (Sunspray 6E; Sun Oil and Refining Co.

ORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

es,ners

l4forn

ds

.)ayrs

.,org inlin),es,tetr

e-le.e

tlyte-ric totil-91

by

adergeand

ree.on-a-n-

ithnotat-ks,chat-riverge

eertyeur-en10hech

ofra.ngsas

ge

t.,n

ec

h

e)

e

Fig. 1. (left) Standard-form apple tree compared to (right ) columnar-form apple tree.

Fig. 2. Large insect-exclusionary cages placed over five columnar apple trees.

Marcus Hook, Pa.) to control overwinterinpests such as mites, aphids, and leafhoppOn 19 Apr., soil around the newly plantetrees was treated with 0.95 liters of dilumetalaxyl (Ridomil 2E; Ciba-Geigy CorpGreensboro, N.C.) (0.63 ml a.i./tree) to cotrol diseases caused by Phytophthora andPythium spp.

Additional chemicals were applied to trein the large cages with a hand-operated bapack sprayer delivering ≈0.11 liters of dilutematerial per tree. Chemicals used were eit

HORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

rs.de

-

sk-

er

nondisruptive to beneficial arthropods or weselective for certain arthropod pests. On May, 21 May, and 4 June 1991 and 6 May,May, and 28 May 1992, all trees receivapplications of myclobutanil [Nova (40% WRohm and Haas Co., Philadelphia] (0.007a.i./tree) for control of apple scab [Venturiainaequalis (Cooke)], powdery mildew[Podosphaera leucotricha (Ell. & Ev.)], andcedar apple rust (Gymnosporangium juniperi-virginianae Schwein). Treatment six receiveseason-long insecticide applications. Th

included Bacillus thuringiensis (Javalin WG;Sandoz Crop Protection Corp., Des PlainIll.) (0.012 g a.i./tree) 20 May 1991 and 4 Ju1992 for control of obliquebanded leafrolle[Choristoneura rosaceana (Harris)]; chlor-pyrifos [Losban (50% W); Dow ChemicaCo., Midland, Mich] (0.08 g a.i./tree) on June and 31 July 1991 and 1 July 1992 control of obliquebanded leafrollers, and aaphid complex composed of apple aphi(Aphis pomi De Geer) and spirea aphids (A.spiraecola Patch); oxamyl (Vydate L; E.I. duPont de Nemours & Co., Wilmington, Del(0.026 ml a.i./tree) on 1 July 1991 and 14 M1992 for control of spotted tentiform leafmine(Phyllonorycter blancardella Fabricius); andendosulfan [thiodan (50% W); FMC CorpPhiladelphia] (0.07 g a.i./tree) 6 Aug. 1992 fcontrol of apple and spirea aphids. Durin1991, herbicide was used to control weedsand around the cages. On 20 May, oryza(Surflan; Elanco Products Co., Indianapolis0.34 liters a.i./ha was sprayed around the cagand on 4 June a 25% solution of glyphosa(Roundup; Monsanto Agricultural Co., SainLouis) was applied with a hand-held rolledirectly to weeds inside the cages.

In 1992, the large-mesh material was rplaced because it was not ultraviolet stabThe replacement material (“PC-33,” whitpolyester, pore size 3.0 × 3.0 mm; Apex MillsCorp.) had the same pore size but slighlarger thread diameter than the original marial. In that same year, landscaping fab(Belton Industries, Belton, S.C.) was usedprevent weed growth inside the cages. Ferizer requirements were assessed from 19leaf samples, and fertilizer was applied broadcasting to all blocks.

Small cages and mesh coverings were mfrom the same materials described for the lacages. The small cages were rectangular measured 40.5 cm × 40.5 cm × 2.1 m high (Fig.3). Each cage housed one potted apple tSeven treatments were examined: three csisted of fine-, medium-, and large-mesh mterial that fully enclosed the trees; three cosisted of fine, medium, and large mesh wopen tops; and the control treatment was caged or sprayed with insecticides. The trements were randomized in each of six blocin a randomized complete-block design. Eablock contained one replicate of each trement. The caged trees were placed in the drows of a source orchard that harbored a lapopulation of arthropod pests.

‘Starkspur Compact Mac’/MM.106 appltrees were potted on 15 May 1990 in 19-litplastic pots containing a mixture of 50% silloam and 50% Cornell artificial mix. Thesuncaged, potted trees were placed outside ding the 1990 growing season and were thoverwintered in a cool greenhouse. On Mar. 1991, the potted trees bloomed in tgreenhouse and were hand-pollinated. Eatree was fertilized on 10 Mar. 1991 with 75 g 14–14–14 Osmocote fertilizer (Grace-SierHorticultural Product Co., Milpitas, Calif.)The potted trees, cages, and mesh coveriwere placed in the source orchard when it wat the full-bloom stage on 10 May 1991.

re1414d, g

dse

1009

CROP PRODUCTION

l

c

nyal

ei,

t

nl

e

t

ca

or

e 13ekpi-ly,20ct-es

i-gtkne

ed

tsresdleschger

to-n-

ser-n,-

n-)

he-estsr

ngn-2.asee.sd

htheh,od

-),,

ren

991

d

Fig. 3. Small insect-exclusionary cages placed over potted columnar apple trees.

In 1992, the trees were removed from costorage on 10 Apr. Nutrients were applieaccording to recommendations from leaf anases in 1991. On 24 Apr. 1992, when all treesthe experiment were at the green-tip stage,potted trees and mesh coverings were plain the source orchard in the same locatiothey had occupied during the 1991 season.18 May, the potted and source orchard trewere at full bloom, and the potted trees wehand-pollinated. As in the other study, thlarge-mesh coverings were replaced in 19with an ultraviolet stable material (PC-33).

Potted trees were sprayed with myclobuta[Nova (40% W)] (0.007 g a.i./tree) on 10 Ma21 May, 4 June, and 13 June 1991 and 6 M14 May, and 28 May 1992 for control of appscab, powdery mildew, and cedar apple rucaptan [Captan (50% W); ICI AmericasWilmington, Del.] (0.41 g a.i./tree) 13 Jun1991 for control of apple scab; and proparg(Omite 6E; Uniroyal Chemical Co.Middlebury, Conn.) (0.06 ml a.i./tree) 4 an13 May 1991 for control of European red mi[Panonychus ulmi (Kock)], apple rust mite[Aculus schlectendali (Nalepa)], and two-spot-ted spider mite (Tetranychus urticae Koch).

Each potted tree received ≈8 liters of waterevery other day. The trees and mesh coveriwere removed from the field after leaf faduring 1991 and 1992. The metal frames weleft in the field to ensure that potted trewould be placed in the same location in 199

Environmental characteristics were mea-sured in the large-cage study throughout 1991 and 1992 growing seasons. Tensioeters were installed between the middle trein all plots, and readings were taken twiweekly until 17 June 1991, when the soil wtoo dry to obtain accurate readings, and unthe end of Aug. 1992. Rain gauges were plac

1010

lddy- intheednsOnesree92

il,y,

est;,

te

de

gslres2.

hem-esestiled

between the middle tree and the tree immeately to the north in all plots. Rain volumewere recorded and the gauges were empafter every measurable rain. Light intensitevaporation, relative humidity, and soil anair temperature were measured four timduring 1991 and twice during 1992. Evaportion was measured by filling white plasticontainers (16 × 16 × 10 cm) with 1000 ml ofwater and placing them in all treatments in oblock at 0830 HR and then measuring thevolume of water remaining at 1630 HR. Theother environmental measurements were corded at 0900, 1200, and 1500 HR. Lightintensity was recorded with a LI-COR quantum/radiometer/photometer (model Li-185ALI-COR, Lincoln, Neb.) with the sensor postioned 1 m aboveground in the middle of eatreatment. Air temperature and relative hmidity were measured using a Cole-ParmLCD digital hygrometer (model 3309-50; ColeParmer, Chicago) held 1 m aboveground in tmiddle of each treatment. Soil temperatuwas recorded from a thermometer inserteda depth of 4.5 cm in the middle of each ploContinuous light measurements were recordfrom 12 Aug. through 16 Aug. 1991 in onreplicate of each treatment with a 21micrologger (Campbell Scientific, LoganUtah). Average hourly reading from 040until 2200 HR was recorded.

In the small-cage study, light intensity waexamined using two methods. Light readinwere taken at 1000, 1200, and 1400 HR in alltreatments from two blocks, using the LI-COinstrument positioned in the center of the ca0.5 m below the top on 6 June, 17 July, andAug. 1991 and 3 June and 20 July 199Continuous light readings were recorded the center of the cage 0.5 m below the top fro16 to 20 Aug. 1991 in all treatments in on

H

di-s

tiedy,desa-c

ne

re-

-;

i-chu-er-here tot.edeX,0

sgs

Rge 152.inme

block using the same technique described fthe large cages.

Tree characteristics were recorded for thlarge-cage study throughout the season. OnMay 1991, all trees were marked with whitpaint 10 cm above the graft union. Trundiameter was measured at this mark. Transration rates were measured 19 June, 16 Ju30 July, and 15 Aug. 1991 and 3 June and July 1992 on small potted ‘Starkspur CompaMac’/MM.106 trees that had been placed inside the large-cage treatments. These trewere in 3.8-liter plastic pots wrapped in alumnum foil and watered the night before beinplaced in the field. Trees were weighed a0800 HR, and each mesh treatment in a blochad one potted tree placed inside the cage. Oof the two uncovered groups of trees was usas a control. At 1630 HR on that same day thepotted trees were removed from the treatmenand weighed. A LI-COR portable area mete(model LI-3000) was used to measure 10 leavchosen randomly from each of the 10 pottetrees. The average from the 100-leaf sampwas multiplied by the total number of leaveper tree to obtain the total leaf area of eapotted tree. The amount of water lost durinthe day was expressed as weight loss psquare centimeter of leaf area.

On 25 June 1992, gas exchange [net phosynthesis (Pn), transpiration (E), and leaf coductance (gl)] was measured on two maturebourse-shoot leaves from two in-ground treeper treatment in the large-cage study. Thmeasurements were made with an LCA-2 potable gas-exchange system (ADC, HoddesdoHerts, UK). The system used an open, flowthrough, clamp-on leaf chamber with 6.25 cm2

of leaf area enclosed. Carbon dioxide differetials averaged 20 to 30 ppm. Transpiration (Ewas estimated by the difference between tequilibrium relative humidity and temperature with the leaf enclosed and the relativhumidity and temperature after the leaf waremoved from the chamber. Measuremenwere made in the middle of the day undesunny conditions with leaves at 24 to 26C.

Tree growth was assessed by measuritrunk diameter and current-season shoot elogation on 28 Oct. 1991 and 29 Sept. 199Internode length of current-season growth wmeasured on the latter date. Cumulative trgrowth from both years was also estimated

Fruit from the small-cage experiment warated for color development on the exposeside of the fruit at weekly intervals from 25June through 26 Aug. 1991 and 6 Aug. throug1 Sept. 1992. A scale of 1 to 4 was used wi1= green and 4 = red. During 1991, fruit werallowed to abscise naturally, and length, widtweight, firmness (average force required tpush a 1.2-cm plunger 8.0 mm into the pareflesh of the fruit), level of maturity (starch-iodine test), soluble solids concentration (average of two refractometer readings per fruitand color (<25%, >25% < 33%, >33% < 50%>50% < 80%, and >80% of surface red) wemeasured. In 1992, all fruit were harvested o8 Sept., and the same measurements as in 1were made.

Arthropod populations were estimate

ORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

ro

bt

nhe

le

b

n

ae

v

a

eha

le

uayv

caa8

e

tol-d tonce

achnalgesairityto

ngceeea

itor-red).atewthg

t1ed

w r

ro

o

hrh

me

s a

c

grin

o a

,r.l

-.

ts

-d

n,r

-d-t

s

-,---t

Table 1. Environmental measurements recorded in large insect-exclusionary cages.

Cage pore sizeEnvironmental characteristics Fine Medium Medium-O Large Control

1991Soil water potential (kPa)z 27.7 ay 24.9 a 20.7 a 28.9 a 19.5 aTotal seasonal rain (mm) 90 a 89 a 97 a 92 a 121 bAir temperature (°C) 30.6 a 29.9 a 30.2 a 29.8 a 29.3 aSoil temperature (°C) 27.6 b 26.3 a 26.5 a 27.6 b 29.6 cRelative humidity (%) 39.4 a 40.8 a 41.4 a 40.5 a 42.3 aEvaporation (% of control) 71.1 a 68.5 a 74.2 a 77.5 a 100 bLight (% of control) 79.4 b 65.8 a 81.3 b 84.4 b 100 c

1992Soil water potential (kPa) 12.4 a 10.8 a 10.3 a 12.5 a 11.6 aTotal seasonal rain (mm) 239 a 240 a 264 b 246 a 348 cAir temperature (°C) 29.2 c 28.3 bc 27.8 ab 27.9 ab 26.9 aSoil temperature (°C) 20.1 b 19.1 a 20.0 b 19.8 ab 22.3 cEvaporation (% of control) 67.0 a 66.7 a 75.7 a 65.6 a 100 bLight (% of control) 73.8 a 67.1 a 80.0 a 71.6 a 100 bzMeasurements taken from 15 May through 17 June.yMean separation in rows by LSD at P ≤ 0.05 (rain 1991: F = 43.81; treatment df = 4, error df = 22; 1992: F= 92.15; treatment df = 4, error df = 22; air temperature 1992: F = 8.79; treatment df = 4, error df = 4; soiltemperature 1991: F = 46.75; treatment df = 4, error df = 8; 1992: F = 38.21; treatment df = 4, error df = 4;evaporation 1991: F = 23.42; treatment df = 4, error df = 8; 1992: F = 21.98; treatment df = 4, error df = 4;light 1991: F = 17.18; treatment df = 4, error df = 24; 1992: F = 30.96; treatment df = 4, error df = 4).Nontransformed data presented.

throughout the study in insect-exclusionacages, and a complete account of all arthropexamined is given by Lawson et al. (1993Only phytophagous insect data recorded frothe large insect-exclusionary cages will presented. Leaf damage caused by the spotentiform leafminer was evaluated on 29 Juand 13 Aug. 1991 and 6 Aug. 1992 by couing the total number of mines per leaf on tfive leaves immediately above the most recyear’s growth ring on all five trees per treament. Leaf damage caused by the apple skeletonizer Choreutis pariana (Clerk) wasevaluated on 10 July 1991 and 6 Aug. 1992counting the total number of leaves damagon all five trees per treatment. On 12 Ju1991, the number of mines caused by Lyonetiaspeculella Clemens on the 10 most terminleaves from each of the five trees per treatmwas counted. Potato leafhopper Empoascafabae (Harris) nymphs were counted on ththree most terminal leaves from each of fitrees per treatment on 17 and 28 June andAug. 1991 and on the four most terminleaves 6 Aug. 1992.

The effects of insect-exclusionary cagon tree performance during each year of tstudy were examined separately. Separanalyses were made for two reasons. First,different-aged trees (newly planted vs. 1 yeold) used in this study were morphologicaldifferent (e.g., root zone and canopy sizwhich was expected to influence plant perfomance. Second, weather patterns were sstantially different between years; 1991 wsunny, hot, and dry, while 1992 was cloudcool, and wet. The effect these distinct enronmental conditions had on photosyntheswater status, and overall plant performancould be influenced by the morphologicdifferences between newly planted and 1-yeold trees (Jackson, 1980; Jones et al., 19Schulze et al., 1974).

Statistical analyses. Nonpercent data wernormalized with a log10(x + 1) transformation,and percent data were transformed using arcsin square-root transformation. Data clected on more than one date were subjectea repeated measures analysis of varia(ANOVA). If a significant treatment × timeinteraction existed, data were analyzed at eindividual date or, if appropriate, as seasoaverages (Abacus Concepts, 1989). Averaof the three daily measurements of light, and soil temperature, and relative humidwere analyzed. Data were subjected ANOVA, and means were compared usiFisher’s protected least significant differen(LSD) test (Abacus Concepts, 1989). On-trfruit color ratings were analyzed using weighted ANOVA (SAS Institute, 1985). Frucolor at harvest was evaluated using a proptional-odds model, and means were compausing a t test approximation (McCullagh, 1980For all analyses, the comparisonwise error rwas <0.05. The regression between tree groand light intensity was generated usinSYSTAT (Wilkinson, 1989). Continuous lighintensity data and fruit quality data from 199could not be analyzed because they lackreplication.

HORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

yds).metedlyt-entt-af

yede

lnt

ee 27l

siste

theary),r-b-s,i-is,elr-5;

he

Results

Environmental characteristics. Soil waterpotential, as measured by tensiometers, shono significant differences among treatmentseither 1991 or 1992 (Table 1). Rain penettion was significantly reduced within all cagein 1991 and 1992. In both years, the uncovecontrol plots received more rain than any the covered treatments, and in 1992 the mdium-O plots received more rain than another caged plots.

Air temperature inside the large insecexclusionary cages was similar for all treaments during 1991. However, in 1992, atemperatures in the fine- and medium-mescovered cages were higher than for the uncered control (Table 1).

Soil temperature varied significantly amontreatments in both years. During 1991, tuncovered control had the highest tempeture and the medium and medium-O cages the lowest temperature. During 1992, the ucovered control also had the highest soil teperature and only subtle differences occurramong the caged treatments (Table 1).

Evaporation rates were similar among acovered treatments, but significantly lowethan those for the uncovered control durinboth years (Table 1).

Light measurements recorded continuouin the large cages revealed results similarthose recorded three times during the dHowever, continuous readings showed ththe greatest differences in light intensity ocurred between 1000 and 1400 HR. In bothyears, light intensity varied significantly amonthe treatments, being highest in the uncovecontrol and lowest under medium mesh 1991; in 1992, there was no difference amothe covered cages, all of which received lelight than the uncovered control (Table 1Light measurements in small cages showsimilar results between the two methods measurement. Continuous measurementsvealed the largest differences among tre

edina-sedfe-y

t-t-irh-v-

gea-adn-

-d

llrg

lytoy.at-

edng

ss).edf

re-t-

ments occurred between 1100 and 1300 HR.Significant differences in light intensity wereobserved during both years (Fig. 4). In 1991the control and all open-top cages had highelight intensities than the fully enclosed cagesAmong the open-top treatments, the controand fine-O had the highest light intensity.Among the fully enclosed treatments, themedium mesh had the lowest and the largemesh cages had the highest light intensityDuring 1992, the control had the highest lightlevel followed by all open-top and then thefully enclosed treatments.

Plant attributes. Whole-tree transpirationrates were similar for all treatments in bothyears, as were gas-exchange measuremen(Pn) in 1992 [17.9–19.9 (µmol(CO2)/m2/s)].

Significant differences were recorded forfoliar N concentrations during 1991, whichwere higher in trees from the insecticide treatment compared to the uncovered–unsprayecontrol and large-mesh plots. During 1992, Nconcentrations were lower in trees grown inthe uncovered–unsprayed control plots than ithe insecticide-treated control, fine-, medium-and medium-O-mesh-covered cages. Othefoliar nutrient concentrations were variableamong trees grown in the various plots, withno clear trends evident (Table 2).

Shoot elongation during 1991 was similaramong trees in the various treatments; however, in 1992 all mesh-covered trees producelonger shoots than trees in either of the uncovered plots. The relationship between shooelongation and percent full sunlight is ex-pressed in the regression line that indicatethat shoot growth was negatively related tolight intensity (Fig. 5), which accounted for55.0% of the total variation in shoot growthduring 1992.

Trunk diameter was unaffected by the treatments in 1991 and 1992 (Table 2). Howeverwhen total shoot and trunk growth measurements for the two years were combined, significant differences among trees in the treatments were evident (Table 2). Total shoo

1011

CROP PRODUCTION

ognor

le

ereooh

eeseafer

fdafd–dro)

m,

aft-edel

inanesste

and

eeshileheun-at-adr or

dn-ot

r-yhat--itn,hteedatayn-

t.ingused

ntes.i-ersht

ck-).onreopyones

Table 2. Measurements recorded for columnar apple trees grown in large insect-exclusionary cag

Cage pore size ControlsPlant characteristics Fine Medium Medium-O Large Unsprayed Insectic

1991Transpiration

(% of control) 120 az 136 a 90 a 111 a 100 a ---N (% dry weight) 2.1 a–c 2.2 bc 2.1 a–c 2.0 a 2.0 ab 2.3 cCa (% dry weight) 0.9 a–c 1.0 bc 1.0 c 0.9 ab 0.8 a 0.9 aB (ppm) 30.7 bc 27.5 a 29.2 ab 30.1 bc 31.7 c 29.7 bShoot elongation (cm) 22.7 a 19.1 a 20.6 a 19.1 a 14.8 a 20.7 Trunk diameter (mm) 5.2 a 4.7 a 5.0 a 4.6 a 3.9 a 4.6 a

1992Transpiration

(% of control) 100 a 91 a 102 a 87 a 100 a ---N (% dry weight) 2.4 c 2.4 c 2.4 bc 2.2 ab 2.1 a 2.4 bcB (ppm) 31.1 ab 30.4 a 33.2 a–c 36.6 c 35.2 bc 34.7 bShoot elongation (cm) 42.7 b 42.1 b 41.4 b 40.8 b 31.3 a 33.4 aTrunk diameter (mm) 23.4 a 21.3 a 20.9 a 20.3 a 20.8 a 21.2 Internode length (cm) 1.4 a 1.5 a 1.4 a 1.5 a 1.3 a 1.6 aBranch number 4.6 a 3.6 a 3.8 a 3.1 a 3.8 a 3.9 aTotal growth 1991–92

Shoot elongation (cm) 65.4 c 61.2 c 62.0 c 59.9 bc 46.0 a 54.1 Trunk diameter (mm) 28.6 b 25.9 a 25.8 a 24.9 a 24.7 a 25.7 a

zMean separation in rows by LSD at P ≤ 0.05 (for all analyses: treatment df = 5, error df = 27; N 1991: F2.67; 1992: F = 3.97; B 1991: F = 3.62; 1992: F = 2.73; Ca 1991: F = 2.67; shoot elongation 1992: F total shoot elongation: F = 8.26; total trunk diameter: F = 3.13). Nontransformed data presented.

s

hdmifah

yen

int

pt

d o

Fig. 4. Percentage of full sunlight recorded in fully enclosed, small, insect-exclusionary cages with finmedium (M), and large (L) pore size; in cages with the tops open (F-O, M-O, and L-O); and uncovered control (C). Mean separation for each year by LSD at P ≤ 0.05 (for all analyses: treatment d= 6, error df = 6; 1991: F = 8.34; 1992: F = 43.55).

elongation was greatest for all trees growinside insect-exclusionary cages comparedtrees grown as the uncovered–unsprayed ctrol treatment. Total trunk diameter was larest for trees grown in the fine-mesh treatmeNo significant differences were recorded finternode length or branch count, which wemeasured only in 1992 (Table 2).

On-tree fruit color ratings revealed that thgreatest color variation occurred among appgrown in the various treatments early in thseason and decreased as the season progr(Fig. 6). At the end of the season, fruit colorated while the apples were still on the trewas the same among all treatments during byears (Fig. 6). However, differences in colwere significant for apples harvested from t

1012

nton--t.re

es

essed,,thre

various treatments in 1992. Fruit from treeuncovered plots had the largest proportionfruit in the >80% red category (Table 3). Tlowest proportion of fruit in the >80% recategory came from trees grown in mediuand medium-O-mesh cages. The only signcant difference detected in fruit quality mesurements from 1992 was fruit firmness. Tfirmest fruit were harvested from the unspraand fine-O-mesh treatments (data not show

Phytophagous insect counts. Foliar dam-age caused by the spotted tentiform leafmdiffered significantly among treatments in boyears (Table 4). In 1991, the fewest minesleaf (zero) were recorded on insecticide-treatrees, while the most mines were observeuncovered–unsprayed trees and trees gr

H

inside large-mesh-covered cages. In 1992, trin the fine-mesh cages had more mines per lthan those in any of the other cages. Anothleafminer, L. speculella, caused a pattern ofoliar damage similar to that of the spottetentiform leafminer during 1991; greatest ledamage occurred on trees in the uncageunsprayed control, followed by the large anmedium-O cages, and the least damage (zewas recorded from trees in the fine, mediuand insecticide treatments (Table 4).

Foliar damage caused by the apple leskeletonizer differed significantly among treaments both years. The uncovered–unspraycontrol trees consistently had the highest levof damage (Table 4). In 1991, foliar damageall other treatments was similar, but less ththe uncaged–unsprayed control. In 1992, trein the fine and medium mesh had the leafoliar damage (zero), followed by trees in thlarge-mesh cages and those uncaged sprayed with insecticides.

Potato leafhopper densities varied in thlarge-cage study during 1991 (Table 4). Trein the medium-, medium-O-, and fine-mestreatments had the lowest densities, whhigher densities were found on trees in tlarge-mesh, uncovered–unsprayed, and covered and insecticide-sprayed control trements. In 1992, all mesh-covered trees hzero infestation, which was significantly lowethan that obtained with either the unsprayedinsecticide-sprayed trees.

Discussion

Although subtle differences occurreamong treatments for many of the enviromental variables measured, many were nsignificant. Light intensity was the characteistic most affected by insect-exclusionarcages. Lakso and Seeley (1978) stated tsolar radiation is the predominant factor influencing the microclimate of fruit crops, because it affects photosynthesis, flowering, fruset, and fruit quality (Barden, 1977; Jackso1980; Robinson et al., 1983). The lowest ligintensities recorded in our study were in thmedium-mesh cages, where light was reduc34% (Table 1). Heinicke (1965) stated thmaximum photosynthetic rates in apple mbe reached at one-half to one-third full sulight, and Rom (1991) reported maximumphotosynthesis at 45% to 55% full sunlighPhotosynthetic measurements taken dur1992 showed no reductions inside the varioexclusionary cages when compared to uncagtrees, indicating that the light saturation poifor apple had been achieved inside the cagPlant form does affect photosynthetic effciency, and small trees are more efficient usof light than large trees because of better ligpenetration through the outside canopy (Jason, 1980; Mika and Antoszewski, 1972Jones et al. (1991) stated that all leaves columnar-form trees have good light exposubecause there is no shading from outside canleaves. Therefore, even with a 34% reductiof ambient sunlight, these columnar-form trewere still receiving sufficient light to maintainmaximum photosynthesis.

es.

ide

–cca

c

a

b

== 9.31;

inofe

-i--ed).

erheredonwn

e (F),in anf

ORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

HORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

Fig. 5. Regression of columnar apple tree shoot elongation for 1992 as a function of percent full s

Fig. 6. On-tree fruit color ratings from columnar apples in small insect-exclusionary cages. Color s= green and 4 = red (for all analyses: treatment df = 24; 1991: F = 1.64; 1992: F = 2.21, P ≤ 0.05, weightedanalysis of variance).

n-olarsevehe et

at-ineheside fores.

byar.redct-ov-edn

77)eestly%tersst.n,th,ts.usede-edas

on-re,gesry.

rolere

ryar.ret ineeheesm-re-ichlu-0%els re-omal.,urthe asis

sulted2;

-ls,

unlight.

cale: 1

Although evaporative rates and light intesity were highest in the uncovered contrduring both years, transpiration was similfor potted whole trees, likely because thewell-watered trees were capable of effectiosmoregulation under conditions of higevaporative demands (Jones, 1983; Schulzal., 1974).

Rain accumulation varied among the trements, but soil water potential was similar all treatments. Either water running off thcages moved through the soil and into tcages, or decreased evaporation rates inthe insect-exclusionary cages compensateddecreased rain accumulation inside the cag

Tree growth was not adversely affected insect-exclusionary cages during either yeDuring 1992, tree performance, as measuby shoot growth, was better inside inseexclusionary cages compared to either uncered control. Priestly (1969) reported reducshoot and root growth from apple trees growunder 77% to 90% shade, and Barden (19reported decreased stem diameter in trgrown under 80% shade. However, Pries(1969) reported that trees grown under 90shade grew more than expected. He lashowed that the new shoot growth was lewoody than that of trees grown in full sunlighThese findings may be the result of etiolatiowhich is easily measured by internode lengand in our study was similar for all treatmen

Foliar damage caused by phytophagoinsects was similar for the insecticide-spraytrees and those covered with fine- and mdium-mesh cages. Lawson et al. (1994) showthat direct fruit damage caused by insects wmost severe in the uncovered–unsprayed ctrol and in the large-mesh cages. Therefocontrol of direct and indirect fruit-damaginarthropod pests without the use of insecticidcan be achieved using insect-exclusionacages covered with fine and medium mesh

Although the uncaged–unsprayed conttrees in the large-cage study sustained sevleaf damage, no significant effect of herbivoon tree growth was apparent in either yeHowever, the cumulative effects of sevefoliage damage by arthropods were evidenthe uncaged–unsprayed control when trgrowth was factored over both years of tstudy. The uncaged–unsprayed control trehad the least shoot elongation, but trunk diaeter was unaffected. Ferree et al. (1986) ported that simulated insect damage, whremoved 7.5% of the leaf area, had no infence on photosynthesis. However, when 1of the leaf was removed, photosynthetic levbegan to decrease. Other authors haveported similar decreases in photosynthesis frplants damaged by arthropods (Ames et 1984; Hall and Ferree, 1975). However, in ostudy, photosynthetic rates measured in uncovered–unsprayed control trees werehigh as those in the sprayed control. Thunaltered photosynthetic rate may be the reof the plant compensating for leaf injury causby arthropod feeding (Layne and Flore, 199Syvertsen et al., 1986).

Nutritional status of all trees in the largecage study was within recommended leve

1013

CROP PRODUCTION

do-am-d),

tts.shctnh-y,e tom-

leemaye

t,hey-

testehen-n

allern-edusootongeas

o-ct-n-tu-t

s.teeeew

anlestla-le

s

4.he

n-htde.

.esr.

Table 3. Columnar apple color ratings at harvest from small insect-exclusionary cages (1992).

Cage pore % Fruit in each color categoryz

size n <25% >25% < 33% >33% < 50% >50% < 80% >80%Fine 62 0.0 24.8 43.5 25.3 6.4Fine-O 50 13.4 10.8 29.6 34.3 11.9Medium 52 14.6 22.9 19.6 39.2 3.8Medium-O 64 9.8 22.3 47.6 20.4 0.0Large 34 14.0 4.0 38.7 31.9 13.2Large-O 50 0.0 6.1 33.6 45.2 13.4Unsprayed control 14 0.0 0.0 0.0 50.0 50.0Insecticide control 52 0.0 0.0 5.5 30.0 64.5zFruit color rating is based on the total proportion of the fruit with red color (χ2 = 95.16; df = 7; P ≤ 0.001,proportional-odds model).

Table 4. Phytophagous insect abundance on columnar apple trees with large insect-exclusionary

Cage pore size ControlsInsect Fine Medium Medium-O Large Unsprayed Insecticid

1991Spotted tentiform leafminerz 0.6 by 0.5 b 0.9 c 2.2 d 1.8 d 0.0 aLyonetia speculellaz 0.0 a 0.0 a 0.2 b 0.2 b 0.6 c 0.0 aApple skeletonizerx 0.2 a 0.1 a 0.1 a 0.4 a 2.1 b 0.0 aPotato leafhopperx 0.1 b 0.0 a 0.1 ab 0.2 d 0.3 d 0.2 c

1992Spotted tentiform leafminerz 1.6 b 0.3 a 0.1 a 0.1 a 0.2 a 0.4 aApple skeletonizery 0.0 a 0.0 a 0.4 ab 0.8 b 6.9 d 3.0 cPotato leafhopperw 0.0 a 0.0 a 0.0 a 0.0 a 0.4 c 0.2 bzMean number of mines per leaf.yMean separation in rows by LSD at P ≤ 0.05 (for all analyses: treatment df = 5, error df = 27. Spotted tentifoleafminer 1991: F = 30.97; 1992: F = 2.57; L. speculella 1991: F = 7.94; apple skeletonizer 1991: F = 14.21992: F = 29.24; potato leafhopper 1991: F = 17.89; 1992: F = 7.39). Nontransformed data presenxMean number of leaves damaged per tree.wMean number of insects per leaf.

ythsn

maloiae

rm

vyy

dcoio-

-ame

n

b-intinr

idetr

nfi

y

9 eo

ln)

igirvin wruhea

g

uon

except for Ca, which was below them in 199(Shear and Faust, 1980; Stiles and Reid, 199Nutrient levels appear to be related to phtophagous arthropod damage. Trees in uncovered–unsprayed control and large-mecovered treatments, which had lower nutrielevels, sustained the most severe arthropdamage. Trees in the insecticide and mediumesh treatments, which sustained the learthropod damage, had higher nutrient leveFerree et al. (1986) reported decreased N ccentrations in plants damaged by mites, whAmes et al. (1984) reported lower foliar Cconcentrations in trees with mite damagBodnar et al. (1983) showed spotted tentifoleafminer damage reduced foliar Ca and Mconcentrations. Therefore, lower nutrient leels in the large-mesh and uncovered–unspracontrol trees were likely due to removal bphytophagous arthropods.

Even though the microclimate differeamong treatments, no effect on disease indence was apparent. The only disease served was an occasional apple scab leswhich was effectively controlled by applications of myclobutanil.

Although branch count did not differ significantly among treatments, more laterbranching occurred than we expected. Colunar-form trees are characterized by reduclateral branching (Fisher, 1970; Kelsey aBrown, 1992). Most branches observed duing 1992 were longer than 20 cm. The comnation of the relatively vigorous MM.106 rootstock coupled with high moisture supply 1992 caused many buds to produce vegetashoots. Although branches showed the stroupright growth characteristic of ‘Starkspu

1014

11).-e

h-t

od-

sts.n-

le

.

g-ed

i-b-n,

l-d

dr-i-

veg

Compact Mac’, when trees are planted insinsect-exclusionary cages or at very high dsities, any branching will interfere with lighpenetration and fruit color. Mechanical havesting of these trees, as discussed by Quiand Tobutt (1990), would also be more difcult on branched trees.

Fruit bud formation is directly affected blight interception and does not occur in potions of apple trees receiving <30% of ambiesunlight for prolonged periods (Faust, 198Jackson, 1980). In our study, potted treesthe small-cage study formed fruit in all cagduring 1992, indicating no adverse effect shading on fruit bud formation.

Fruit color recorded at harvest from treessmall insect-exclusionary cages appearedbe directly related to light levels (Fig. 4, Tab3). Optimal fruit color is achieved if radiatioremains >70% of full sunlight (Faust, 1989The medium mesh caused the greatest lreduction and the poorest fruit color early both seasons. All other mesh coverings duced light intensity compared to the uncoered control, and fruit from trees grown mesh-covered cages was greener early inseason compared to uncovered fruit. Hoever, by the end of both growing seasons, fcolor, rated while the apples were still on ttree, was similar among all treatments. Yexamination of harvested fruit revealed ththe best-colored fruit came from treatmenwith the highest light levels. The small-caexperiment was established in the drive rowa source orchard, and shading from these latrees certainly contributed to reduced frpigmentation. All apples harvested had shstems, which kept them close to the tree tru

H

resulting in most fruit having a poorly coloregreen side. However, new columnar gentypes produced by the apple breeding prograt the New York State Agricultural Experiment Station have longer stems (unpublisheso this problem would be reduced.

Fruit firmness was the only other fruicharacteristic influenced by the treatmenThe uncovered–unsprayed and fine-O-metreatments produced the firmest fruit. Insedamage, particularly internal feeding, ofteresults in fruit softening due to increased etylene production (Lau, 1985). In our studmuch of the early insect damage to fruit in thuncovered–unsprayed treatment appearedcause hard scar tissue around the site of daage, resulting in firmer fruit. The small sampsize, resulting from heavy fruit drop beforharvest because of severe insect damage, not accurately reflect fruit firmness in thuncaged–unsprayed control.

In summary, air and soil temperature, lighrain accumulation, and evaporation were tenvironmental characteristics significantlmodified inside exclusionary cages. Exclusionary cages appeared to have the greaeffect on light and evaporation, which werreduced by 15% to 35% when compared to tuncovered control. Even with these enviromental modifications, no negative effects otree growth were observed. During 1992, trees grown in exclusionary cages had longshoots than the controls but not longer interodes. Nutrient concentrations were reduconly in the treatments where phytophagoinsect damage was most severe. Only shelongation, when evaluated over the duratiof the study, was reduced by severe foliadamage caused by insects. Fruit color wreduced in proportion to light reduction.

Lawson et al. (1994) reported that arthrpod suppression obtained with some inseexclusionary cages was similar to that of isecticides. In the absence of insecticides, narally occurring enemies provided significancontrol of many foliage-feeding arthropodThe results presented in this paper indicathat although the microenvironment insidinsect-exclusionary cages was modified, trgrowth was unaffected. These findings shothat insect-exclusionary cages may provideinsecticide-free orchard environment suitabfor detailed evaluations of integrated pemanagement thresholds, predator–prey retionships, and may be used to facilitate appproduction without pesticides.

Literature Cited

Abacus Concepts. 1989. SuperANOVA. AbacuConcepts, Berkeley, Calif.

Ames, G.K., D.T. Johnson, and R.C. Rom. 198The effect of European red mite feeding on tfruit quality of ‘Miller Sturdeespur’ apple. J.Amer. Hort. Sci. 109(6):834–837.

Barden, J.A. 1977. Apple tree growth, net photosythesis, dark respiration, and specific leaf weigas affected by continuous and intermittent shaJ. Amer. Soc. Hort. Sci. 102(4):391–394.

Bodnar, J.M., J.T.A. Proctor, J.E. Laing, and R.ACline. 1983. Nutrient changes in apple leavdue to the spotted tentiform leafminer. J. AmeSoc. Hort. Sci. 108(5):685–688.

cages.

e

rm5;ted.

en-

-lan-

r-nt;insf

in toe

.ht

ne--

the-itet,t

tseofrgeitrtk,

ORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

t

eo

-

e

t

.

i

d

7

e

.k1

n

e

e

.o

9

poc

n

a

8

nl

.

,

i-

.

e

es

,

,-l

s.

.y-

.

.

.

Collyer, E. 1958. Some insectary experiments wipredacious mites to determine their effect othe development of Metatetranychus ulmi(Koch) populations. Entomol. Expt. Appl.1:138–146.

Croft, B.A. 1978. Potentials for research and implmentation of integrated pest managements deciduous tree-fruits, p. 101–115. In: E.H. Smithand D. Pimentel (eds.). Pest control strategie.Academic, New York.

Faust, M. 1989. Physiology of temperate zone frutrees. Wiley, New York.

Ferree, D.C., F.R. Hall, and M.A. Ellis. 1986. Influence of mites and diseases on net photosyntheof apple leaves. The regulation of photosynthsis in fruit trees, p. 56–62. In: A.N. Lakso and RLenz (eds.). Symp. Proc. Publ. New York StaAgr. Expt. Sta., Geneva.

Fisher, D.V. 1970. Spur strains of McIntosh discovered in British Columbia, Canada. Fruit VarHort. Digest 24:27–32.

Hall, F.R. and D.C. Ferree. 1975. Influence otwospotted spider mite populations on photosynthesis of apple leaves. J. Econ. Entomo68(4):517–520.

Heinicke, D.R. 1965. The effect of natural shade ophotosynthesis and light intensity in Red Delcious apple trees. J. Amer. Soc. Hort. Sci. 88:1–8

Herbert, H. and K.H. Sanford. 1969. The influencof spray programs on the fauna of apple orcharin Nova Scotia: Apple rust mite, Vasatesschlechtendali, a food source for predators. CanEntomol. 101:62–67.

Jackson, J.E. 1980. Light interception and utilization by orchard systems. Hort. Rev. 2:208–26

Jones, H.G. 1983. Estimation of an effective sowater potential at the root surface of transpirinplants. Plant Cell Environ. 6:671–674.

Jones, H.G., A.N. Lakso, and J.P. Syvertsen. 198

HORTSCIENCE, VOL. 29(9), SEPTEMBER 1994

hn

-n

s

it

sis-.e

-

f-l.

n-.es

.

-.ilg

5.

Physiological control of water status in tempate and subtropical fruit trees. Hort. Rev. 7:30344.

Jones, H.G., J.W. Palmer, C.R. Rom, and AMackenzie. 1991. Water relations of ‘WijicMcIntosh’ apple trees. J. Hort. Sci. 66(3):31317.

Kelsey, D.F. and S.K. Brown. 1992. ‘McIntosWijcik’: A columnar mutation of ‘McIntosh’apple proving useful in physiology and breediresearch. Fruit Var. J. 46(2):83–87.

Lakso, A.N. and E.J. Seeley. 1978. Environmtally induced responses of apple tree photosthesis. HortScience 13(6):646–650.

Lau, O.L. 1985. Harvest indices for B.C. applBritish Columbia Orchardist 7(7):1A–20A.

Lawson, D.S., W.H. Reissig, J.P. Nyrop, and SBrown. 1994. Management of arthropods columnar apple trees using exclusionary cagCrop Protection (In press.)

Layne, D.R. and J.A. Flore. 1992. Photosynthecompensation to partial leaf area reductionsour cherry. J. Amer. Hort. Sci. 117(2):27286.

Looney, N.E. and W.D. Lane. 1984. Spur-tygrowth mutants of McIntosh apple: A review their genetics, physiology and field performanActa Hort. 146:31–46.

McCullagh, P. 1980. Regression models for ordidata. J. Royal Stat. Soc. Ser. B. 42(2):109–1

Mika, A. and R. Antoszewski. 1972. Effect of leposition and tree shape on the rate of photosthesis in the apple tree. Photosynthetica 6(4):3386.

Priestley, C.A. 1969. Some aspects of the physogy of apple rootstock varieties under reducillumination. Ann. Bot. 33:967–980.

Quinlan, J.D. and D.R. Tobutt. 1990. Manipulatifruit tree structure chemically and genetica

r-1–

D.

–

h

g

n-yn-

s.

K.n

es.

tic in–

efe.

al42.f

yn-1–

iol-ed

gly

for improved performance. HortScience 25:60–64.

Reissig, W.H., R.W. Weires, C.G. Forshey, W.LRoelofs, R.C. Lamb, H.S. Aldwinckle, and S.R.Alm. 1984. Management of the apple maggotRhagoletis pomonella (Walsh) (Diptera:Tephritidae) in disease-resistant dwarf and semdwarf trees. Environ. Entomol. 13:684–690.

Robinson, T.L., E.J. Seeley, and B.H. Barritt. 1983Effect of light environment and spur age on‘Delicious’ apple fruit size and quality. J. Amer.Soc. Hort. Sci. 108(5):855–861.

Rom, C.R. 1991. Light thresholds for apple treecanopy growth and development. HortScienc26(8):989–992.

Sanford, K.H. and H.J. Herbert. 1967. The influencof spray programs on the fauna of apple orchardin Nova Scotia. Can. Entomol. 99:689–696.

SAS Institute. 1985. SAS user’s guide: Statisticsversion 5 ed. SAS Inst., Cary, N.C.

Schulze, E.D., O.L. Lange, M. Evenari, L. Kappenand U. Buschbom. 1974. The role of air humidity and leaf temperature in controlling stomataresistance of Prunus armeniaca L. Under desertconditions. I. A simulation of the daily course ofstomatal resistance. Oecologia 17:159–170.

Shear, C.B. and M. Faust. 1980. Nutritional rangein deciduous tree fruits and nuts. Hort. Rev2:142–163.

Stiles, W.C. and W.S. Reid. 1991. Orchard nutritionmanagement. Cornell Coop. Ext. Bul. 219.

Syvertsen, J.P., L.G. Albrigo, and C.C. Childers1986. Photosynthesis and water use efficiencof spider mite damaged citrus leaves. The regulation of photosynthesis in fruit trees, p. 52–56In: A.N. Lakso and R. Lenz (eds.). Symp. ProcPubl. New York State Agr. Expt. Sta., Geneva

Wilkinson, L. 1989. SYSTAT: The system for sta-tistics. SYSTAT, Evanston, Ill.

1015