Biological Control 76 (2014) 41–51

Contents lists available at ScienceDirect

Biological Control

journal homepage: www.elsevier .com/locate /ybcon

Management to control citrus greening alters the soil food weband severity of a pest–disease complex

http://dx.doi.org/10.1016/j.biocontrol.2014.04.0121049-9644/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author at: Citrus Research and Education Center (CREC), University of Florida (UF), 700 Experiment Station Road, FL 33850, USA. Fax: +1 863 9E-mail addresses: r.camposherrera@ufl.edu, raquel.campos@ica.csic.es (R. Campos-Herrera).

R. Campos-Herrera a,b,⇑, F.E. El-Borai a,c, T.E. Ebert a, A. Schumann a, L.W. Duncan a

a Citrus Research and Education Center (CREC), University of Florida (UF), 700 Experiment Station Road, FL 33850, USAb Instituto de Ciencias Agrarias, CSIC, Serrano 115 Dpdo, Madrid 28006, Spainc Plant Protection Department, Faculty of Agriculture, Zagazig University, Egypt

h i g h l i g h t s

� Advanced production system (APS)speeds citrus tree growth andmaturity.� APS increased abundance of

herbivorous arthropods, nematodesand oomycetes.� APS reduced numbers of

steinernematid entomopathogenicnematodes.� Modification of APS to mitigate these

non-target effects may be feasible.

g r a p h i c a l a b s t r a c t

An advanced production systems (APS) was designed to grow citrus trees more quickly than conventionalcitriculture (CC) methods to mitigate the impact of the bacterial disease of citrus huanglongbing. Changesin the soil physico-chemical properties caused by daily fertigation in APS increased the abundance of theroot herbivore Diaprepes abbreviatus (A) and the root rotting oomycete Phytophthora nicotianae (B). APShad additional non-target effects on some species of native entomopathogenic nematodes (EPNs; C)reported to modulate population growth of both D. abbreviatus and P. nicotianae.

a r t i c l e i n f o

Article history:Received 5 February 2014Accepted 25 April 2014Available online 6 May 2014

Keywords:Bacterial disease huanglongbing (HLB)CitrusEntomopathogenic nematodeIrrigation systemNematophagous fungiPolypropylene mulchSoil food web

a b s t r a c t

Since 2005, the Florida citrus industry has faced the need to control the devasting bacterial diseasehuanglongbing (HLB). Advanced production systems (APS) were designed to grow citrus trees morequickly than conventional citriculture (CC) methods in order to mitigate the impact of HLB. Daily fertiga-tion required by APS produces changes in the soil physical–chemical properties compared to those inconventionally managed orchards. We used real-time PCR in an ongoing field experiment to comparethe effects of APS and CC on more than a dozen metazoan and microorganism species in soil food websthat affect larvae of a major arthropod pest of citrus, Diaprepes abbreviatus. Soil chemical properties,citrus performance, weevil occurrence, and abundance of free-living and plant-parasitic nematodes werealso evaluated. The effects of polypropylene mulch that provides a barrier to soil entry by D. abbreviatuslarvae were also investigated in each of the two cultural systems. Trees grew significantly larger in APSand mulching increased tree growth and reduced tree mortality, thereby increasing fruit yield per ha. APSincreased the fruit yield in 2011; however, by 2013 the number of fruit per tree was not affected by any ofthe treatments. Root mass density increased in APS, but decreased under mulch. The numbers of plant-parasitic and free-living nematodes and some natural enemies of nematodes such as Catenaria sp. andPaecilomyces lilacinus were more abundant in the treatments with greater root mass density. Bothorganisms in the D. abbreviatus–Phytophthora nicotianae pest–disease complex were more abundant in

56 4631.

42 R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51

APS than in CC, whereas fewer steinernematid entomopathogenic nematodes (EPNs) that prey on insectlarvae occurred in APS. By contrast, heterorhabditid EPNs tended to be more numerous in APS than in CC,although they comprised <25% of the EPN communities in any treatment. Major differences between APSand CC in almost all of the measured physical and chemical soil properties provide a basis for controlledstudies to understand why EPN taxa responded differently to these treatments and how APS and soilsgenerally might be modified to conserve the beneficial activities of nematodes.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Entomopathogenic nematodes (EPNs) belonging to the familiesHeterorhabditidae and Steinernematidae are obligate pathogens ofsoilborne arthropods (Georgis et al., 2006). These nematodes arevirtually ubiquitous in soils of all continents except Antarctica(Adams et al., 2006). In Florida, native entomopathogenic nema-todes (EPNs) are important natural enemies of the key citrus pestDiaprepes abbreviatus (Dolinski et al., 2012). The larvae of thisweevil reside exclusively in the soil where, during several monthsof development, they cause extensive damage to the fibrous andwoody roots of citrus trees and facilitate infection by the phyto-pathogenic oomycetes Phytophthora nicotianae and Phytophthorapalmivora (Graham et al., 2003). Naturally occurring EPNs killedweevil larvae buried in orchards at rate exceeding 50% per weekand variability in the regional abundance of weevils may be linkedto the species diversity and community composition of native EPNspecies (Duncan et al., 2003, 2007, 2013; Futch et al., 2005;Campos-Herrera et al., 2013a).

Advanced production system (APS) is a term applied to an ‘openhydroponic’ method of citriculture in which trees are fertigateddaily via drip or microsprinker irrigation systems. The new systemis being studied as a method to bring young trees to maturity morerapidly to recoup the investment in an orchard as quickly aspossible, before trees become non-productive due to the bacterialdisease huanglongbing (citrus greening) (Schumann et al., 2013).First detected in Florida in 2005, the bacterial pathogen and itspsyllid vector are ubiquitous in the citrus growing regions wherethey threaten the existence of the industry (Graham et al., 2013).If widely adopted, APS will fundamentally change the physical/chemical properties of orchard soil, compared to soil underconventional citriculture (CC) where trees are irrigated only peri-odically and fertilized just 3–4 times per year. These changes couldalter soil food webs in ways that affect the health of the trees. Forexample, APS was detrimental to native Steinernema diaprepesi andexotic Steinernema riobrave applied to soil (Campos-Herrera et al.2013b). However, augmented Heterorhabditis indica persistedequally well in APS and CC plots. The bacterium Paenibacillus sp.,a species-specific phoretic associate of S. diaprepesi that impairsnematode motility (El-Borai et al., 2005; Enright and Griffin,2005), was more abundant on the cuticles of S. diaprepesi in APSplots (Campos-Herrera et al., 2013b). These bacteria may havecontributed to reducing S. diaprepesi in APS, whereas molecularmonitoring of several species of nematophagous fungi recoveredfrom the nematode samples did not implicate any of the fungalnatural enemies as potentially affecting the EPNs differently inthe two citriculture systems (Campos-Herrera et al., 2013b).

S. diaprepesi was more effective than other native EPN species atprotecting citrus seedlings from D. abbreviatus in long-term green-house trials (El-Borai et al., 2007, 2012). The nematode is com-monly encountered on Florida’s central ridge eco-region whereweevil abundance is typically low, but not in coastal flatwoodseco-regions where D. abbreviatus are most abundant (Campos-Herrera et al., 2013a). Understanding the edaphic factors that affectthe abundance of S. diaprepesi might reveal ways in which soil can

be modified to enhance and conserve the biocontrol potential ofthe nematode. Because APS is detrimental to steinernematids(Campos-Herrera et al., 2013b), we conducted studies in the ongoingtrial of Schumann et al. (2013) to determine whether non-targeteffects of the two citriculture systems (APS and CC) on EPNs canaffect biological control of D. abbreviatus, citrus health and fruityield. In addition, we installed and evaluated the biotic and abioticeffects of landscape fabric mulch that enhances citrus growth andprevents weevil infestation of the soil beneath the tree canopy(McKenzie et al., 2001; Duncan et al., 2009). We hypothesized that(i) APS would continue to decrease the numbers of S. diaprepesiand, thereby, increase the numbers of D. abbreviatus emergingfrom the soil and Phytophthora sp. infecting roots and (ii) fabricmulch would decrease numbers of D. abbreviatus, thereby, increas-ing root mass density, tree size and fruit yield in both APS and CC.

2. Materials and methods

2.1. Field experiment design and treatments

The study were conducted in an ongoing experiment locatedin the central ridge eco-region (Auburndale, Florida, USA,81:48:44.65 W, 28:06:53.98 N, 49 m elevation above sea level).Four factorial treatments were established to compare the effectsof two different horticultural systems and the use of fabric mulchon both citrus growth and the soil food web. Treatments included:(i) the horticultural management systems conventional citriculture(CC) or advanced production system (APS) and (ii) the applicationof landscape fabric (LSF) or not (bare soil, BS) as a mulch on the soilsurface. The horticultural systems were established in spring 2008,each one replicated four times in plots containing four rows (3 m inrow � 6 m between rows) of 35–40 trees (Hamlin orange onSwingle citrumelo rootstock) in a randomized complete blockdesign. Details of the managements, crop history, map representa-tion and establishment were described by Schumann et al. (2012,2013) and Campos-Herrera et al. (2013b). Each plot (n = 4) wasdivided into 3 subareas, in each of which soil beneath 3 adjacenttrees was covered with landscape fabric (Lumite 994GC, wovenpolyester landscaping fabric, Synthetic Industries, Gainesville,GA) in September 2009. Two pieces of fabric (1.8 m � 9 m long)were stretched over the soil surface on each side of the three treesand secured with steel staples pounded into the soil. Slits cut intothe fabric allowed each piece to overlap the other by 30 cm at thetree line while tightly surrounding the tree trunks. The finaldimensions of the mulch were 3 m wide by 9 m long. Three treeplots immediately adjacent to the mulched trees were selectedfor the bare soil comparison. At each sampling time, the abioticand biotic variables measured in each of the three subareas wereaveraged (e.g., n = 4 plots per treatment).

2.2. Sampling methods and chemical analysis

Soil in each subarea was sampled five times during two years:July 27th 2011, December 7th 2011, March 27th 2012, May 29th2012 and August 21st 2012. On each date, the steel staples were

R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51 43

pulled from the soil on the east side of the trees to allow samplingbeneath the mulch, after which the fabric was reinstalled the sameday. Four cores of soil (2.5 cm diam. � 30 cm deep) were takenfrom the drip-irrigated zones beneath each of three trees (12 coresper sample) in the APS treatments. Soil samples were recoveredfrom the same relative location in the CC plots. Samples weretransported in insulated chests to the laboratory and processedthe same day (see following section).

Soil pH (1:1 soil–water suspension) was determined for eachsample and 50 g of oven dried soil (70 �C for 48 h) were alsoanalyzed for nutrients by Applus Agroambiental (S.A. Sidamón,Lleida, Spain). Soil variables included electric conductivity (EC)(1:5 soil–water suspension, dS m�1), organic matter (OM, %) by theWalkley–Black method (Walkley, 1935), nitrate (NO3, mg kg�1),available phosphorous (P2O5, mg kg�1) by Olsen’s extractionmethod, and K, Ca, Mg, and Na content by the ammonium acetatemethod (mg kg�1). Soil moisture was determined gravimetricallyby drying (70 �C for 48 h) and reweighing 180 g fresh soil.

2.3. Tree performance and rhizosphere communities

Treatment effects on citrus trees were assessed by measuring (i)the trunk diameter on four occasions between October 2010 andAugust 2012, following Duncan et al. (2009), (ii) the root mass den-sity during each sampling event (see description below), and (iii)the number of fruit per tree in November 2011 and 2013.

Emergence of adult weevils from soil was monitored using 14ground-traps per main plot (Duncan et al., 2001). The traps wereinstalled with at least five trees between traps in the same rowsused to study the effects of management regimes and mulchingon soil communities. Each trap consisted of a conical wire meshstructure (0.65 m2 surface at the base) topped by a boll weevil trap(Great Lakes IPM, Vestaburg, MI, USA). Traps were installed in May2011, and checked bi-weekly during spring, summer and autumnof 2011–2012.

Free-living nematodes (FLNs) and plant-parasitic nematodes(PPNs) were enumerated from Baermann funnels for all the

Table 1Target organisms identified and quantified by real time qPCR experiments.

Type of organism/species Material for standard curve Unit ofmeasurement

Citrus pathogenPhytophthora nicotianae Pure culture pg of DNA

Entomopathogenic nematodesHeterorhabditis indica Infective juveniles IJsHeterorhabditis zealandica Infective juveniles IJsSteinernema diaprepesi Infective juveniles IJsSteinernema riobrave Infective juveniles IJsSteinernema scapterisci Infective juveniles IJsSteinernema sp. glaseri-

groupInfective juveniles IJs

Competitor Free-living nematodesAcrobeloides-group Nematodes (different stages) ng of DNA

Nematophagous fungi*

Catenaria sp. ITS rDNA sequence + pDrive pg of DNAArthrobotrys dactyloides Pure culture pg of DNAGamsylella gephyropagum Pure culture pg of DNAHirsutella rhossiliensis Pure culture pg of DNAPaecilomyces lilacinus Pure culture pg of DNA

Ectoparasitic bacteriaPaenibacillus nematophilus 16S rDNA sequence of

490 bp + pUC57Paenibacillus sp. 16S rDNA sequence of

1515 bp + pDrive

* The ‘‘infection/infestation rates’’ was calculated for all the nematophagous fungi (NF)sample.

sampling events except the first one (July 2011). Aliquots of60 cm3 fresh soil were prepared for all the samples (n = 48) persampling event, incubated at room temperature (22–25 �C) for7 days, and recovered for counting under the dissecting micro-scope. Counts were expressed as number of nematodes per100 cm3 of soil. The citrus fibrous roots were recovered by rinsingthe entire 500 cm3 soil sample through a 1 mm sieve followingnematode extraction (Jenkins, 1964) (see description below). Rootswere dried and weighed.

For real time qPCR quantification, nematodes and associatedmicroorganisms were extracted from 500 cm3 soil by sucrose cen-trifugation (Jenkins, 1964). The organisms in the water suspensionwere allowed to settle in test tubes overnight at 4 �C. Thereafter,most of the water was aspirated and samples were transferred to1.5 mL Eppendorf tubes and stored at �20 �C until DNA extraction(Campos-Herrera et al., 2011a). Six EPN species, five nematopha-gous fungi (NF) species, free-living bacterivorous nematodes(FLBNs) from the Acrobeloides-group, two ectoparasitic bacterialspecies Paenibacillus spp. and a citrus pathogen P. nicotianae weretargeted (Table 1). All of the species are commonly encounteredin Florida citrus groves (Graham et al., 2003; Duncan et al., 2003,2007; Campos-Herrera et al., 2011a, 2011b, 2012, 2013a; Pathaket al., 2012). Morphological, morphometric and molecularidentifications were performed to confirm the identities of all theorganisms used to produce the standard curves (El-Borai et al.,2005; Nguyen, 2007; Campos-Herrera et al., 2011a, 2011b;Pathak et al., 2012). Maintenance of cultures and development ofstandard curves were performed according to Campos-Herreraet al. (2011a, 2011b, 2012, 2013a) and Pathak et al. (2012). TheUltraClean™ Soil DNA Extraction Kit (MoBio) was employed toextract the DNA in each nematode sample and the quality andquantity of each sample was assessed in duplicate using theNanodrop System 1000 v.3.3.0 (Thermo Scientific, Wilmington,DE, USA). Aliquots of the original DNA were stored at �80 �C untilanalysis. For each organism, species-specific primers and probe,protocols and concentrations, annealing temperature and numberof cycles were adjusted following Atkins et al. (2005), Zhang

Source Reference for primers/probe sequence,protocols

J.H. Graham and D. Bright Hung et al. (2010)

Authors Campos-Herrera et al. (2011a)Authors Campos-Herrera et al. (2011a)Authors Campos-Herrera et al. (2011a)Authors Campos-Herrera et al. (2011a)Authors Campos-Herrera et al. (2011b)Authors Campos-Herrera et al. (2011a)

Authors Campos-Herrera et al. (2012)

Authors Pathak et al. (2012)Authors Pathak et al. (2012)Authors Pathak et al. (2012)R.A. Humber Zhang et al. (2006)R.A. Humber Atkins et al. (2005)

GenBank: AY480936 Campos-Herrera et al. (2011b)

Authors, Genbank:JF317562

Campos-Herrera et al. (2011b)

by dividing the DNA quantity of each NF species by the total amount of DNA in a

44 R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51

et al. (2006), Huang et al. (2010), Campos-Herrera et al. (2011a,2011b, 2012, 2013a) and (Pathak et al., 2012). In all the runs,negative controls (sterile de-ionized water) and positive controls(the corresponding completed standard curve) were included, withall the samples (unknown and control) run in duplicate. Post-rundata analysis was performed as described by Campos-Herreraet al. (2011a, 2012).

2.4. Ecological indices and statistical analysis

Data for soil properties and organisms were transformed beforestatistical analysis. Numbers of EPNs, FLNs and Paenibacillus sp.copies were log(X + 1) transformed. NF parasitism of nematodeswas estimated by dividing the DNA quantity of each NF speciesby the total amount of DNA in a sample and then transformed tosquare root (Campos-Herrera et al., 2012; Duncan et al., 2013).P. nicotianae DNA quantity was divided by the quantity of citrusfibrous roots to estimate an infection rate before transformationto square root for analysis. Similarly, PPNs were divided by the rootmass density to establish the number of nematodes per g of rootprior to transformation (log(X + 1)). In the case of NF, to compareinfection rates between all species and to estimate the total NFinfection rate in a sample, the units of measurement betweenspecies were standardized (0–1) by dividing all data within a

LSF BS LSF BS

APS CC

0

20

40

60

80

100

LSF BS

APS

NO

3(m

g.K

g-1 )

0

2

4

6

8

LSF BS LSF BS

APS CC

0

10

20

30

40

LSF BS

APS

0

10

20

30

40

LSF BS

APS

(a)

(c)

(b)

(d)

(f)

(e)

P2O

5(m

g.K

g-1 )

K (m

g.K

g-1 )

LSF BS LSF BS

APS CC

pH

2.0

1.5

1.0

0.5

0.0

0.20

0.15

0.10

0.05

0.00

EC

(dS

.m-1

)O

M (%

)

Fig. 1. Soil properties measured in conventional citriculture (CC) and an advanced produ(BS). (a) pH; (b) organic matter (OM, %); (c) electrical conductivity (EC, dS m�1); (d) nitrcalcium (Ca, mg kg�1); (h) magnesium (Mg, mg kg�1); (i) sodium (Na, mg kg�1). Values

species by the highest measurement for that species (De Rooij-van der Goes et al., 1995). The ecological indices calculated were(i) species richness (number of species, S) and (ii) the Shannon–Wiener diversity index (H0 =

Ppi ln pi, where pi is the quantity of

the ith-species as a proportion of the entire community).Treatment effects were determined by a split plot, repeated

measurements ANOVA (SAS 9.3) in which mulch treatments weresubplots and cultural systems were whole plots. The whole plotdesign was randomized complete block (4 blocks) with the meansquare error for the block � cultural method interaction used asthe error term in the model. Spearman’s correlation coefficient(r) was used to measure the strength of relationships betweenthe cumulative abundance during the course of the experimentof weevils and EPNs detected in the whole plots. All data arepresented as mean ± SEM of untransformed values.

3. Results

3.1. Soil properties

Cultural system affected all measured physical and chemicalsoil properties, with the exception of OM and Mg (Fig. 1). Soil pHand levels of Ca and Na were significantly higher in APS, whereasN, P and K were higher in CC. All of the variables except Na were

0

5

10

15

20

LSF BS LSF BS

APS CC

0

20

40

60

80

LSF BS LSF BS

APS CC

0

100

200

300

400

500

LSF BS LSF BS

APS CC

LSF BS

CC

LSF BS

CC

LSF BS

CC

(g)

(i)

(h)

Ca

(mg.

Kg-

1 )M

g (m

g.K

g-1 )

Na

(mg.

Kg-

1 )

ction system (APS) in combination with a landscape fabric mulch (LSF) or bare soilogen (NO3, mg kg�1); (e) phosphate (P2O5, mg kg�1); (f) potassium (K, mg kg�1); (g)statistically significant (P < 0.05) highlighted in bold, and no significative (n.s.).

R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51 45

more abundant in bare soil than under mulch (Fig. 1i). Cultural sys-tem interacted with mulch in the case of N (Fig. 1b), where thatnutrient differed between bare and mulched soil more in CC thanin APS.

Soil in the APS irrigated zone was wetter than that under CC.The bare soil treatment was on average 38% wetter than mulchedsoil under APS (8.5% vs 6.2%) and 12% wetter than mulched soilunder CC (5.0% vs 4.5%).

3.2. Citrus performance

Twenty one months after planting trees and 1 year after install-ing the fabric mulch, the trunk cross-sectional area of trees grownin CC was 67% of that for trees under APS (829 mm3 vs 1228 mm3;P < 0.001) (Fig. 2a). However, the difference in tree size betweenthese treatments decreased with time until 33 months post mulch-ing when CC trees had 87% the girth of APS trees (3057 mm3 vs3515 mm3). Mulch did not affect tree size significantly (P < 0.05)until 2 years after fabric installation when mulched trees were11% larger than trees in bare soil. The beneficial effects of mulchingwere most apparent in comparison with the beneficial effects ofAPS. At 2 years post mulching, the girth of APS trees in bare soilwas 24% larger than CC trees in bare soil and 10% larger than CC

500

1000

1500

2000

2500

3000

3500

4000

APS LSFAPS bareCC LSFCC bare

Month

S O N D J F M AM J J A S O N D J F M AM J J A0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

2010 2011 2012

Trun

k cr

oss

sect

iona

l are

a (m

m2 )

(a)

(b)

Fibr

ous

root

s (g

. 50

0 cm

-3 s

oil)

Fig. 2. Citrus growth in conventional citriculture (CC) and an advanced productionsystem (APS) in combination with a landscape fabric mulch (LSF) or bare soil (BS).(a) Trunk cross sectional area (mm2); (b) fibrous roots (g. 500 cm�3 soil).

trees in mulched soil. Three years post mulching, the trunks ofAPS trees in bare soil were 14% and 6% larger compared to CC treesin bare and mulched soil, respectively.

Two years after the mulch was installed, the fibrous root massdensity in CC plots, irrespective of mulch, was 44% of that in APSplots (Fig. 2b). However, the root mass in all treatments declinedthroughout the experiment and differences between these twotreatments eventually disappeared. Initially, the trees in bare soilhad more roots than those under mulch, regardless of managementregime (P < 0.01). The effect of mulch on root mass declined withdecreasing root mass until it was no longer evident during the finaltwo sampling events (data not shown).

The number of fruit and the weight of fruit in 2011 were unaf-fected by the fabric mulch, but were both 47% greater in APS plotsthan in CC plots (P < 0001; Fig. 3). In 2013, the number of fruit inmulched plots was 13% greater (P = 0.05) than in plots with baresoil, because there were 8% fewer surviving trees (2.75 vs 3.0;P = 0.01) in bare soil than in mulched soil. Among surviving trees,the number of fruit per tree was not affected by cultural systemor by mulching in 2013, nor did the cultural system affect thenumbers of fruit per plot or tree survival rates.

3.3. Primary consumers

The Diaprepes–Phytophthora pest–disease complex was affectedby the citriculture systems. More adult Artipus floridanus weevilsthan D. abbreviatus weevils were captured in ground traps duringthe course of the experiment (Fig. 4a and b). The cumulativenumber of D. abbreviatus captured in APS plots was twice that inCC plots (P = 0.05). Forty-nine percent more A. floridanus were cap-tured in APS compared to CC, but the difference was not significant.P. nicotianae expressed per unit soil or per gram of fibrous rootaveraged 26-fold higher in APS than in CC treatments, regardlessof whether LSF was present or not (Fig. 4c).

The total number of PPNs were significantly higher (P = 0.05) inAPS than in CC (data not shown) and the total numbers of PPN pergram of root also tended to be more abundant in the APS treatment(P < 0.10). Individually, neither absolute numbers nor numbers per

BS

Num

ber f

ruit

per p

lot

0

50

100

150

200CoveredBare

Cultural system

APS CC0

500

1000

1500

2000

2011

2013

LSF

Fig. 3. Citrus yield in 2011 and 2013 in conventional citriculture (CC) and anadvanced production system (APS) in combination with a landscape fabric mulch(LSF) or bare soil (BS).

Phyt

opht

hora

nic

otia

nae

(ng

of D

NA

. g ro

ot)

P = 0.05

P = n.s.

P = 0.02

CCAPS

CCAPS

CCAPS

(a)

(b)

(c)

Fig. 4. Abundance of primary citrus consumers in conventional citriculture (CC)and advanced production system (APS). (a) Diaprepes abbreviatus adult emergence;(b) Artipus floridensis adult emergence; (c) Phytophthora nicotianae (ng DNA. g root).Values statistically significant (P < 0.05) highlighted in bold and no significative(n.s.).

46 R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51

gram of root of Xiphinema vulgare nor Belonolaimus spp. wereaffected significantly by any treatment.

3.4. Higher order trophic levels

Four of the six targeted EPN species were detected in all thesampling events: two natives, S. diaprepesi and H. indica and twoexotics, S. riobrave (introduced for weevil management in spring2009) and Steinernema scapterisci. Heterorhabditis zealandica wasonly detected in 3 blocks under CC management in May 2012and Steinernema sp. glaseri-group was not detected in this study.Only the four EPN species detected at every sampling time wereconsidered for ANOVA.

Steinernematid EPNs as a group were more numerous in CCthan in APS (P = 0.04, Fig. 5). In contrast, six times as many H. indicaIJs were detected in APS as in CC, although the trend was notsignificant. The number of detected IJs of every steinernematidspecies was greater in CC than in APS plots on every sampling date;however, this trend was significant (P = 0.06) only for S. scapterisci.Steinernematids also comprised >99% of all of the EPNs detected inCC plots (except in May 2012 when they comprised 86% of all

EPNs) and they dominated the EPN communities in the APS plots,averaging 73% of those populations during the experiment (datanot shown). During the course of the experiment, S. diaprepesi, S.riobrave and S. scapterisci accounted for 58%, 31% and 11% of thesteinernematids detected.

Mulching the soil surface with landscape fabric reduced thenumbers of S. scapterisci and H. indica (P 6 0.05, Fig. 5c and d),but did not measurably affect the other EPN species (Fig. 5a andb). Interactions between the treatments occurred for S. scapterisci,where the differences in IJ numbers between bare and mulchedplots were greater in CC than in APS.

The NF Paecilomyces lilacinus, Arthrobotrys dactyloides andCatenaria sp. were detected at all the sampling times and were,therefore, subjected to ANOVA. Gamsylella gephyropaga was neverdetected and Hirsutella rhossiliensis was detected in just 2 CC plotsin March 2011 and one ACP plot in May 2011. The experimentaltreatments had significant effects on only P. lilacinus and Catenariasp. (Fig. 6). Recovery of both species as well as the total NF werereduced by the use of the fabric mulch (P < 0.05), whereas no NFspecies was significantly affected by the management regime(Fig. 6).

The ectoparasitic bacterium Paenibacillus sp. was detected at allthe sampling events, whereas P. nematophilus was only detected inbare soil in one ACP plot in August 2012. Paenibacillus sp. was four-fold more abundant in bare soil as under mulch (P = 0.007, Fig. 7a),but was not significantly affected by management regime. TheAcrobeloides-group nematodes that sometimes compete with EPNsfor the cadaver were detected in all plots during each samplingevent. These nematodes were most abundant in the CC plots(P = 0.002, Fig. 7b). The management regime interacted withground cover because 55% more Acrobeloides-group nematodesoccurred in the bare soil than under mulch (P = 0.02) in the CCplots, whereas mulch did not affect these nematodes in the APSplots. The FLNs as a group were more numerous in bare soil(P = 0.009, Fig. 7c), but were not clearly affected by managementsystem.

3.5. Ecological indices and species relationships

Communities of both EPNs and NF were richer and more diversein bare soil than under fabric mulch (Fig. 8). These characteristicsof EPN and NF communities were unaffected by the managementregime.

No species of EPN was significantly related to the numbers ofadult D. abbreviatus captured in ground traps, but the totalnumbers of EPNs detected in each plot were inversely related tothese weevils (r = �0.93, n = 8 P = 0.004; Fig. 9). The numbers ofD. abbreviatus were also inversely related to numbers of Acrobeloides-group nematodes (r = �0.90, P = 0.001). The later relationship maybe because the total number per plot of Acrobeloides-group nema-todes that sometimes compete with EPNs were positively relatedto the number of EPNs in these plots (r = 0.76, P = 0.03). Numbersof A. floridanus were inversely related to S. diaprepesi (r = �0.76,P = 0.03), but unrelated to any other nematode species orassemblage of species.

4. Discussion

The advanced production system evaluated in this studyaffected EPNs in ways that may have reduced the biological controlof D. abbreviatus and P. nicotianae, thereby increasing the severityof this pest–disease complex compared to that in citrus grownconventionally. The EPN responses to both citriculture systemswere consistent with their behavior at this site in 2009–2010(Campos-Herrera et al., 2013b). S. diaprepesi was the only naturallyoccurring EPN species detected above trace levels in the previous

0

10

20

30

40

50

LSF BS LSF BS

APS CC

0

10

20

30

40

LSF BS LSF BS

APS CC

0

10

20

30

40

LSF BS LSF BS

APS CC

0

10

20

30

40

LSF BS LSF BS

APS CC

(a)

(c)

(b)

(d)

(f)

(e)

S. d

iapr

epes

i IJs

. 50

0 cm

-3so

il

0

10

20

30

40

LSF BS LSF BS

APS CC

S. ri

obra

ve IJ

s . 5

00 c

m-3

soil

S. s

capt

eris

ci IJ

s . 5

00 c

m-3

soil

0

10

20

30

40

50

LSF BS LSF BS

APS CC

H. i

ndic

a IJ

. 50

0 cm

-3so

ilSt

eine

rnem

atid

sIJ

s . 5

00 c

m-3

soil

Het

eror

habd

itids

IJs

. 500

cm

-3so

il

Fig. 5. Abundance of entomopathogenic nematodes (IJs. 500 cm�3 soil) in conventional citriculture (CC) and an advanced production system (APS) in combination with alandscape fabric mulch (LSF) or bare soil (BS). (a) Steinernema diaprepesi; (b) S. riobrave; (c) S. scapterisci; (d) Heterorhabditis indica; (e) steinernematids; (f) heterorhabditids.Values statistically significant (P < 0.05) highlighted in bold, marginally significant in italics (P < 0.1) and no significative (n.s.).

R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51 47

study and was significantly more abundant in CC than in APS plotsat that time. Campos-Herrera et al. (2013b) also showed that thepersistence of augmented S. riobrave was higher in CC than inAPS, whereas H. indica persisted equally well in either system.The natural prevalence of both S. riobrave and H. indica increasedat this site since the previous study. The abundance of S. riobravein the current study was similar to that of S. diaprepesi andalthough these two species dominated the EPN community in bothcitriculture systems, they were more than twice as abundant in CCon every sampling occasion. By contrast, the APS treatment tendedto favor H. indica, either directly and/or by creating a niche withfewer steinernematid competitors. These three species are all

pathogens of D. abbreviatus, but their relative effectiveness againstthe weevil is reported to be S. diaprepesi > S. riobrave > H. indica(El-Borai et al., 2007, 2012; Duncan et al., 2013). The use ofcommercially formulated EPNs in citrus orchards to reduceD. abbreviatus injury to the root cortex can also reduce P. nicotianaelevels in soil (Duncan et al., 2010). Therefore, the higherabundance of D. abbreviatus and P. nicotianae in APS could haveresulted in part from fewer steinernematids in APS than in CC.

Because the two citriculture systems affected H. indica and thesteinernematid species differently, it is likely that soil propertiescan be manipulated through agricultural practices to conserveparticular EPN species. The profound differences between APS

(a)

(b)

(c)

(d)

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

1e-3

1e-4

1e-5

1e-6

1e-7

1e-8

1e-9

P. li

laci

nus

infe

ctio

n ra

te .

500

cm-3

soil

1e-3

1e-4

1e-5

1e-6

1e-7

1e-8

1e-9

1e-3

1e-4

1e-5

1e-6

1e-7

1e-8

1e-9

Cat

enar

ia s

p.in

fect

ion

rate

. 5

00 c

m-3

soil

A. d

acty

loid

es in

fect

ion

rate

. 50

0 cm

-3so

il

0.6

0.4

0.2

0

Tota

l NF

norm

aliz

ed .

500

cm-3

soil

Fig. 6. Infection rate of nematophagous fungi (NF) (per 500 cm3 soil) in conventional citriculture (CC) and an advanced production system (APS) in combination with alandscape fabric mulch (LSF) or bare soil (BS). (a) Paecilomyces lilacinus; (b) Arthrobotrys dactyloides; (c) Catenaria sp.; (d) total nematophagous fungi (standardized). Valuesstatistically significant (P < 0.05) highlighted in bold, and no significative (n.s.).

48 R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51

and CC in most of the measured physical soil properties providemyriad possibilities for more highly controlled studies toward thisend. Campos-Herrera et al. (2013b) measured more DNA copies ofPaenibacillus sp. per IJ S. diaprepesi in APS and speculated thatgreater encumbrance by the bacterium reduced the fitness of thenematode in APS. The soil pH was higher in APS than in CC plotsand Paenibacillus sp. spores were subsequently found to detachfrom the S. diaprepesi cuticle at rates inversely related to the pHof aqueous solutions or soils in laboratory studies (El-Borai et al.,unpublished data). In the current study there was a non-significanttrend for fewer S. diaprepesi in APS plots, but there was no evidencethat encumbrance by the bacteria differed between treatments.Soil pH in APS and CC reported by Campos-Herrera et al. (2013b)was 7.2 and 5.6, respectively. In the current study, the averagepH remained low (5.4) in the CC treatment, but regular use ofdilute sulfuric acid to clean the irrigation drippers since 2011decreased pH to 6.5 in the APS treatment, which could havereduced the spore burden on S. diaprepesi in APS compared to theearlier study. This possibility is consistent with S. diaprepesiabundance per half liter soil in the two treatments which remainedsimilar for CC in this and the previous study (13.7 and 12.2 IJs,respectively), but which tripled in APS from 2.2 IJs previously to

6.5 IJs in the present study. Nevertheless, regardless of whethersoil pH caused the different ratios of bacteria to EPNs in the twostudies, factors other than Paenibacillus sp. were responsible forthe different treatment effects on EPNs in the current study. Someof the other differences between the soils in this study have beenassociated with EPN abundance elsewhere. The level of soil Pwas greater in CC than in APS throughout the current study, butnot during that of Campos-Herrera et al. (2013b). S. diaprepesiwas positively correlated with P content of soil in both this studyand in a geospatial survey of EPNs in Florida orchards (Campos-Herrera et al., 2013a). Surveys involving species other than S. dia-prepesi have measured inverse relationships between soil P andEPN abundance (Campos-Herrera et al., 2008; Hoy et al., 2008).The geospatial survey by Campos-Herrera et al. (2013a) alsorevealed associations between properties affecting soil moisture(depth to ground water, water holding capacity, clay and organicmatter content) and species composition of EPN communities.Wetter soils tended to support EPN communities heavily domi-nated by H. indica, whereas S. diaprepesi often dominatedcommunities in the drier soils that also contained significant num-bers of H. zealandica and H. indica. These associations were verymuch in line with the EPN species composition in this study where

0

1

2

3

LSF BS LSF BS

APS CC

0

10

20

30

40

LSF BS LSF BS

APS CC

(a)

(b)

Paen

ibac

illus

sp. c

opy

num

bers

. 50

0 cm

-3so

ilAc

robe

loid

es-g

roup

ng

DN

A . 5

00 c

m-3

soil

0

200

400

600

800

LSF BS LSF BS

APS CC

(c)

Num

ber o

f FLN

s . 1

00 c

m-3

soil

Fig. 7. Abundance of some natural enemies of entomopathogenic nematodes inconventional citriculture (CC) and an advanced production system (APS) incombination with a landscape fabric mulch (LSF) or bare soil (BS). (a) Paenibacillussp. (No. copies. 500 cm�3 soil); (b) Acrobeloides-group nematodes (ng DNA.500 cm�3 soil); (c) total number of free-living nematodes (No. nematodes.100 cm�3 soil). Values statistically significant (P < 0.05) highlighted in bold,marginally significant in italics (P < 0.1), and no significative (n.s.).

R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51 49

the wetter APS soil tended to favor H. indica and the drier CC soilconsistently supported more S. diaprepesi as well as S. riobraveand S. scapterisci. Soil was also drier and H. indica fewer beneaththe fabric mulch in both cultural systems.

Mulching increased the size of trees in both APS and CC, andfabric mulch conferred about half as much additional tree growthas was achieved by using APS. However, the increased tree sizedid not produce a corresponding increase in fruit yield pertree, perhaps because mulching reduced the root mass density.

The increased tree size and survival in mulched plots suggests thatmulching had the intended effect of preventing the neonate weevillarvae from entering the soil. In a previous study, the installation oflandscape fabric as a mulch beneath mature citrus trees reducedthe emergence from soil of D. abbreviatus by 99% compared tonon-mulched trees (Duncan et al., 2009). Nevertheless, the treesin the present study were not destructively sampled and thereforeit is not known whether reduced root herbivory is one of the ben-efits that mulch provided the trees. Indeed, mulching did notreduce the levels of P. nicotianae which might be expected if thefabric reduced weevil herbivory. Fabric was not installed until9 months after planting and effects might be greater if trees wereplanted into mulched soil. Mulching had the additional benefit ofeliminating the considerable expense of herbicide applications.

It is interesting that while APS greatly increased the fibrous rootdensity compared to that in CC, fabric mulch decreased the fibrousroots in both cultural systems, despite growing larger trees.Therefore, the efficiency of roots beneath fabric mulch was muchgreater than that of roots in bare soil. The extent to which thegeneral decline in root density over time was due to the effectsof huanglongbing, weevil herbivory, and/or to repeated samplingis unknown. Huanglongbing reduces fibrous root density by asmuch as 40% prior to symptom expression in the tree canopy(Graham et al., 2013). Visual symptoms of huanglongbing in thisorchard increased from 10% to 40% of trees between January2012 and August 2013; therefore, given the latency between infec-tion and symptoms, the incidence of the disease was likely veryhigh during this trial (Schumann et al., 2013). Regardless, the den-sity of roots likely accounted for many of the treatment effects onsoil organisms. For example, the abundance of plant parasiticnematodes, free living nematodes, the Acrobeloides-group nema-todes and some of the natural enemies of nematodes such asP. lilacinus and Catenaria sp. were highest in treatments with thegreatest root mass density, where resources for root parasites aremost numerous and where root exudates support the microbialfood base for free living nematodes. As reported in other studies,the Acrobeloides-group nematodes in each of the main plots werepositively related to EPNs, supporting the possibility that these freeliving nematodes compete with EPNs for resources in insectcadavers (Campos-Herrera et al., 2012, 2013a).

Despite causing non-target effects on some natural enemies ofD. abbreviatus, APS produced higher fruit yields in a much shortertime than can be achieved using conventional citriculture practices(Schumann et al., 2013). Current projections show that orchardscan become profitable in seven years using APS, compared to10–12 years if managed conventionally. To date, the APS treat-ments with the highest densities of trees have produced twice asmuch fruit using half as much water and substantially less fertil-izer than the conventional system. Nevertheless, it is unknownwhether an increased severity of the Diaprepes–Phytophthoracomplex in APS affected the growth and production of the treesin that treatment. The increased growth of APS trees in responseto fabric mulch weevil barriers demonstrated the possibility thatnon-target effects on steinernematids may have reduced thebenefit of APS. Moreover, tree decline caused by weevils in thistrial could become more pronounced in APS than in CC in futureyears because the cortical damage to major roots caused by weevilfeeding does not heal and is cumulative. Aboveground declinesymptoms are not apparent until the loss of root cortex reaches athreshold and the tree age at which this occurs depends in parton the local EPN community (Duncan et al., 2013). Because APSis being widely adopted as a tool to manage citrus greeningdisease, it is important to understand the mechanisms by whichAPS affected the various EPN species in this orchard. Withthat knowledge it should be possible to (i) either modify APStechnology to mitigate the problem or implement pest–disease

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

LSF BS LSF BS

APS CC

(c)

(d)(b)

(a)4

3

2

1

0

4

3

2

1

0

0.8

0.6

0.2

0.0

0.4

0.8

0.6

0.2

0.0

0.4

Fig. 8. Ecological indices estimated for entomopathogenic nematodes (EPN) and nematophagous fungi (NF) in conventional citriculture (CC) and an advanced productionsystem (APS) in combination with a landscape fabric mulch (LSF) or bare soil (BS). (a) EPN richness S; (b) NF richness S; (c) EPN diversity H0; (d) NF diversity H0 . Valuesstatistically significant (P < 0.05) highlighted in bold, marginally significant in italics (P < 0.1), and no significative (n.s.).

Num

ber o

f adu

lt D

iapr

epes

(c

umul

ativ

e)

Number of IJ EPNs(mean per sample event)

02

4

6APSCC

8

10

12

14

16

18

20

10 10 30 40 50

Fig. 9. Relationship between the total number of adult Diaprepes abbreviatusemerging from soil and the average number of infective juvenile of entomopath-ogenic nematodes during 2011–2012.

50 R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51

management tactics proactively and (ii) discover ways in which toconserve various EPN species in this and other agricultural systems.

Acknowledgments

The authors thank John Strang for his collaboration and for pro-viding the site and many of the resources used for this study. We

appreciate the technical assistance of Joshua Fluty and TonyMcIntosh in the field and laboratory. Also, we would like to thankRobin J. Stuart for the initial work installing landscape fabric andground-traps and James H. Graham for providing laboratory spaceand equipment. This study is supported by a USDA–CSREES SpecialGrant (TSTAR) and U.S.–Egypt Science and Technology Joint Fund(338). R. Campos-Herrera was supported by a Marie CurieInternational Outgoing Fellowship within the 7th EuropeanCommunity Framework Programme (FP7-PEOPLE-2009-IOF-252980).

References

Adams, B.J., Fodor, A., Koppenhöfer, H.S., Stackenbrandt, E., Stock, S.P., Klein, M.G.,2006. Biodiversity and systematics of nematode–bacterium entomopathogens.Biol. Control 38, 4–21.

Atkins, S.D., Clark, I., Pande, S., Hirsch, P.R., Kerry, B.R., 2005. The use of real-timePCR and species-specific primers for the identification and monitoring ofPaecilomyces lilacinus. FEMS Microbiol. Ecol. 51, 257–264.

Campos-Herrera, R., Gómez-Ros, J.M., Escuer, M., Cuadra, L., Barrios, L., Gutiérrez, C.,2008. Diversity, occurrence, and life characteristics of naturalentomopathogenic nematode populations from La Rioja (Northern Spain)under different agricultural management and their relationships with soilfactors. Soil Biol. Biochem. 40, 1474–1484.

Campos-Herrera, R., Johnson, E.G., El-Borai, F.E., Stuart, R.J., Graham, J.H., Duncan,L.W., 2011a. Long-term stability of entomopathogenic nematode spatialpatterns measured by sentinel insects and real-time PCR assays. Ann. Appl.Biol. 158, 55–68.

Campos-Herrera, R., El-Borai, F.E., Stuart, R.J., Graham, J.H., Duncan, L.W., 2011b.Entomopathogenic nematodes, phoretic Paenibacillus spp., and the use of realtime quantitative PCR to explore soil food webs in Florida citrus groves.J. Invertebr. Pathol. 108, 30–39.

Campos-Herrera, R., El-Borai, F.E., Larry, W.D., 2012. Wide interguild relationshipsamong entomopathogenic and free-living nematodes in soil as measured byreal time qPCR. J. Invertebr. Pathol. 111, 126–135.

R. Campos-Herrera et al. / Biological Control 76 (2014) 41–51 51

Campos-Herrera, R., Pathak, E., El-Borai, F.E., Schumann, A., Abd-Elgawad, M.M.M.,Duncan, L.W., 2013a. New citriculture system suppresses native and augmentedentomopathogenic nematodes. Biol. Control 66, 183–194.

Campos-Herrera, R., Pathak, E., El-Borai, F.E., Stuart, R.J., Gutiérrez, C., Rodríguez-Martín, J.A., Graham, J.H., Duncan, L.W., 2013b. Geospatial patterns of soilproperties and the biological control potential of entomopathogenic nematodesin Florida citrus groves. Soil Biol. Biochem. 66, 163–174.

De Rooij-van der Goes, P.C.E.M., van der Putten, W.H., van Dijk, C., 1995. Analysis ofnematodes and soil-borne fungi from Ammophila arenaria (Marram grass) inDutch coastal foredunes by multivariate techniques. Eur. J. Plant Pathol. 101,149–162.

Dolinski, C., Choo, H.Y., Duncan, L.W., 2012. Grower acceptance ofentomopathogenic nematodes: case studies on three continents. J. Nematol.44, 226–235.

Duncan, L.W., McCoy, C.W., Stansly, P.A., Graham, J.H., Mizell, R.F., 2001. Estimatingthe relative abundance of citrus root weevils with modified Tedders traps.Environ. Entomol. 30, 939–946.

Duncan, L.W., Graham, J.H., Dunn, D.C., Zellers, J., McCoy, C.W., Nguyen, K., 2003.Incidence of endemic entomopathogenic nematodes following application ofSteinernema riobrave for control of Diaprepes abbreviatus. J. Nematol. 35, 178–186.

Duncan, L.W., Graham, J.H., Zellers, J., Bright, D., Dunn, D.C., El-Borai, F.E.,Porazinska, D.L., 2007. Food web responses to augmenting theentomopathogenic nematodes in bare and animal manure–mulched soil. J.Nematol. 39, 176–189.

Duncan, L.W., Stuart, R.J., Lapointe, S.L., 2009. Use of landscape fabric to manageDiaprepes root weevil in citrus groves. Fla. Entomol. 92, 74–79.

Duncan, L.W., Dewdney, M., Graham, J.H., 2010. Remember Diaprepes? It’s still aproblem. Citrus Ind. Mag. 91 (7), 10–14.

Duncan, L.W., Stuart, R.J., El-Borai, F.E., Campos-Herrera, R., Pathak, E., Graham, J.H.,2013. Modifying orchard planting sites conserves entomopathogenicnematodes, reduces weevil herbivory and increases citrus tree growth,survival and fruit yield. Biol. Control 64, 26–36.

El-Borai, F.E., Duncan, L.W., Preston, J.F., 2005. Bionomics of a phoretic associationbetween Paenibacillus sp. and the entomopathogenic nematode Steinernemadiaprepesi. J. Nematol. 37, 18–25.

El-Borai, F.E., Brentu, C.F., Duncan, L.W., 2007. Augmenting entomopathogenicnematodes in soil from a Florida citrus orchard: non-target effects of a trophiccascade. J. Nematol. 39, 203–210.

El-Borai, F.E., Stuart, R.J., Campos-Herrera, R., Pathak, E., Duncan, L.W., 2012.Entomopathogenic nematodes, root weevil larvae, and dynamic interactionsamong soil texture, plant growth, herbivory, and predation. J. Invertebr. Pathol.109, 134–142.

Enright, M.R., Griffin, C.T., 2005. Effects of Paenibacillus nematophilus on theentomopathogenic nematode Heterorhabditis megidis. J. Invertebr. Pathol. 88,40–48.

Futch, S.H., Duncan, L.W., Zekri, M., 2005. Validation of an area-wide extensionprogram to estimate the seasonal abundance of adult citrus root weevils withun-baited pyramidal traps. Proc. Fla. State Hort. Soc. 117, 143–147.

Georgis, R., Koppenhöfer, A.M., Lacey, L.A., Bélair, G., Duncan, L.W., Grewal, P.S.,Samish, M., Tan, L., Torr, P., van Tol, R.W.H.M., 2006. Successes and failures inthe use of parasitic nematodes for pest control. Biol. Control 38, 103–123.

Graham, J.H., Bright, D.B., McCoy, C.W., 2003. Phytophthora-Diaprepes weevilcomplex: Phytophthora spp. relationship with citrus rootstocks. Plant Dis. 87,85–90.

Graham, J.H., Johnson, E.G., Gottwald, T., Irey, M., 2013. Pre-symptomatic fibrousroot decline in citrus trees caused by Huanglongbing and potential interactionwith Phytophthora spp. Plant Dis. 97, 1195–1199.

Hoy, C.W., Grewal, P.S., Lawrence, J.L., Jagdale, G., Acosta, N., 2008. Canonicalcorrespondence analysis demonstrates unique soil conditions forentomopathogenic nematode species compared with other free-livingnematode species. Biol. Control 46, 371–379.

Huang, J., Li, C., Xiao, C., Wang, G., 2010. Detection of Phytophthora nicotianae in soilwith real-time quantitative PCR. J. Phytopathol. 158, 15–21.

Jenkins, W.R., 1964. A rapid centrifugal-flotation technique for separatingnematodes from soil. Plant Dis. Rep. 48, 492.

McKenzie, C.L., Lapointe, S.L., Duncan, L.W., 2001. Landscape fabric as a physicalbarrier to neonate Diaprepes abbreviatus (Coleoptera: Curculionidae). Fla.Entomol. 84, 721–722.

Nguyen, K.B., 2007. Methodology, morphology and identification. In: Nguyen, K.B.,Hunt, D.J. (Eds.), Entomopathogenic Nematodes: Systematics, Phylogeny andBacterial Symbionts, Nematology Monographs and Perspectives, vol. 5. Brill,Leiden–Boston, pp. 59–119.

Pathak, E., El-Borai, F.E., Campos-Herrera, R., Jonhson, E.G., Stuart, R.J., Graham, J.H.,Duncan, L.W., 2012. Use of real-time PCR to discriminate predatory andsaprophagous behavior by nematophagous fungi. Fungal Biol. 116, 563–573.

Schumann, A., Hostler, K., Waldo, L., Mann, K., 2012. Advanced production systemsfor Florida citrus: research update, huanglongbing impacts and productionforecasts. Citrus Ind. 93, 6–12.

Schumann, A., Hostler, K., Waldo, L., Rayburn, N., 2013. Update on advanced citrusproduction system trials. Citrus Ind. 94 (8), 6–12.

Walkley, A., 1935. An examination of methods for determining organic carbon andnitrogen in soils. J. Agric. Sci. 25, 598–609.

Zhang, L.M., Liu, X.Z., Zhu, S.F., Chen, S.Y., 2006. Detection of the nematophagousfungus Hirsutella rhossiliensis in soil by real-time PCR and parasitism bioassay.Biol. Control 36, 316–323.