Ptant Pathology (1991) 40, 445 -456

Interactions between Fusarium oxysporum f. sp. tracheiphilumand Meloidogyne spp. in Vigna unguiculata. 1. Effects of differentinoculum densities on Fusarium wilt

A. R. HARRIS* and H, FERRISDepartment of Nematology. University of California. Davis 95616. iSA

The effects of chlamydospores and conidia of Fusarium oxysporum I sp, Ircuheiphilum at different initialspore concentrations were compared in the wilt-susceptible cowpea cultivar California Blackeyc No, 5(CB5), In glasshouse experiments with one inoculum density of either \fcl(ii<l(igyne ineognita or \fjavanica, chlamydospores resulted in greater ineidence and severity of Fusarium vsilt than conidia at thesame inoculum densities. Wilt symptoms also increased on wilt-resistant cultivar r B 3 as inoculumdensities of M. javaniea were increased. When three cultivars were infested with moderate or highdensities of both F. o. traeheiphilwn and M.javanua. only CB5 developed severe wilt at either inoculumdensity. The wilt-tolerant cultivar Grant had mild uilt symptoms in most plants at moderate inoculumdensities, and a tenfold increase in inoculum did not increase wilt ratings, CB3, however, had higherincidence and severity of Fusarium wilt symptoms at high inoculum densities, although 60% of theplants survived for 9 weeks.

I N T R O D U C T I O N

Interactions between \teloidogyne spp, andFusarium oxysporum have been reported in mnn\crops, and the literature on this subject v\asreviewed recently (Mai & Abawi, 1987), Mostpapers on these interactions reported the use ofinoculum at either unspecified concentrations ordensities that were unnaturally high. Roots com-monly were wounded, either by cutting, trans-planting, or making inoculation holes in the soil.Further, conidia generally were used as fungalinoculum, although chlamydospores are thenatural survival structures and primary inoculumof F. oxysporum (Nelson, 1981), Studies of theeffects of different inoculum densities havedemonstrated an increase in susceptibility toFusarium wilt at moderate or high levels ofMeloidogyne spp, and moderate levels of F.oxysporum in watermelon (Sumner & Johnson.1973) and cotton (Garber et ai, 1979; Starr et ai,1989), Levelsof inocula of A/, inc(>f;niiii up lo 10"juveniles per plant had no effect on Fusarium wiltsymptoms or on infection of tomato plants by F.oxysporum (Jones et ai, 1976; Sidhu & Webster.1981), Higher nematode inoculum levels

'Present address: CSIRO Division of Soils, CflenOsmond, South Australia 5064,

increased chlorosis and propagule count ofFusarium in two tomato eultivars resislanl to F.oxy^parum (Sidhu & Webster, I'-'M ), hut did notreduce the monogenic resistance in lour othercultivars (Abawi & Barker, 1^X4),

M. javanica has twice been reported to enhanceFusarium wilt symptoms in ccivvpeas (\ignatotgtiieulaia). Thomason ei ai (14SM) reportedthat the presence of \t iuianicn increased \vleninecrosis in stems of the wilt-tolerant cultivarGrant and the vvilt-susceptible cultivar Chino ?when grown in pots containing loam infested withF. oxysporum f. sp, traeheiphihtm. hut thev did notquantify their inocula, Swanson (l'J,^4i foundthat the addition of \t. iavanica to sand infestedwith eonidia of f o. truehciphitum increasedvascular discoloration at the priniarv node inboth the wilt-resistant cultivar California Black-eye No, 3 (CB3) and the uilt-susceptihlc cultivarCalifornia Blackcye No. 5 (CB5), In CB3, vascu-lar discoloration increased with increasing inocu-lum densities of both eggs of \t. Javuniea andconidia of F. <i. tnuheiphituni. Hovsevcr, v.isculardiscoloration was slight e\cept at high inoculumconcentrations 15 * 10 conidia ,ind over 2 5 v UVnematode eggs per plant). Swanson concludedthat a predisposing interaction is possible in (, B3in the field, Further, Thomason i7 .// (19.'̂ 9)demonstrated that soil lumigation in covvpca

446 A.R. Harris and H. Ferris

fields reduced root-knot and Fusarium wilt in theFi«a;7Mw-tolerant cultivar Grant,

Before future studies could be conducted onthe nature of reported interactions between F. o.traeheiphilum and Meloidogyne spp, on cowpeas.it was necessary to determine realistic inoculumdensities and inoculation methods that wouldgive moderate levels of disease with low variabi-lity. The objectives of these investigations were: tocompare conidia and chlamydospores of F. o.traeheiphilum as inoculum; to develop techniquesfor inoculating cowpea with Meloidogyne spp,and chlamydospores of Fusarium withoutwounding roots; to determine whether Meloido-gyne spp, can increase susceptibility to F. <>.traeheiphilum at realistic inoculum densities ofboth the nematode and fungus; and to determinethe identity and density of F, o. traeheiphilum andMeloidogyne spp, in a field of Fusarium wilt-resistant cowpea that showed Fusarium wiltsymptoms. Part of this research has been reportedbriefly (Harris & Ferris, 1988a),

MATERIALS AND METHODSAll experiments were conducted in a glasshouse(24 + 5 C) or lahoratory at the University ofCalifornia, Davis, from 1986 to 1989, The follow-ing materials and methods were common to allglasshouse experiments, Cowpea seeds were sur-facc-disinfested in 0 5"/,, NaOCl for 10 min. thenrinsed by soaking in several changes of steriledistilled water for 10 min. Seeds were germinatedhy placing in a humid chamber for 1 -2 days, thenseedlings of similar size were selected and one wassown in aerated-steam-treated sand in each pot.The sand was maintained at 27±5 C by keepingpots in either constant-temperature water bathsor environmental growth chambers. Each plantwas fertilized with half-strength Hoagland's solu-tion (Hoagland & Arnon, 1950), generally onalternate days.

The isolate of F. o traeheiphilwn (160 si 28)used in all experiments was race 3, identified usingrace differential cowpea cultivars (Harris, 1989),This isolate was collected from a cowpea field inTulare County in 1985, Chlamydospores wereproduced hy adding a suspension from cultureson potato-dextrose agar (PDA) to XO g ofchopped, sterile, cowpea straw plus 500 ml ofnutrient solution (5 mM KNO,-(-0 5'".. glucose) in2 8-1 flasks. After 2 weeks, infested stravv wastransferred aseptically into sterile paper bags andair-dried for 2 weeks The dry straw was milled(Wiley mill, 40 mesh [0 42-iiim| screen) and usedas inoculum by mixing with dry aerated-steam-

treated sand in a sterilized concrete mixer. Inocu-lum was quantified by dilution plating the strawon yeast-dextrose agar medium (YDA) (Nelsonet ai, 1986) in petri dishes and counting resultantcolonies.

Plants were assessed for Fusarium wilt inci-dence and severity with two rating systems. Fordisease rating, wilt symptoms were scored on ascale of 0-4, where 0 = healthy; 1= stunted;2 = stunted, swollen basal stem, primary leafchlorosis, one primary leaf may have fallen, andvascular discoloration; 3 = stunted, swollen basalstem, both primary leaves fallen, and vasculardiscoloration; 4 = dead. This disease rating meth-od was modified from Rigert & Foster (1987),Vascular discoloration in cross-sections of theprimary node (first unifoliolate node) was scoredon a scale pre-transformed to arcsines, where0 = 0-9, 1 = 10-34, 2 = 35-64, 3 = 65-89. 4 = 90-99 and 5 = 100% of vessels discoloured (modifiedfrom Swanson. 1984), Seedlings with diseaseratings of I do not develop other disease symp-toms and usually recover from stunting (Rigert &Foster. 1987), Disease ratings of 0 or 1. orvascular discoloration scores of 0-2, indicatedresistance; higher ratings indicated susceptibility,A few seedlings died or failed to develop due togenetic abnormalities, so the number of plantssurviving or wilted by the end of each experimentwas counted.

Effect of inoculum type and density of F. o,tracheiphilum and of inoculation method

Cowpea cultivar CB5 was inoculated with F. o.traeheiphilum. to which it is susceptible, in twoexperiments, Conidial inocula were prepared byplacing small pieces of dried fungus, from single-spore cultures stored at 4 C, onto PDA in petridishes. Wild-type colonies were subcultured ontoPDA slants. Wild-type cultures were selectedagain, washed with sterile distilled water, andfiltered through cheese cloth, Conidial concentra-tions were adjusted with the aid of a haemacyt-ometer.

In the first experiment, conidial suspensions ordried chlamydospore-infestcd straw were mixedwith sand in a sterilized concrete mixer. Calcu-lated initial spore concentrations and controltreatments arc listed in Table 1, Infested sand foreach treatment vsas placed in 500<ni' (lO-cm-diameter) plastic pots approximately .̂ 75 cm' ofsand per pot), which (hen wcrc immersed in thesame type of sand (uninfested) in plastic tubs.Each tub contained three pots of the same

Inoculum densities o/Fusarium oxysporum and Meloidogyne in eowpea 447

Table 1. Recovery of Fusarium oxvsporum f, sp, tracheiphilum from sterilized, air-dry sand in pots at 27 C 19 daysafter inoeulation, and stained Meloidogyne incognita in cowpea cultivar CB5 roots 5 weeks after inoculation with 10'

eggs per plant

Type ofinoculum andnumber of sporesper pot

Initial concentration of Logiij(,V-t-1) c,fu,,gF. o. tracheiphilum (initial concentration of dry sand Number of mature

(spores g of of spores, g of 19 days after \t incognitaair-dry sand) air-dry sand) inoculation per plant

Control (no F. o. tracheiphilum)

Conidia:

10'10'

Chlamydospores:

No M. incognita:Conidia, 10'Chlamydospores, 10'

0

13130

131013070

262602610

13102610

0000

1 3222 30333014301

1 6132 6033 602

3 3013-602

0

1015100

11025(1

4050

53

2127?139

232331

00

treatment and had no drainage holes. The tubswere immersed in constant-temperature waterhaths. Each tub (experimental unit) of the 10treatments was replicated five times in a com-pletely randomized design. Pots were left withoutfertilizing or watering for 2 weeks hefore planting,to allow fungus populations to stabilize and formresting structures. To assess survival of spores,sand was sampled from each pot 19 days afterinfestation with F. o. traeheiphilum. Samples werecollected at a depth of 4 5 cm with an 11-mm-diameter cork borer. Sand from the three pots ineach tub was bulked and air-dried in paper bags,A dilution series was made of each bulked sampleof sand in sterile 0 1 % agar water, and I ml ofeach dilution was pipetted onto petH dishes ofPCNB agar (Nash & Snyder. 1962), Each dilutionwas replicated five times. Colonies were countedafter 3 days' incubation at room temperatureunder fluorescent lights. Two weeks after infes-tation with F. o. traeheiphilum, 10' eggs of race Iof M. incognita (isolate I 265) were pipetted into ahole (12 mm diameter, 20 mm deep) made in thesand in each pot, Nematode eggs were extractedwith 0 5% NaOCI (Hussey & Barker, 1973) froma single-egg-mass culture maintained on tomato(Lycopersicon escutentum 'Rutgers') in a glass-house. CB5 is moderately susceptible to someisolates of A/. //icr̂ n̂/Va (Swanson. 1984), Effects

of \l incognita on Fusarium wilt were measuredby including eonidial and ehlamvdospore treat-ments (10'' spores per pot) without A/ ineogniia.A germinated seed of cowpea cultivar CB5 wasplanted immediately ia each hole (one seed perpot). Four weeks after planting, each seedling wasassessed for Fusarium wilt incidence and severity.The method of Byrd et ai (1983) was used to stainroots with acid fuchsin, except that lactic acid wassubstituted for HCl, and mature St incognitawere counted. Data on effects of inoculum con-centration were analysed separately for each typeof spore with the general linear models (GLM)procedure in SAS (Statistical Analysis System;SAS Institute Inc, 1985), and data on effects ofM incognita were subjected to analysis of vari-ance.

In the second experiment, the same materialsand methods were used, except as detailed below ,Because ofthe low infection rate of eonidia in theprevious experiment, the lowest inoculum con-centration of conidia (lO-* per pot) was omitted.Further, to increase the chances o\ infection hvconidia, seedling roots were cut with a knilc in situ2 weeks after sowing, immcdiatelv before theconidial and nematode suspensions were pouredinto the knife slit. Roots were wounded only inthe conidial treatments to conlirtn that conidiaadded to soil can inlVxt cowpeas On the da\ oi

448 A. R. Harris and H. Ferris

inoculation with conidia, germinated seeds ofcowpea cultivar CB5 were planted in sandinfested with chlamydospores at four differentrates. One seed was planted in each 460-cm' (14-oz) polystyrene cup. immediately followed byinoculation with 10-' eggs of M. javanica (isolateJ255, a subculture of isolate J-7c-54 describedlater), M. javaniea was used in this experiment todetermine whether a putative virulent nematodeisolate (Harris & Ferris, 1991a) could increaseincidence or severity of Fusarium wilt in asusceptible cowfjea cultivar, because M. ineognitahad no effect on wilt in the previous experiment.Treatments were conidia at 2 x 10-, x 10'. andX lO'' per cm' of sand, chlamydospores at 2 x 10',X 10-, X 10', and x 10'' per cm' of sand (all withM. javaniea), control (M. javaniea but no F. o.traeheiphilum) and 2x10 ' conidia or chlamydos-pores without M. javaniea. Plants were grown for5 weeks after infestation of sand. Roots werestained as described for the previous experiment,weighed, and the number of females, egg massesand galls counted. Regression coefficients andintercepts of the disease models for conidia andchlamydospores were compared by /-test.

Inoculum density of M, javanica

Germinated seeds of wilt-resistant cowpea culti-var CB3 were planted in sand in 500-cm' plasticpots (one seed per pot). Juveniles of A/, javanieawere pipetted immediately into each planting holeat rates of 0. 10. I0-. 10'or 10" nematodes per pot,Nematodes were obtained from single-egg-masscultures of isolate J-7c-54 of M. javaniea ontomato cultivar UC82 by placing infected roots ina Seinhorst mist apparatus (Seinhorst, 1950),Juveniles were collected at 8-h intervals andstored for up to 3 days at 10 15 C with aeration,until enough were obtained for the inoculations.This putative virulent nematode isolate was col-lected in 1954 by 1, J, Thomason from cowpeas insouthern California, and was used in the interac-tion experiments of Thomason et ai (1959) andSwanson (1984), and in our subsequent experi-ments(Harris& Ferris, 1991a, b). Each treatmentwas replicated 12 times in a randomized completeblock design with split plots, with environmentalgrowth chambers as blocks and position withinthe chambers (front, middle or back) as mainplots Treatments were completely randomizedwithin main plots. After 4 weeks, seedlings weretransplanted carefully, to minimize root damage,into 1320-cm' (13-5-cm-diameter) plastic pots.Each pot contained approximately 900 cm' of

Table 2. Effect of initial inoculum density of juveniles ofMeloidogyne javanica on the incidence of Fusarium wilt,and on final galling index values, of wilt-resistantcowpea cultivar CB3 inoculated with Fusarium oxys-porum f. sp, tracheiphilum at 2 x 10* cf.u, per cm' of

sand

Initialnumber ofM. javanica

010

1001,000

10,000

Proportion ofplants with

Fusarium wilt

0090-400670-751-00

Galling indexvalue

(mean ± S.D,)

0±036-1-1639 ±2772 ±1688±6

sand infested with chlamydospores to give adensity of2 x lO"* colony-forming units (cf.u.) percm' of sand. Plants were scored for disease rating,vascular discoloration at the primary node, androot galling (Daulton & Nusbaum, 1961)8 weeksafter inoculation with F. o. tracheiphilum (i.e.when the cowpeas were 12 weeks old) (Table 2),Galling index values were: O = free from galls;1 = trace, less than 5 galls; 5 = very slight, trace to25 galls; 10 = slight, 26 to 100 galls; 25 = moder-ate, galls numerous, mostly discrete; 50 = moder-ately heavy, galls numerous, many coalesced;75 = heavy, galls very numerous, mostly coa-lesced, root growth slightly retarded; 90 = veryheavy, mass invasion, slight root growth;100 = extremely heavy, mass invasion, no rootdevelopment. Data for wilt ratings were analysedwith the GLM procedure in SAS,

One stem section was sampled from each blockfor each treatment, surface-disinfested in 0-5%NaOCl. and placed in petri dishes of YDAmedium (Nelson et ai, 1986), Emergent colonieswere examined microscopically, and the presenceof F. oxysporum was confirmed in the stems ofdiseased plants. Root samples also were collectedfrom each block for each treatment, adult femalenematodes were extracted from galls, and peri-neal patterns were cut (Hartman & Sasser. I98S)and examined to confirm that galling was causedby M. javaniea.

Comparison of moderate and high inoculumdensities of both F, o. irackeipkihtm andM, javanica on three cultivars

Three cowpea cultivars were tested, and theirreported reactions to F. o. iracheiphilum are CB3.resistant: CBS. susceptible; and Grant, tolerant

Inoculum densities o/Fusarium oxysporum and Meloidogyne in cowpea 449

(Thomason et ai, 1959), Grant is a selection ofChino 3 and is the cultivar that first was reportedto be more susceptible to Fusarium wilt in thepresence of M. javaniea (Thomason et ai, 1959),All three cultivars are susceptible to M. javaniea.Each cultivar was inoculated at a high, moderate,or zero level of F. o, tracheiphilum and M.javanica. The high level (2x10-^ c,f,u, of F. o.tracheiphilum per cm' of sand plus 10"̂ nematodeeggs per plant) approximated those that Swanson(1984) reported reduced resistance to F. o. tra-cheiphilum in CB3, The moderate level (2x KJ-*c,f,u, per cm' of sand plus 10' nematode eggs perplant) was considered optimal from the aboveexperiments, based on disease level, variabilityand densities occurring naturally in the field. Thedensity of F. oxysporum was higher than thatoccurring naturally in the field (Rigert, 1985), butthe nematode density approximated densitiesthat could be found in soil of severely infestedfields (Harris, 1984),

Polystyrene cups (460 cm') were filled withchlamydospore-infested sand (as describedearlier), then one germinated seed was planted ineach cup. Eggs of M. javaniea (isolate J-7c-54)were obtained as described above, and 5 ml of eggsuspension was pipetted into each planting hole.Sterilized straw and distilled water were added tothe sand in control treatments (zero level). Eggsof M. javaniea were used as inoculum becausethey are surface-sterilized during bleach extrac-tion, and are easier to extract. Further, eggspresumably have the potential to infect over alonger period than second-stage juveniles, andasynchronous infection undoubtedly occurs inthe field. To determine nematode viability, asample of eggs was placed on a 500-mesh (0-026-mm) sieve in tap water, and hatched juvenileswere collected and counted every I -3 days. Over38 days, 9-1% ofthe eggs hatched. Cups with thesame treatment were nested in groups (tuhs) ofthree, and each group was replicated five times, asdescribed earlier. Tubs were randomized withinwater baths (blocks) in a complete factorialdesign. Foliar Fusarium wilt symptoms wererecorded daily, and as plants wilted and died, theywere assessed for disease rating, vascular discolo-ration, root galling, stem length, and numbers ofpods and seeds. Nine weeks after inoculation, allsurviving plants were assessed. Results for eachcultivar were analysed by one-way analysis ofvariance in the SAS GLM procedure, andweighted least-squares techniques were used fordisease rating and vascular discoloration to mini-mize the residual sum of squares. This conserva-

tive adjustment was necessary to compensate forunequal variances in wilt ratings. Stem sectionsfrom plants with doubtful symptoms were platedonto PCNB agar, and resultant colonies weresuhcultured onto YDA slants to confirm infec-tion hy F. oxysporum. Root samples also werecollected from each treatment, stained as de-scrihed earlier, and root-knot nematodescounted.

Identity and density of F. a. tracheiphilum andMeloidogyne spp, in a cowpea field

One field of cowpea cultivar CB3 that showedFusarium wilt symptoms was lound in TulareCountv, California in 19XS This field was sam-pled to determine the identities and naturaldensities of F. o. traeheiphilum and Sfclaulogynespp, in field soil. Small sections were cut from thelower and middle stem of plants that showedvascular discoloration. The seetions were surface-sterilized in 1-0".. NaOCl lor I min, then placedon PCNB agar. Fungi were suhcultured ontoPDA and carnation Icaf-piccc agar (Fisher et ai.1982), and identified hy colony appearance,microscopic morphology, and differential hosttest (Nelson et ai, 1986), Root-knot nematodesalso were isolated from root galls and single-egg-mass cultures were established on tomato cultiv arUC82, The nematodes were identified hv peri-neal-pattern morphology and differential hosttest (Hartman & Sasser, 1985), Two to 3 monthsafter harvest, four composite soil samples werecollected from areas where Fusarium wilt symp-toms had been severe, Eaeh sample was compo-sited from 20 25 cores, 5 30 cm deep. The soilwas a sandy loam, classed as Futric Fluvisol witha medium textured surface layer (FAO L'ncsco.1975), Ectoparasitic and juvenile nematodes wereextracted from soil with a modified semiautoma-tic elutriator and sugar flotation technique (Bvrdet ai. 1976), Endoparasitk nemalodcs and nev,lyhatched juveniles were extracted by placing theorganic matter collected from the elutriator ontoBaermann funnels in a Seinhorst mist apparatus(Seinhorst, 1950) for 7 davs Nematodes werecounted separately lor each extraction techniquewith a dissecting microscope, and the totalnumber of ncmalodes of each genus wa> calcu-lated for each sample, F. oxysporum also v\asisolated from each composite soil sample, hvdilution plating 5 g of air-dry soil onto PCNBagar by the methods of Nash iV Snvder (1962),Colonies that appeared to be F. o\v.sporiim weresubcultured onto PDA in petri dishes, incu-

450 A. R. Harris and H. Ferris

3

2-5

2

1-5

1

0-5

O Vascular discoloration (mean)Vascular discoloration

V= -1-12-)-0-94Xp = 0-37*

vf Coefficient of variation^ y = 305-76 - 76-73X-

' , f^ = 0-86***

200 I

100

0 1 2 3 4 5Log,o(X-i- 1) (Chlamydospores/cm' of soil)

Fig, 1. Effect of inoculum density of chlamydospores of Fusarium oxysporum f, sp, tracheiphilum on vasculardiscoloration at the primary node of wilt-susceptible cowpea cultivar CB5 infected with Meloidogyne incognita (firstexperiment). Each data point is the mean of five replications, but regressions were not performed on means,(•), (*•*) Indicate significant coefficients of determination at /• = 0-05 and 0-001, respectively.

bated at 24-25 C in the dark for 3 days, and thencolony diameter was measured. Colonies withdiameter greater than 2 cm were macerated in 5ml of sterile, distilled water, and each isolate wastested for pathogenicity by inoculating two to fiveseedlings of cowpea cultivars CB3 and CB5, asfollows. Germinated seeds were sown in sand intrays. After 11 days, seedling roots were cut, andthe seedlings were soaked in the fungal suspen-sion for at least 2 min, before they were trans-planted into aerated-steam-treated U, C, soil mix(Baker, 1957) in 190-cm' (6 oz) polystyrene cups.The plants were maintained in a glasshouse(approximately 27 C). and Fusarium wilt symp-toms were recorded weekly after 3 weeks,

RESULTS

Effect of inoculum type and density of F. o.tracheiphilum and of inoculation method

When conidia were added to sand in the firstexperiment, the disease ratings and vasculardiscoloration ratings for all treatments were low( < 0-54). and were not different from controls (noF. o. traeheiphilum)(P<005). Dilution plating ofthe sand collected from pots just after plantingseedlings indicated that survival of conidia waslow (Table I), The concentration of chlamydo-spores. however, had a significant effect onvascular discoloration, and a linear model bestdescribed the relationship between inoculatedtreatments on a ]ogH,{X-^-1) scale (/><0-01) (Fig,I), The coefficient of variation was inverselyproportional to the logia(A'-f I) of the chlamydo-

spore inoculum density. When disease rating wasused as a measure of Fusarium wilt, there was adifference (f < 0-01) between the control (no F. o.tracheiphilum) and all /i«ar/M/w-infested treat-ments. The linear model comparing infestedtreatments was significant at /'=0-053. Linearcontrasts indicated that controls (no F. o. trachei-philum) had more M. incognita (P<0-05) thantreatments infested with either conidia or chlamy-dospores (Table 1), M. incognita did not affect theseverity of Fusarium wilt at one inoculum level(10* conidia or chlamydospores jjer pot), eventhough some nematodes infected roots andmatured.

In the second experiment, the inoculum densit-ies of conidia and chlamydospores affected bothdisease rating (Fig, 2) and vascular discoloration(Fig, 3) (P < 0-01), Treatments infested with chla-mydospores had greater disease ratings and vas-cular discoloration ratings than conidial treat-ments with root slicing at each inoculum density.The number of plants showing Fusarium wiltsymptoms also increased with Fusarium inoculumdensity. For both disease rating and vasculardiscoloration, the intercept for chlamydosporeswas higher than that for conidia (P<0-0OI). butthe regression coefficients did not differ(P<0 05). The coefficients of variation wereinversely proportional to the logio(A'-t-1) of sporeconcentrations, as for the previous experiment.The equations for the linear models, and theircoefficients of detennination. for Fusarium wiltratings and coefficients of variation are given inthe legends for Figs 2 and 3. Addition oT M.javanica had no effect (at P»O-OS) on mean

Inoculum densities o/Fusarium oxysporum and Meloidogyne in cowpea

3r

451

2-5

o) 2

I 1-5 -

5 1 -

0-5 -

A Conidia (mean)O Chlamydospores (mean)

Chlamydospores

OA

Conidia

200 I-cta>

'o

100 Ioo

0 1 2 3 4 5

Log,o(X-H 1) (Spores/cm^ of soil)

Fig. 2. Effect of inoculum density of either conidia or chlamydospores oi Fusarium oxysporum i. sp, iracheiphilum ondisease rating of wilt-susceptible cowpea cultivar CB5 infected with M eloulogyne javanica (second experiment). Eachdata point is the mean of five replications, but regressions were not performed on means.For chlamydospores:

disease rating Y= -0-76-1-0 70Jk'(/?' = 0 52, /'<0OOI);coefficient of variation y = 225-20-48 70A'(/?- = 0 61, /'<0001),

For conidia:disease rating K= - 2 36-I-1 04X1/?- = 0 66, P<OOa\)\coefficient of variation f=250 22-51-09JI (/?- = 0 90, /'<OOOI).

•r'V

1,i

1010

u(/}

3

asc

3

2-5

2

1-5

1

0-5

A Conidia (mean)O Chlamydospores (mean)

Chlamydospores

O

o

6-

A

~ Conidia

c^100 I

0 ^ 1 2 3 4 5Log,o(X + 1) (Spores/cm^ of soil)

Fig. 3, Effect of inoculum density of either conidia or chlamydospores oi Fusarium oxysporum ( sp, iracheiphilum onvascular discoloration at the primary node of wilt-susceptible cowpea cultivar CB5 infected with Sleloul>'i;\ncjavanica (second experiment) Each data point is the mean of five replications, but regressions were not performed onmeans.For chlamydospores:

vascular discoloration f = - I 05 + 0-87,V (/?• =0 52, P<OOt}\y,coefficient of variation y' = 243-06-51-87,V (/?- = 0 XV P<000\).

For conidia:vascular discoloration f =-2-85- f 1 I9,\ (/?- = 0 62, P<000\):coefficient of variation f = 169 37-?2 4(),V (/?-=0 99, P<000\)

disease rating or vascular discoloration in thesusceptible cowpea cultivar (CB5), although thenumber of plants showing Fusarium wilt symp-toms increased from two to six in the chlamydo-spore treatments. Equivalent conidial treatmentsshowed no differences in disease incidence or

seventy. Linear contrasts indicated that controls(St. javaniea but no F. o. traeheiphiltim) had moreM. javaniea per g of stained root (mean = 157)than the conidial treatments (mean = 75)(P < 0-001), However, these controls had no moreM. javanica per g (at ^ = 0-05) than chlamydo-

452 A. R. Harris and H. Ferris

2-5

ig 1-5

«0-5

OVascular discoloration (mean)

Vascular discolorationy = -0-69 + 0-75X,f^ = 0-24"

Coefficient of variation''9= 201-71 - 35-20XR' = 0-82"*

200

0 1 2 3 4Log,o(X-(- 1) (M. yavan/ca/Plant)

Fig. 4. Effect of inoculum density of juveniles of Meloidogyne javanica on vascular discoloration at the primary nodeof wilt-resistant cowpea cultivar CB3 inoculated with Fusarium oxysporum f. sp, tracheiphilum. Each data point is themean of 12 replications, but regressions were not performed on means, ( " ) , (•••) Indicate significant coefficients ofdetermination at P = 0-01 and 0-001, respectively.

Table 3. Effect oi Fusarium oxysporum i. sp, Iracheiphilum and Meloidogyne javanica inoculum densityon Fusarium wilt ratings, plant death, root galling, and pod and seed production in three cowpea

cultivars 9 weeks after inoculation with both pathogens

Cultivarandputativeresistanceto wilt

Grant(tolerant)

CB3(resistant)

CB5(susceptible)

Inoculumlevel

0ModerateHigh

0ModerateHigh

0ModerateHigh

Diseaserating

0 3 B''

1 5 A

2-2 A

00 c1 8 B

2 8 A

1 2 B

3 9 A

3 7A

Vasculardiscoloration

at primarynode

0 2 B

21 A2 7 A

0 0 c1 7 B

3 9 A

1 4 B

4 9 A

4-7 A

Gallingindexvalue

0 0 B

30 4 A

49 6 A

0-0 c28 0 B

50-2 A

0 0c18 6 B

28-4 A

Proportionof plants

dead

0-000-07013

0000-13040

020093093

Number ofpods

3-2 A4 1 A3-7 A

3 1 A3 3 A

2 2 B

3-4 A1 7 B

0 9 B

Number ofseeds

14 1 A14 1 A

I 9 B

15 3 A

11-3 A5 5 B

12 6 A5 0 B

2 3 B

"Within cultivars and columns, means not followed by a common letter differ significantly (i>< 0-05)according to a r-test.

spore treatments (mean =135), There also was aninverse linear relationship between chlamydo-spore inoculum density and total number of M.javanica (significant at P = 005), when controls(no F. o. tracheiphilum) were omitted from theanalysis.

Inoculum density of M.javanica

1- usarium wilt symptoms were slow to develop, sothat by the time of assessment some plants were

partially defoliated and chlorotic. due to naturalsenescence and damage caused by two-spottedmites. Therefore, only data for vascular discolo-ration at the primary node, and not for diseaserating, are considered reliable measures of Fusar-ium wilt disease, and are presented in Fig, 4.There was a difference (/•< 0-001) in vasculardiscoloration between treatments with M. javo-nica and controls (no M.javanica), at this moder--ately high fungal density of 2 x 10* c.f.u, per cm'of sand. A linear model best described the

Inoculum densities o/Fusarium oxysporum and Meloidogyne in cowpea 453

relationship between the different inoculum levelsof M. javanica and vascular discoloration at theprimary node (Fig. 4). As initial nematode inocu-lum density increased, the proportion of plantsshowing Fusarium wilt symptoms increased, asdid the final galling indices (Table 2). The regres-sion of final galling index on logio(A'-t-1) of theinitial number of M. javanica fitted the linearmodel. f=3-8-(-21-6A'(/?^ = 0-76, P<0-001),

Comparison of moderate and high inoculumdensities of both F. o. tracheiphilum andM, javanica on three cultivars

In the susceptible cultivar CB5, both moderateand high inoculum densities off", o. traeheiphilumand M. javanica resulted in severe wilt, death of14/15 plants, and fewer pods and seeds (P<00])(Table 3), In the wilt-tolerant cultivar Grant andthe wilt-resistant cultivar CB3, moderate inocu-lum densities caused mild wilt symptoms in 10/14and 12/15 plants, and only one and two deadplants, respectively, and no reduction in numberof pods or seeds (P<005). A tenfold increase ininoculum density of F. o. traeheiphilum and A/javanica increased wilt ratings in CB3 (P<005),but not in Grant, The number of pods and seedswas reduced in CB3, but only seeds were signifi-cantly fewer in Grant, Inoculum density had noeffect at P = 005 on stem length. The meangalling index value for CB5 was lower (P<005)than the means for either CB3 or Grant, whichdid not differ from each other. Two Grant and sixCB5 control plants became infected with F. o.tracheiphilum, possibly due to airborne contami-nation, since care was taken to avoid splashingduring watering.

Identity and density of F. o. tracheiphilum andMeloidogyne spp. in a cowpea field

The fungi and nematodes isolated from cowpeassampled from the naturally infested field inTulare County were identified, respectively, as F.o. tracheiphilum race 3 and M. ineognita race 1,The four soil samples from this field had densitiesof F. o. tracheiphilum of 335, 252, 361 and 178cf.u. per cm' of air-dry soil. The same samples,respectively, had densities of 3331, 8677, 13 063and 3074 Meloidogyne spp, (all stages) per 1000cm' of moist soil. Three of the samples alsocontained low numbers of Pratytenchus spp,, twocontained Criconemetla spp, and Trichodorusspp,, and one contained Hoplolaimus spp.

DISCUSSION

In the first experiment, only 17% (8/47) of thesurviving plants of cultivar CB5 infested withconidia showed Fusarium wilt symptoms, asmeasured by disease ratings greater than 1, Thislow infection of the susceptible cultivar wasapparently due to the low survival of conidiawhen mixed with sand (Table I), In contrast, 57"/.,(20 35) of surviving plants infested with chlamy-dospores showed Fusarium wilt symptoms. Thehigher number of M. ineognita in roots of controlplants (no F. o. traeheiphilum) could have beenbecause of larger, healthier root systems thatcould support more nematodes. Some of theplants infested with F. o. traeheiphilum were deador dying, and roots were already decaying whensampled, so SI. ineognita may have been lost. Thenumber of A/, ineognita per unit of root may notnecessarily have differed between treatments, soroots were weighed in the second experiment. Theinability of SI ineognita to affect Fusarium wiltwas confirmed in two subsequent experiments(Harris, 1989; Harris & Ferns, 1988a, 1991a, b).

In the second experiment, chlamydosporesresulted in greater disease than conidia, eventhough the latter were added near wounded roots.Because plants infested with conidia were 2 weeksolder at the time of infestation than those infestedwith ehlamydospores, it is possible that plants inthe conidial treatments were more resistant to F.o. traeheiphilum due to excess \ascular capacitN(Beckman, 1987), However, it is more likely thatchlamydospores simply had greater survival andinoculum potential than conidia, Chlamydo-spores are the natural survival structures andprimary inoculum of F. iixyspurum (Nelson,1981; Beckman, 1987), Therefore, it is appro-priate to use chlamvdospores rather than eonidiaas inoculum for this type of experiment. Inocu-lum densities of 2 x ItJ-' spores per cm' of sandwere required to produce Fusarium wilt svmp-toms in at least 75".i ofthe plants. This inoculumdensity was the only level used that resulted inmean disease ratings and vascular discolorationratings considered to reflect susceptibility. It alsowas the only level with coefficients o\' variationless than 20",,, Greater variabilitv between plantswould necessitate using impracticallv largenumbers of replicate plants. Maximum diseaselevels were not achieved in either this or theprevious experiment, but even 2 x U)-" t, f,u percm' is high compared to densities occurringnaturally in the field (Rigert, 1985; Harris, I989>On the basis of those results, it was decided to use

454 A. R. Harris and H. Ferris

chlamydospores at 2 • KV c.f.u. per cm' of soil asthe standard inoculum density in subsequentexperiments (Harris, 1989; Harris & Ferris,1988a. b, 1991a, b).

In both experiments, nematodes were fewer inthe presence of F. o. traeheiphilum, even whennematodes per g root were considered. Thehypothesis that F. o. tracheiphilum reduces infec-tion by SIeloicloi;yne spp, needs further investiga-tion. More Heterodera juveniles penetratehealthy roots of sugar beet than roots infectedwith F. oxysporum (Jorgenson, 1970), Further,several surveys showed that Ftisarium commonlycolonizes cysts and eggs of Heterodera spp,(Morgan-Jones & Rodriguez-Kabana, 1987),

Cowpea cultivar CB3 remained virtually resis-tant to F. o. traeheiphilum race 3 in the absence ofSf javaniea (one plant showed some leaf chlorosisand slight vasctilar discoloration below the pn-mary node). As initial inoculum densities of SIjavaniea were increased, the final galling indicesalso increased. Final nematode populations werenot counted, however, because plants were heldlong enough (12 weeks) for the nematodes toproduce several generations. As nematodenumbers and galling increased, so did the inci-dence and severity of Fusarium wilt, until, at lO''nematodes, all surviving plants were diseased Sf.Javanica. therefore, apparently reduced the resis-tance of cowpea eultivar CB3 to F. o. traeheiphi-lum using these methods, even when as few as 10nematodes were added to each plant. This inter-action between St. lavanua and F. c traeheiphi-lum in CB3 was confirmed in six subsequentexperiments (Harris, 1989; Harris & Ferris,IMHMa, b, 1991a, b). These results support thoseof Garber ('( ai (1979), Sidhu & Webster (1981)and Swanson (1VX4). v\ho used different methodsand inoculum type. The results suggest that aquantitative factor associated with the nematodeslowers resistance to F oxysporum. MaximumFusarium wilt severity was not attained m thisexperiment, but even lO'' Sf javanica in thisvolume of soil (375 cm ) is considered a very highlevel m the field (even alter allowing for only 2025"" infection by added juveniles). A concentra-tion of 10'juveniles per plant (or pot of approxi-mately 500 cm' of soil) is closer to the densitiesfound in infested lields that show plant damage(Harns, 19X4), This inoculum level (10' nema-todes per plant) was used as the standard in otherexperiments (Harris, 1989; Harris iV: Ferns,1988a. b, 1991a, b)

At moderate inoculum densities of F. o. irmhet-philum and St jatantea (2 > 10'' cl u, per cm' of

sand plus 10' nematode eggs per plant), almost allCB5 plants died within 8 weeks of infestationunder our experimental conditions. Most plants(> 70%) of Grant and CB3 showed mild Fusar-ium wilt symptoms and moderate root galling,but no significant (P<0Q5) reduction in yield inthe glasshouse. The same inoculum densitiesproduced mild to moderate Fusarium wilt symp-toms in wilt-resistant CB3 in five other experi-ments conducted under similar conditions (Har-ris, 1989; Harris & Ferris, 1988a, b, 1991a, b). Athigh inoculum densities (2 x 10' c,f,u, [>er cm' ofsand plus 10' nematode eggs per plant). Grantremained tolerant to Fusarium wilt, and had onlymoderate vascular discoloration, although seedproduction decreased. These results are at vari-ance with those of Thomason et ai (1959), whoreported that, at a similar temperature range,xylem necrosis increased greatly if Sf. javanicawere added to soil in addition to spores of F. o.traeheiphilum. However, they did not quantifytheir inoculum, so comparison of results is diffi-cult, CB3 showed higher (P<005) incidence andseverity of Fusarium wilt symptoms at highinoculum densities, and all plants had somevascular discoloration at the primary node. Six ofthe 15 plants died and produced no seed pods, butthe surviv ing plants still appeared healthy after 9weeks, and apparently remained resistant. Theincrease in vascular discoloration at high levels ofinoculum is similar to that reported by Swanson(19S4), v\ho concluded that a predisposing inter-action is possible in CB3 in the field. However,both Swanson s results and ours suggest thatmost plants remain tolerant of Fusarium wilt,despite infection, at realistic inoculum densities.This conclusion is supported bv field observationsin California in recent years. Onlv two fields ofCB3 showing Fusarium wilt svmptoms werelocated and reported b> extension specialistsfrom 1985 to 1987, In 1988, letters were sent to 20extension officers, researchers and cowpeaspecialists throughout California, requestinginformation on fields with potential interactionson 1 usarium wilt-resistant cowpeas, Fusariumwilt was reported in only one additional field ofcowpeas The sampling from this field of CB3,reported above, revealed high densities (Harns.1984) of Steloidogyne spp, (some identified as M.tnei'Kiuta race 1) in the soil. The dcnsitv of F.owsporum in field soil samples (2-3 months afterharvest), however, was approximatciv 100 timeslower than the standard iniKulum density used inthe above experiments possibly because ofclumped distribution m the Held, Root-knot

Inoculum densities o/Fusarium oxysporum and Meloidogyne in cowpea 455

nematodes isolated from other California cowpeafields also were M. incognita, and most isolates ofF. o. tracheiphilum were race 3 (Harris, 1989).

Under the conditions of our experiments,infestation of soil with chlamydospores at 2 x lO''c.f.u. per cm' and with M. javanica at 10'nematodes per plant gave moderate levels ofdisease. These inoculum densities were a com-promise between densities occurring naturally infield soils, and densities required to obtain maxi-mum Fusarium wilt with minimum variabilitybetween plants. The above densities and inocula-tion methods which minimize root woundingwere used in subsequent studies of the nature ofinteractions (Harris. 1989; Harris & Ferris.1988b, 1991a, b),

ACKNOWLEDGEMENTSWe thank Drs N, H, Willits and D, M, Eastburnfor advice on statistical analyses, Dr I, J, Thoma-son for supplying a nematode culture, C, Frateand F, Workneh for collecting field samples ofsoil and cowpeas. and Drs J, E, DcVay and J, D,McDonald for reviewing the manuscript,

REFERENCESAbawi G,S, & Barker K,R, (1984) Effects of cultivar,

soil temperature, and population levels of Meloido-gyne incognita on root necrosis and Fusarium wilt oftomatoes. Phytopathology 14, 433 43X

Baker K,F. (Ed.) (1957) The U. C System for ProducingHeallhy Container-Grown Plants. California Agric,Exp, Stn, Ext, Serv, Manual 23

Beckman C H , (1987) The Nature of Hilt Di.sea-.e.\ ofPlants. APS Press, St Paul, Minnesota,

Byrd D,W,. Jr,, Barker K,R,, Ferris H,, Nusbaum C,J.,Griffin W,E,, Small R H , & Stone C A , (1976) Tvsosemi-automatic elutriators for extracting nematodesand certain fungi from soil. Journal of Nematology S,206 212,

Byrd D,W,, Jr,, Kirkpatrick T, & Barker K,R, (1983)An improved technique for clearing and stainingplant tissues for detection of nematodes. Journal ofNematotogy 15, 142-143.

Daulton R,A,C, & Nusbaum CJ , (1961) The effect ofsoil temperature on the survival of the root-knotnematodes Meloidogyne javanica and SI. hapla.Nematologica 6, 280 294,

FAO-Unesco(1975)5o(/A/apo///it' World. I StldlltltltlVol. 2. North America. Unesco, Paris.

Fisher N,L,, Burgess L.W,, Toussoun T A , & NelsonP,E, (1982) Carnation leaves as a substrate and forpreserving cultures of Fusarium species. Phytopatho-logy 12, \5\Mi.

Garber R H , , Jorgensen E C , Smith S, & Hyer A H ,(1979) Interaction of population levels of Fusariumoxysporum f, sp, rasmlectum -and Meloidogyne incog-

nita on cotton. Journal of Nematology I I , 133 137,

Harris A,R, (1984) Distribution of plant parasiticnematodes in horticultural crops in the Gol Gol,Mildura, Nangiloc, Robinvale and Swan Hill dis-tricts, Australasian Plant Pathology 13, 52 55,

Harris A,R, (1989) Interactions between Fu.sariumoxysporum i. sp, tracheiphilum and Meloidogyne spp,in Vigna unguicutala. PhD Thesis. University ofCalifornia, Davis

Harris A,R. & Ferris H, (1988a) Interaction of Mclonhi-gyne spp, and Fusarium oxysporum on cowpea (Vignaunguiculata). (Ahslr.) Journal o/ Neiiuiiolngy 20, 639

Harris A,R. & Ferris H, (1988b) Nature of interactionsbetween Fusarium o\ysporutn and Mclouloi;yncjara-nica in cowpea, (Abstr.) Phytopathology 78. 1575,

Harris A R , & Ferns H, (1991a) Interactions betweenFusarium oxysporum f, sp. iracheiphilum and Sfetoi-dogync spp, in I'igna unguiculala. 2 Specificity otdifferent taxa, Ptant Pathology 40, 457 464,

Harris A R , & Ferris H, (1991b) Interactions betweenFusarium oxysporum i. sp, tracheiphilum and Meloi-dogyne spp, in Vigna unguieulala. 3, Pathogenesis byF o. Iracheiphilum as dfiecled by SI /aianuadnd hostcultivar. Plant Pathology 40, 465 476.

Hartman K M & Sasser J,N, (1985) Identification ofSteloidogyne species on the basis of differential hosttesi and perineal-pattern morphology ln: .4nAdvanccil Treatise on SleloiJogyne. Vol. ii Slethod-ology (Ed by K R Barker, C C. Carter & J \Sasbcr), pp. 69 77 Coop. Publ Dept Plant Palhol ,North Carolina State L'ni\ersit\, and LI S AgencyInt. Dc\ , Raleigh, North Carolina

Hoag landD,R& Arnon D.I (1950) TIK Water-cultureStclluullor Oro» ing Plants Hilhotit Soil Calif .-XgricExp. Stn, Circ. 347

Hussey R S , & Barker K,R, (1973) A companson ofmethods of collecting inocula of Steloidogyne sppincluding a new technique. Plant Disease Reporter 51,1025-1028.

Jones J P , Overman A J. & Crill P (1976) Failure ofroot-knot nematode to affect Fusarium wilt resis-tance in tomato. Phytopathology 66, 1339 1,̂ 40

Jorgenson F C (1970) Antagonistic interaction ofHeUrodera schachtii Schmidt and Fusarium o\\-,v/><)rum (Wolf) on sugarbeets Journal ol \cniatology2, 393-39K

Mai W,F, & Abawi G,S, (1987) Interactions amongroot-knot nematodes and Fusarium will fungi onhost plants .Annual Review of Phyiopaihi>loi;y 25,317-338,

Morgan-Jones G, & Rodriguez-Kabana R (I'^S"')Fungal biocontrol for the management of nema-todes In: listas on .\eniatology: .4 Commemorationof the Twenty-fifth .4nniicr.sary of the Soeuty ofNematologi.sls(^d. by J A. Veech \ D W DicKson),pp, 94 99 Society of Ncmatologists, Inc.. H\,ilts-villc, Maryland

Nash S, M. Se. Snyder W,C, (1962) Quantitative estima-tions by plate counts of propagules of the bean rootrot Fu.sarium in field soils, Phytopatholi'gy 52, 567572,

Nelson P I (1981) Life csclc and epidcmiolog> of

456 A. R. Harris and H. Ferris

Fusarium oxysporum. In: Fungal Wilt Disea.tes ofPlants (Ed, by M, E, Mace, A, A, Bell & C, H,Beckman), pp, 51-80, Academic Press. New York,

Nelson P,E,, Toussoun T,A,, Burgess L,W,, MarasasW,F,O, & Liddell C M , (1986) Isolating, identifying,and producing inoculum of pathogenic species ofFusarium. In: Methods for Evaluating Pesticides forControl of Plant Pathogens (Ed, by K, D, Hickey), pp,54-59, APS Press, St, Paul. Minnesota,

Rigert KS, (1985) The inheritance of Fusarium wiltresistance in cowpea (Vigna unguieulala (L) Walp,),PhD Thesis, University of California, Davis,

Rigert K,S, & Foster K,W, (1987) Inheritance ofresistance to two races of Fusarium wilt in threecowpea cultivars. Crop Science 27, 220-224,

SAS Institute Inc, (1985) 5/15" V.ser's Guide: Statistics,Version 5 Edition. Cary, N C : SAS Institute Inc,

Seinhorst J,W, (1950) De betekenis van de toestand vande grond voor het optreden van aantasting door het

stengelaaltje (Ditylenchus dipsaci (Kuhn) Filipjev).Tijdschrift Plantenzieklen 58, 289-348.

Sidhu G.S, & Webster J.M. (1981) Influence of popula-tion levels of root-knot nematode on Fusarium wiltseverity of tomato. Phytoprotection 62, 61-66.

Starr J.L., Jeger M.J., Martyn R.D. & Schilling K.(1989) EfTects of Metoidogyne incognita and Fusariumoxysporum f. sp. vasinfectum on plant mortality andyield of cotton. Phytopathotogy 79, 640-646.

Sumner D.R, & Johnson A.W. (1973) EfTect of root-knot nematodes on Fusarium wilt of watermelon.Phytopathotogy 63, 857-861.

Swanson T,A, (1984) Root-knot nematode and Fusar-ium wilt diseases of cowpea and soybean, PhD Thesis,University of California, Riverside,

Thomason LJ,, Erwin D,C, & Garber MJ, (1959) Therelationship ofthe root-knot nematode, Metoidogynejavanica, to Fusarium wilt of cowpea. Phytopatho-logy 49, 602-606,