Profitability of Sensor-based Irrigationin Greenhouse and Nursery Crops

Erik Lichtenberg1,3, John Majsztrik2, and Monica Saavoss1

ADDITIONAL INDEX WORDS. economics, disease management, gardenia production,precision irrigation, production time management, sensor technology, wirelesssensor networks

SUMMARY. Improvements in sensor technology coupled with advances in knowledgeabout plant physiology have made it feasible to use real-time substrate volumetricwater content sensors to accurately determine irrigation timing and applicationrates in soilless substrates in greenhouse and container production environments.Sensor-based irrigation uses up-front investments in equipment and systemcalibration in return for subsequent reductions in irrigation water use andassociated costs of energy and labor, spending on fertilizer, and disease losses. It canalso accelerate production time. We present formulas for assessing profitabilitywhen benefits and costs are separated in time and apply those formulas using datafrom an experiment on production of gardenia [Gardenia augusta ‘MADGA 1’(Heaven ScentTM)]. Sensor-controlled irrigation cuts production time and croplosses by more than half. Annualized profit under the wireless sensor system wasover 1.5 more than under the nursery’s standard practice, with the bulk of theincrease in profit due to the reduction in production time. These results indicate thatcontrolling irrigation using wireless sensor systems is likely to increase profitabilitysubstantially, even if efficiency gains are not as high as those achieved underexperimental conditions.

Improvements in sensor technol-ogy coupled with advances inknowledge about plant physiology

have made it feasible to use real-timesubstrate volumetric water contentsensors to accurately determine irri-gation timing and application ratesin soilless substrates in greenhouseand container production environ-ments (Belayneh et al., 2013; Burnettand van Iersel, 2008; Lea-Cox et al.,2010). Wireless transmission of sen-sor data allows for real-time manage-ment of the irrigation system. Wirelesssensor-based irrigation managementoffers significant potential benefits togreenhouse operators (Nemali et al.,2007; Scoggins and van Iersel, 2006).By matching water applications withmoisture availability and plant uptakerates, sensor-based irrigation can reduceirrigation water use without riskingadverse consequences from under- or

overwatering. Better matching of mois-ture availability with plant demandscan also reduce leaching of nutrients,resulting in fertilizer savings as well aswater savings. In the many watershedswhere runoff of nutrients causes eu-trophication, reductions in nutrientleaching can benefit the environmentas well as contribute to growers’ bot-tom lines (Majsztrik et al., 2013).Wireless soil moisture sensor systemsprovide more accurate measurementsof substrate moisture status than qual-itative methods (weight, appearance,length between irrigation cycles, etc.)and require less labor (Lea-Cox et al.,2010; Majsztrik et al., 2011; Nemaliand van Iersel, 2006; van Iersel et al.,2009).

Savings in irrigation water, labor,energy, and fertilizer expendituresare obvious potential benefits of soilmoisture sensor systems. Other po-tential benefits of sensor systems areless obvious. Greater precision in main-taining soil moisture at desired levelscan lower disease pressure, often causedby precautionary overwatering (Chappell

et al., 2012). Lower disease pressuremeans reduced crop losses in additionto less fungicide use. Better matchingof water availability to plant uptake hasalso been shown to accelerate growthin some instances (Chappell et al.,2013). Shorter production times meanhigher profits, just as lower costs do.

Achieving these potential bene-fits requires installation and calibra-tion of equipment and software(Kohanbash et al., 2013). In essence,sensor-based irrigation systems sub-stitute capital for water and associatedvariable inputs such as energy, labor,and fertilizer (Shani et al., 2009). Theprofitability of investing in wirelesssensor systems thus depends on therelative magnitudes of benefits andcosts. Those benefits and costs are in-curred at different times: Investmentin sensor systems is made up front,while reductions in spending on wa-ter, energy, labor, fertilizer, and pes-ticides accrue later on, as do anybenefits from shortening productiontime or reducing disease losses. Thispaper presents a methodology for cal-culating the profitability of investing insensor-based irrigation managementthat takes these differences in timinginto account. It then applies that meth-odology to data from gardenia pro-duction in a Georgia nursery as part ofa project focused on implementationof wireless irrigation sensor networksfor ornamental plant production.

Materials and methodsWhen benefits and costs accrue

at different points in time, calculatingprofit—or, indeed, comparing them inany way—requires putting benefits andcosts on a common time footing. Themost convenient method is convertingall revenues and costs to constant peri-odic payments; e.g., annualizing them.We begin by discounting all revenuesand costs to convert them to theirpresent values. We then calculate thepresent value of profit, which we con-vert to a constant annual payment (orloss). Finally, we calculate profit (orloss) per unit area to permit scaling upor down.

UnitsTo convert U.S. to SI,multiply by U.S. unit SI unit

To convert SI to U.S.,multiply by

0.0929 ft2 m2 10.76390.0283 ft3 m3 35.31473.7854 gal L 0.2642

This paper is part of a series of manuscripts describingthe research and development completed by theSCRI–MINDS (Managing Irrigation and Nutritionthrough Distributed Sensing) project. The authorsgratefully acknowledge funding and support from theUSDA–NIFA Specialty Crops Research Initiative;Award #2009-51181-05768.

1Department of Agricultural and Resource Econom-ics, 2102 Symons Hall, University of Maryland,College Park, MD 20742

2Department of Plant Science and Landscape Archi-tecture, 2102 Plant Science Building, University ofMaryland, College Park, MD 20742

3Corresponding author. E-mail: elichten@umd.edu.

770 • December 2013 23(6)

Formally, let qit denote thequantity sold at time (year, month,week) t using production system i, ptthe unit price of the crop received attime t and eijt expenditures on input jat time t, where i = {s,n} denotes asensor-based or non-sensor-basedsystem. Net profit (or loss) earned attime t using the sensor-based system

is pt qst �X

j

esjt . The present value of

net profit (or loss) earned at time tusing the sensor-based system ispt qst �

Pj esjt

ð1þ rÞt, where r is the periodic

interest rate. The present value of netprofit (or loss) earned at time t usingthe non-sensor-based system ispt qnt �

Pj enjt

ð1þ rÞt.

Let K denote the initial invest-ment cost of the sensor-based system,f t the sensor equipment expendituresat time t, Cs and Cn the crop produc-tion costs at planting incurred underthe sensor-based and non-sensor-based systems, and Ts and Tn therespective times until the end of theproduction cycle under the sensor-based and non-sensor-based systems.The present value of profit earnedover the entire production cycleunder the sensor-based system is

V s¼XT s

0

pt qst� f t �PJ

1 esjt

ð1þ rÞt�Cs �K .

The present value of profit earnedover the entire production cycle un-der the non-sensor-based system is

V n ¼XT n

0

pt qnt �PJ

1 enjt

ð1þ rÞt� Cn.

Annualization of profit may re-quire adjustments in time period ac-counting. For many nursery operations,it makes sense to account for revenuesand costs on a less-than-yearly basis;e.g., monthly or weekly. In such cases,the periodic interest rate r is not anannual interest rate. If revenues andcosts are tracked monthly or weekly,then the respective monthly andweekly interest rates are 1/12 and1/52 of the annual interest rate; i.e.,i ¼ 12r or i ¼ 52r. Similar adjust-ments may be needed for time ofproduction. If revenues and costs aretracked monthly, the production pe-riod on a yearly basis is A ¼ T =12 ; ifrevenues and costs are tracked ona weekly basis, the production periodon a yearly basis is A ¼ T =52.

Annualized profit under the sensor-

based system is then Z s ¼iV s

1� 1

ð1þ iÞAs

while annualized profit under thenon-sensor-based system is Z n ¼

iV n

1� 1

ð1þ iÞAn

. The annualized cost

of the sensor system is k ¼i K þ

PT s

t¼0 f t

h i

1� 1

ð1þ iÞAs

.

The change in annualized profitfrom switching to the sensor-basedsystem from the non-sensor-basedsystem is Z s � Z n. The annual rateof return on capital invested in the

sensor-based system isZ s � Z n

k. The

payback period, expressed in years, isthe inverse of the annual rate of

return,k

Z s � Z n.

We use this procedure in a casestudy of the likely profitability ofinvesting in a wireless sensor systemfor scheduling irrigation in greenhouseproduction of ‘MADGA 1’ gardenias(Heaven Scent�). Estimates of theeffects of using a wireless sensor systemwere taken from experimental data. Acharacterization of production activi-ties and estimates of sales and costsincurred with and without the wirelesssensor system was obtained from thegrower cooperating in the experiment.

The experiment involved usinga wireless network to monitor envi-ronmental conditions and control sub-strate water content of 1-gal gardeniasin 10 bays covering 20,000 ft2 in anunheated greenhouse at a large com-mercial nursery in Dearing, GA. Linerswere planted into 1-gal containers filledwith a bark-based substrate in June2010, with each bay containing 2340gardenias. An irrigation controller(Moisture Clik IL200-MC; Dynamax,Houston, TX) that uses a dielectricsoil moisture sensor (SM200; Dyna-max) to measure substrate water con-tent was used to control irrigation.These controllers use a single soil mois-ture probe, which was placed in a con-tainer centrally located within eachblock of plants to avoid edge effects.Irrigation application was triggeredwhen substrate water content droppedbelow 0.20 m3�m–3. A 24-h timer wasused to power off the Moisture Clik

controllers between 1700 and 0800 HR

to prevent irrigation at night. A watermeter attached to each bay was used tomonitor water use. Irrigation volumeswere recorded monthly. Other thanirrigation, production followed thestandard cultural practices of the nurs-ery (Chappell et al., 2013). All plantswere harvested by the end of May2011. Changes in time of productionwere estimated by comparing actualsales of the experimental crop with salesfrom the nursery’s historical record.

Nursery personnel controlled ir-rigation in the remaining five bays ofthe greenhouse, and were told toirrigate according to their regular prac-tices. A previous experiment at the samenursery showed that water use couldbe reduced by over 80% with no ad-verse effects on production (van Ierselet al., 2009). Perhaps inspired by thatexperiment, nursery personnel ig-nored instructions and used the watermeters to match the irrigation volumeapplied by the sensors in the five baysthey were controlling. As a result, theexperiment was unable to determinechanges in water use due to the wire-less sensor system. Thus, instead ofcomparing experimental and controlplants in the same greenhouse, produc-tion in the experimental greenhousewas compared with the operation’stypical production practices, outcomes,and costs for this gardenia cultivar.Gardenia production elsewhere in thenursery under the nursery’s standardpractice during the experimental pe-riod conformed to the historic patternof maturation, losses, and sales, in-dicating that this comparison is valid.The comparison data were providedby the cooperating grower, who keepsa detailed accounting of per-plant vari-able costs for business managementpurposes. Water use was not includedin the grower’s accounting, so we wereunable to determine changes in wateruse under the wireless sensor systemby comparing water use in the exper-iment with historical water use.

Although the experiment wasunable to examine changes in wateruse, it did find substantial changes intime of production. Actual sales ofthe experimental crop were comparedwith sales from the nursery’s historicalrecord (Fig. 1). Plants grown withsensor-controlled irrigation were firstready for sale by February of the fol-lowing year, 8 months after planting,and all were sold by the end of the

• December 2013 23(6) 771

spring planting season, 11 monthsafter planting. In contrast, historicalsales records indicate that the firstplants were typically ready for sale byMay of the following year (11 monthsafter planting), missing much of thespring planting season. Sales wereusually slow during the summer, thenpicked up during August through thebeginning of November. The plantswere then overwintered until Februaryof the next year. The full crop was thuscompletely sold by the beginning ofMay, almost 2 years after planting.Thus, under the experimental regime,sales were completed during the samemonth that plants had historically firstbeen ready for sale and the sale of allplants in the crop was completed inhalf the time needed historically. Thesales-weighted average time to sale ofthe experimental crop was 9.7 months,36% below the sales-weighted averagehistoric time to sale of 15.2 months.

The experiment also found sub-stantial changes in crop losses. His-torically, 18.2% of plants planted diedor could not be sold (Fig. 1). Theexperimental regime cut those lossesin half, mainly by eliminating lossesdue to disease (Chappell et al., 2013).

Production practices and theirassociated costs were obtained fromthe nursery operator. Liners cost $0.50each. Pots, substrate, fertilizer, plantinglabor, and herbicide treatment cost$0.7467 per plant. Monthly labor,overhead, and other costs amountedto $0.0568 per plant. Plants receiveda foliar fungicide treatment every 2weeks from April through October ata total cost of $0.025 per plant. Un-der the standard production regime,

each plant received a fungicide drenchevery 6 months at a cost of $0.045 perplant. Use of the wireless sensor systemeliminated the need for (and cost of)this fungicide drench. Finally, eachplant received a top dress of fertilizerafter 12 months. Because all plants pro-duced under the experimental regimewere sold by the end of 11 months, useof the wireless sensor system eliminatedthis cost as well.

Crop losses were assumed to oc-cur at a constant monthly rate duringthe entire time plants were present inthe nursery, beginning 6 months afterplanting and continuing until allremaining plants were sold. The coop-erating grower confirmed that thisassumption was reasonable. Plants soldwere valued at the nursery’s wholesaleprice of $6.50 each.

Revenue, expenditures, and profitwere calculated on a monthly basisdiscounted at an interest rate of 6%(or 0.5% per month), roughly theaverage interest rate paid by smallbusinesses reported by the NationalFederation of Independent Busi-nesses as of Dec. 2012 (Dunkelbergand Wade, 2013). Interest was com-pounded continuously. The presentvalues of revenue, cost, and profitunder the experimental and historicsystems were obtained by adding updiscounted monthly revenue, expen-ditures, and profit over the entireproduction cycle. Those present valueswere then annualized using the same6% annual rate assuming a productioncycle lasting 11 months under theexperimental system and 22 monthshistorically. By putting all revenuesand costs on an equal time footing,

annualization avoids the need to ac-count explicitly for the differentialcost of greenhouse space occupiedby the crop under historic productionprocedures, but freed up under theexperimental sensor-based system.

The wireless sensor system con-sisted of 40 sensors (10HS; DecagonDevices, Pullman, WA) costing $70apiece, connected to eight nodes(Decagon nR5 nodes) costing $675each, which transmitted the sensordata to a single base station costing$60. Other electrical supplies usedwere estimated to cost $1000. Thedata were processed using softwarewhose license was estimated to cost$450. The computer used to run thesoftware and view the data was esti-mated to cost $600. The entire systemwas annualized over a 3-year lifetime,a very conservative assumption thatlikely overstates the true yearly cost.

Results and discussionInvestment in the wireless sensor

system increased the overall present-value cost of production by 3.6% overstandard practice at this nursery(Table 1). The present value of revenueunder the wireless sensor system was12.8% higher than under the nursery’sstandard practice, in part because re-ductions in disease losses increasedthe total number of plants harvestedby 11.1% and in part because salesstarted and ended much earlier. Be-cause more plants were harvested, theproduction cost per plant was 6.7%lower with the wireless sensor systemthan under the standard practice. Asa result, profit was 20.6% higher withthe wireless sensor system than underthe nursery’s standard practice.

This difference in profit under-states the actual increase in profit dueto use of the wireless sensor system.Shortening the production cycle freesup greenhouse space for plantinganother crop. The production statis-tics in Fig. 1 indicate that using thewireless sensor system comes close todoubling the number of plants thatcan be grown in the 22 months neededto finish out a single crop under thenursery’s standard practice. Annualiz-ing revenue, expenditures, and profittakes this timing difference into ac-count, as noted earlier, avoiding theneed for an explicit accounting of thecost of greenhouse space. On an an-nual basis, revenue was 119% higherunder the wireless sensor system than

Fig. 1. Temporal distribution of sales and losses comparing a wireless irrigationsensor system and standard nursery practice for production of gardenias in anunheated greenhouse in Dearing, GA.

772 • December 2013 23(6)

SPECIAL SERIES

under the nursery’s standard practice(Table 2). Annual expenditures onproduction (net of sensor system costs)were 64% higher under the wirelesssensor system than under the nurs-ery’s standard practice. Adding an-nualized sensor system costs bringsproduction costs using the wirelesssensor system to 76% over productioncosts compared with the nursery’sstandard practice. The percentage dif-ference in profit is a weighted sum ofpercentage differences in revenue andcost, where the percentage difference

in revenue is weighted by the ratio ofrevenue to profit and the percentagedifference in cost is weighted by theratio of cost to profit. Revenue underthe industry’s standard practice was185% of profit while cost was 85% ofprofit, so that annual profit was 156%higher under the wireless sensor sys-tem than under the nursery’s standardpractice. The payback period for invest-ment in the wireless sensor system wasless than 1 month (although payback isnot realized until plants are sold, begin-ning 8 months after planting).

Not surprisingly, the increase inprofitability due to the use of the wirelesssensor system was not sensitive to thechoice of a 6% interest rate. In fact,the increase in annualized profit wasslightly higher at higher interest rates:At an interest rate of 10%, the increasein annualized profit was 157% while atan interest rate of 20% the increase inannualized profit was 161%. A higherinterest rate has two opposing effectson the difference in annualized profit.On the one hand, it increases theannualized cost of investing in a wireless

Table 1. Monthly accounting of present-value expenditures, sales, revenues, and profits for the production of gardeniasat a greenhouse operation in Dearing, GA.

MonthTime after

planting (mo.)

Standard practice Sensor-based irrigation

Expenditure($)

Quantitysold

Revenue($)

Profit($)

Expenditure($)

Quantitysold

Revenue($)

Profit($)

June 0 29,173 39,483July 1 1,439 –1,439 1,439 –1,439August 2 1,432 –1,432 1,432 –1,432September 3 1,425 –1,425 1,425 –1,425October 4 1,417 –1,417 1,417 –1,417November 5 1,296 –1,296 1,296 –1,296December 6 2,287 –2,287 2,277 –2,277January 7 1,256 –1,256 1,245 –1,245February 8 1,236 –1,236 1,219 1,861 11,449 10,230March 9 1,216 –1,216 1,194 8,190 50,135 48,941April 10 1,302 –1,302 1,271 6,222 37,898 36,626May 11 1,281 2,499 15,374 14,093 1,244 4,999 30,297 29,052June 12 3,369 1,170 7,162 3,793 0 0 0July 13 1,027 1,042 6,347 5,320 0 0 0August 14 947 4,042 24,497 23,550 0 0 0September 15 696 2,819 17,000 16,303 0 0 0October 16 518 2,127 12,763 12,245 0 0 0November 17 349 372 2,221 1,872 0 0 0December 18 565 0 0 –565 0 0 0January 19 301 0 0 –301 0 0 0February 20 287 1,819 10,698 10,412 0 0 0March 21 179 2,872 16,807 16,628 0 0 0April 22 21 383 2,230 2,209 0 0 0Total 53,019 19,145 115,099 62,080 54,941 21,272 129,779 74,837

Table 2. Annualized revenue, expenditures, and profit: wireless sensor system versus standard practice, based on greenhouseproduction of gardenias for an operation in Dearing, GA.

Standardpractice ($)

Sensor-basedirrigation ($)

Sensor-based irrigation,shorter production

time only ($)Sensor-based irrigation,

no losses only ($)

Annualized revenue 66,297.36 145,505.64 130,956.46 73,666.51Annualized production expenditures 30,539.11 50,039.93 49,995.21 30,926.04Annualized sensor system cost 3,755.24 3,755.24 3,755.24Annualized profit 35,758.24 91,710.47 77,206.01 38,985.23Annualized revenue per square footz 3.31 7.28 6.55 3.68Annualized production

expenditures per square foot1.53 2.50 2.50 1.55

Annualized sensor system costper square foot

0.19 0.19 0.19

Annualized profit per square foot 1.79 4.59 3.86 1.95z$1.00/ft2 = $10.7639/m2.

• December 2013 23(6) 773

sensor system. On the other hand, itreduces the present value of revenueunder the nursery’s standard irriga-tion practice. These two opposing ef-fects come close to cancelling each otherout at reasonable interest rates.

Use of the wireless sensor systemincreases profitability in two ways: byaccelerating production and by re-ducing losses. We conducted a sensi-tivity analysis to estimate the extent towhich the overall increase in profitcould be attributed to each. To esti-mate the extent to which the increasein profit was due to accelerating theproduction cycle, we calculated annu-alized revenue, expenditures, and profitassuming the shorter production timeexperienced under the wireless sensorsystem but kept disease losses at thelevel experienced under the nursery’sstandard practice. To estimate theextent to which the increase in profitwas due to reductions in losses, wecalculated annualized revenue, expen-ditures, and profit assuming lossesexperienced under the wireless sensorsystem but kept the temporal distri-bution of sales at the level experiencedunder the nursery’s standard practice.

Reducing production time wasresponsible for the most of the increasein profit (Table 2). Annual revenue inthe reduced production time scenariowas almost double than that underthe nursery’s standard practice, whereasannual production expenses net of sen-sor system costs remained 64% higher.As a result, annual profit was 116%higher in this scenario than that underthe nursery’s standard practice.

Reducing losses had a muchsmaller impact on profit (Table 2).Annual revenue in the reduced lossesscenario was only 11% higher thanthat under the nursery’s standard prac-tice while annual production expendi-tures (net of sensor system costs) werealmost the same. As a result, annualprofit was only 9% higher in this sce-nario than under the nursery’s standardpractice. Accelerated production timeand reduced losses taken separatelyaccount for a 125% increase in profit.The remaining 31% increase in profit isdue to the combination of the two.

ConclusionSensor-based irrigation manage-

ment can reduce irrigation water useand associated costs of energy andlabor, spending on fertilizer (and leach-ing of excess nutrients), and disease

losses. It can also accelerate productiontime. In essence, sensor-based irrigationsubstitutes capital for variable inputs,trading off up-front investment inequipment and system calibration,for benefits such as reduced input use,reductions in losses, and acceleration ofproduction that accrue later on.

We present formulas for assessingprofitability when benefits and costs areseparated in time and apply those for-mulas using data from an experiment ongardenia production. Controlling irriga-tion using data from moisture sensorsled to substantial reductions in bothproduction time and crop losses: Theweighted average time from planting tosale was over one-third lower while croplosses were reduced by 50%. Annualizedprofit under the wireless sensor systemwas over 1.5 times more than under thenursery’s standard practice, with the bulkof the increase in profit due to thereduction in production time. Theseresults indicate that controlling irrigationusing wireless sensor systems is likelyto increase profitability substantially,even if efficiency gains are not as highas those achieved in this experiment.

Using wireless sensor systemscan have environmental as well asprivate economic benefits (see Majsztriket al., 2013). More efficient water usecan reduce pressure on increasinglyscarce water supplies. More efficientenergy use can reduce carbon emissions,and more efficient fertilizer use canlower runoff and thus nutrient pollutionof waterways. The design and conductof the experiments used in our analysisprevented us from estimating thesepotential benefits to society at large,but this technology clearly has promiseas a win-win combination of economicand environmental improvements.

Literature citedBelayneh, B.E., J.D. Lea-Cox, and E.Lichtenberg. 2013. Costs and benefits ofimplementing sensor-controlled irriga-tion in a commercial pot-in-pot containernursery. HortTechnology 23:760–769.

Burnett, S.E. and M.W. van Iersel. 2008.Morphology and irrigation efficiency ofGaura lindheimeri grown with capaci-tance sensor-controlled irrigation. Hort-Science 43:1555–1560.

Chappell, M., S.K. Dove, M.W. van Iersel,P.A. Thomas, and J. Ruter. 2013. Imple-mentation of wireless sensor networks forirrigation control in three container nurs-eries. HortTechnology 23:747–753.

Chappell, M., M. van Iersel, E. Lichtenberg,J. Majsztrik, P. Thomas, J. Ruter, and S.Wells. 2012. Benefits of Precision Irrigationof Gardenia augusta ‘‘Heaven Scent’’�: Re-ducing Shrinkage, Shortening the CroppingCycle, and Economic Impact. Proc. South-ern Nursery Assn. Res. Conf. 57:321–323.

Dunkelberg, W.C. and H. Wade. 2013.National Federation of Independent Busi-ness: Small business economic trends,January 2013. Natl. Federation Indepen-dent Business, Nashville, TN.

Kohanbash, D., G. Kantor, T. Martin, andL. Crawford. 2013. Wireless sensor networkdesign for monitoring and irrigation control:User-centric hardware and software devel-opment. HortTechnology 23:725–734.

Lea-Cox, J.D., G.F. Kantor, W.L.Bauerle, M.W. van Iersel, C. Campbell,T.L. Bauerle, D.S. Ross, A.G. Ristvey, D.Parker, D. King, R. Bauer, S.M. Cohan, P.Thomas, J.M. Ruter, M. Chappell, M.Lefsky, S. Kampf, and L. Bissey. 2010. ASpecialty Crops Research Project: UsingWireless Sensor Networks and Crop Mod-eling for Precision Irrigation and NutrientManagement in Nursery, Greenhouse andGreen Roof Systems. Proc. SouthernNursery Assn. Res. Conf. 55:211–215.

Majsztrik, J.C., A.G. Ristvey, and J.D.Lea-Cox. 2011. Water and nutrient man-agement in the production of container-grown ornamentals. Hort. Rev. 38:253–297.

Majsztrik, J.C., E.W. Price, and D.M. King.2013. Environmental benefits of wirelesssensor-based irrigation networks: Case-study projections and potential adoptionrates. HortTechnology 23:783–793.

Nemali, K.S., F. Montesano, S.K. Dove,and M.W. van Iersel. 2007. Calibrationand performance of moisture sensors insoilless substrates: ECH2O and Thetaprobes. Sci. Hort. 112:227–234.

Nemali, K.S. and M.W. van Iersel. 2006.An automated system for controllingdrought stress and irrigation in pottedplants. Sci. Hort. 110:292–297.

Scoggins, H.L. and M.W. van Iersel.2006. In situ probes for measurement ofelectrical conductivity of soilless substrates:Effects of temperature and substrate mois-ture content. HortScience 41:210–214.

Shani, U., Y. Tsur, A. Zemel, and D.Zilberman. 2009. Irrigation productionfunctions with water-capital substitution.Agr. Econ. 40:55–66.

van Iersel, M., R.M. Seymour, M. Chappell,F. Watson, and S.K. Dove. 2009. SoilMoisture Sensor-based Irrigation ReducesWaterUseandNutrientLeaching inaCom-mercial Nursery. Proc. Southern NurseryAssn. Res. Conf. 54:17–21.

774 • December 2013 23(6)

SPECIAL SERIES