620

Mycologia, 95(4), 2003, pp. 620629.q 2003 by The Mycological Society of America, Lawrence, KS 66044-8897

Primordia initiation of mushroom (Agaricus bisporus) strains onaxenic casing materials

R. Noble1

T. R. FermorS. LincolnA. Dobrovin-PenningtonC. EveredA. Mead

Horticulture Research International, Wellesbourne,Warwick, CV35 9EF, United Kingdom

R. LiDepartment of Agriculture, Yunnan AgricultureUniversity, Kunming 650201, China

Abstract: The mushroom (Agaricus bisporus) has arequirement for a casing layer that has specificphysical, chemical and microbiological propertieswhich stimulate and promote the initiation of pri-mordia. Some of these primordia then may developfurther into sporophores, involving differentiation oftissue. Wild and commercial strains of A. bisporuswere cultured in axenic and nonaxenic microcosms,using a rye grain substrate covered by a range of or-ganic and inorganic casing materials. In axenic cul-ture, A. bisporus (commercial strain A15) was capableof producing primordia and mature sporophores oncharcoal (wood and activated), anthracite coal, lig-nite and zeolite, but not on bark, coir, peat, rock-wool, silica or vermiculite. Of six strains tested, onlythe developmental variant mutant, B430, producedrudimentary primordia on axenic peat-based casingmaterial. However, none of these rudimentary pri-mordia developed differentiated tissues or beyond 4mm diameter, either on axenic casing material in themicrocosms or in larger-scale culture. In larger-scale,nonaxenic culture, strain B430 produced severelymalformed but mature sporophores in similar num-bers to those of other strains. Typically, 36% of pri-mordia developed into mature sporophores, but sig-nificant differences in this proportion, as well as inthe numbers of primordia produced, were recordedbetween 12 A. bisporus strains.

Key words: aggregate, hypha, mycelium, primor-dium, Pseudomonas putida, sporophore

Accepted for publication December 23, 2002.1 Corresponding author. E-mail: ralph.noble@hri.ac.uk

INTRODUCTION

The initiation of sporophores of the mushroom Agar-icus bisporus (Lange) Sing. begins with the aggrega-tion of mycelium into thick strands or hyphal cords(Umar and van Griensven 1997). This initiation isfollowed by the formation of primordia, defined byWood (1976) as structures greater than 1 mm diam-eter composed of a dense hyphal mesh with a smoothsurface and visibly distinct from strands or knots ofhyphae. Flegg (1979) referred to the latter as my-celial aggregates. In addition to a nutritional sub-strate, e.g., compost (Noble et al 1998), most A. bis-porus strains have a requirement for a separate cas-ing layer that has specific physical, chemical and mi-crobiological properties which stimulate andpromote the initiation of primordia (Eger 1961,Flegg and Wood 1985). Primordia are formed aftermycelial colonization of the nutritional substrate andcasing layer. Development of primordia is stimulatedby a reduction in temperature and CO2 (Flegg andWood 1985) and by the presence of bacteria in thecasing layer, of which Pseudomonas putida (Trevisan)Migula is regarded as most significant (Hayes et al1969, Park and Agnihotri 1969). Long and Jacobs(1974) showed that activated charcoal could replacethe function of stimulatory bacteria in the casing lay-er, possibly by adsorbing compounds that were inhib-itory to initiation (Wood and Blight 1983). After ini-tiation, some primordia may continue to develop insize and by differentiation of tissues into sporophores(Wood 1979).

Strains of A. bisporus have been found to varymarkedly in their ability to form primordia axenicallyon malt-extract agar (MEA). Using 12 strains of com-mercial origin, Wood (1976, 1979) found that suchprimordia were formed asynchronously. Wannet et al(1999) found two wild strains from the Agaricus Re-covery Programme (ARP) collection (Sylvan SpawnLaboratory Inc., West Hills Industrial Park, Kittaning,Pennsylvania), 116 and 155.8, which produced pri-mordia after mycelial colonization of MEA, usually atthe edge of the plate. Except in rare cases, none ofthe above strains produced sporophores in the ab-sence of a casing layer, and none produced primor-dia on defined-agar media. Elliott and Wood (1978)isolated two developmental variant, mutant A. bispo-rus strains (B430 and B431) that produced primordia

621NOBLE ET AL: MUSHROOM PRIMORDIA

FIG. 1. Kilner jar microcosm, A. bisporus strains B430 and A15.

synchronously throughout mycelial growth on de-fined-minimal media but were not shown to producemature sporophores.

The aims of the current work were to determineif: (i) axenic casing materials other than activatedcharcoal could stimulate initiation; (ii) the mutantstrain B430 and other wild A. bisporus isolates couldinitiate primordia and develop sporophores on anaxenic casing layer; (iii) primordia and sporophoresof the above strains develop normally under nonax-enic conditions, compared with commercial A. bis-porus strains.

MATERIALS AND METHODS

Axenic microcosm tests. A microcosm (500 mL, Kilner jar,Ravenhead Glass Ltd, St. Helens, Lancashire, United King-dom) culture system, modified from Rainey (1989), wasused to examine growth in axenic casing materials (FIG. 1).The microcosms were filled with a 10 mm layer (30 g) ofsterile rye grain spawn (Elliott 1985), which served as anutritional substrate, colonized with the appropriate A. bis-porus strain. This substrate was covered with 60 g (about 17mm layer, surface area 64 cm2) of casing material, consist-ing of a mixture of black peat and CaCO3 (4:1 v/v) or othermaterials. Gamma irradiation (Technical Service Consul-

tants Ltd, Heywood, Lancashire, United Kingdom) was usedto produce axenic casing material, to avoid physical andchemical changes caused by autoclaving. The microcosmswere incubated at 25 C until mycelium had started to col-onize the surface of the casing layer, normally 35 d afterthe jars were filled. To promote conditions in the micro-cosms for initiation to occur, the CO2 concentration wasreduced from about 0.4% v/v to 0.080.12% v/v by placingtwo sterile plastic caps, each containing 5 g self-indicating,12.5 mm mesh soda lime granules, in each microcosm.The microcosms then were transferred to a 16 C cabinet;the soda lime was replaced when necessary, and the casinglayer irrigated with up to 15 mL sterile water after about 14d to maintain a matric potential of 21 to 22 kPa (Nobleet al 1999). The matric potential, pH and electrical con-ductivity of casing materials were determined according toNoble et al (1999).

The numbers of primordia $1 mm diameter were re-corded 21 d after transfer to 16 C; axenic or nonaxenicconditions were maintained in the microcosms for a further21 d to determine if primordia developed into mature spo-rophores, i.e., Stage 5 (Hammond and Nichols 1976) pileuswith broken veil and gills exposed.

Initiation on axenic and nonaxenic casing materials. Thetypes and sources of the 12 organic and inorganic casingmaterials used are shown in TABLE I. Peat and lignite weretoo acidic for mycelial growth; CaCO3, therefore, was added

622 MYCOLOGIA

TA

BL

EI.

Typ

es,s

ourc

esan

dpH

valu

esof

casi

ng

mat

eria

ls

Mat

eria

lT

ype,

mes

hsi

ze(m

m)

pHSo

urce

(UK

unle

ssst

ated

)R

efer

ence

Peat

Bar

kC

oir

Act

ivat

edch

arco

al

bulk

blac

k,10

dus

tco

mpo

sted

con

ifer

fin

es,

2du

stco

con

utpi

th,

2du

stgr

anul

ar,

48

4.5a

7.2

5.5

9.4

Har

tePe

atL

td,

Clo

nes

,Ir

elan

dM

elco

urt

Indu

stri

esL

td,

Tetb

ury,

Glo

cs.

Leg

aspi

Oil

Co.

Inc.

,M

akat

i,Ph

ilipp

ines

Sigm

a-A

ldri

chC

o.L

td,

Pool

e,D

orse

t

Nob

leet

al19

99A

llen

1975

Hay

es19

78L

ong

and

Jaco

bs19

74A

nim

alch

arco

alW

ood

char

coal

An

thra

cite

coal

Lig

nit

eZ

eolit

eVe

rmic

ulit

eR

ockw

ool

Silic

age

l

gran

ular

,4

8gr

anul

ar,

48

crus

hed

,2

dust

2du

stcl

inop

tilo

lite,

0.5

2br

own

,2

4re

cycl

ed,

8du

st0.

010

.04

10.3 8.8

7.6

4.5a

7.0

7.6

7.4

5.9

Fish

erSc

ien

tifi

c,L

ough

boro

ugh

,Lei

cs.

Fish

erSc

ien

tifi

c,L

ough

boro

ugh

,Lei

cs.

Cot

radi

ng

Ltd

,H

orsh

am,S

usse

xW

BB

Dev

onC

lays

Ltd

,N

ewto

nA

bbot

En

viro

nm

enta

lM

iner

als

Ltd

,C

ambr

idge

Dup

reVe

rmic

ulit

eL

td,

Her

tfor

dIs

over

Sain

t-Gob

ain

,Par

is,

Fran

ceFi

sher

Scie

nti

fic,

Lou

ghbo

roug

h,L

eics

.

Nob

lean

dG

aze

1995

Nob

lean

dG

aze

1995

aC

aCO

3w

asad

ded

at9%

v/v

tope

atan

dlig

nit

eto

prod

uce

aca

sin

gm

ater

ial

wit

hpH

7.4.

at 9% v/v to achieve a pH of 7.4 (TABLE I). The commercialwhite hybrid A. bisporus strain Sylvan A15 (Sylvan SpawnLtd, Peterborough, United Kingdom) was used. Each of sixreplicates consisted of four blocks, each containing threepairs of jars. Each pair of jars contained the same casingmaterial, one of each pair randomly selected to contain axe-nic material and other nonaxenic material. Casing materialswere allocated to blocks within replicates following an alphadesign, such that pairs of casing materials appeared withina block no more than twice. This allocation ensured thatthe precision of comparisons between pairs of casing ma-terials was as equal as possible across all pairs.

Data from this experiment were analyzed with the re-stricted-maximum likelihood (REML) approach (Thomp-son and Welham 2000), to allow for proper adjustment fordifferences between blocks within replicates. This approachalso enabled comparisons between treatments that allowedfor the variation in the precision of different paired com-parisons.

Axenic and nonaxenic initiation of A. bisporus strains.These A. bisporus strains were used: commercial strain A15,UK wild strains 96-4, 97-3 (HRI Culture Collection, Welles-bourne, Warwick, United Kingdom), American wild strainsJB10, JB157 (deposited as ARP116, ARP155, Sylvan SpawnLaboratory Inc.) and the developmental variant B430 (HRICulture Collection). Strains ARP116 and ARP155 are var.burnettii with mainly tetrasporic basidia (Callac et al 1993);all other strains are var. bisporus with mainly bisporic basid-ia. This experiment was arranged as a split-plot design withsix replicate blocks that each contained six pairs of jars.Each pair of jars contained the same strain, one with axenicand the other with nonaxenic casing material. Data fromthis experiment were analyzed using analysis of variance.

Flask culture. Composted substrate (Noble et al 1998)(400 g) was filled to a depth of 80 mm in 2 L Quickfitmultiadapter flasks (Fisher Scientific, Loughborough, Unit-ed Kingdom) and covered with 180 g of the above peat andCaCO3 casing material to a depth of 28 mm (surface area113 cm2). After autoclaving (axenic flasks only), the sub-strate was inoculated with two 5 mm plugs of A. bisporusmycelium grown on MEA, through holes made in the casinglayer. The flasks were sealed with bacterial air vents andkept at 25 C until mycelium became visible at the surfaceof the casing material, which then was watered with 50 mLsterile water. The flasks then were transferred to a room at18 C and connected to a filtered air flow of 15 L h21 tomaintain a CO2 concentration of 0.060.12% v/v. Primordiaand sporophores were recorded in the same way as for cul-ture in microcosms.

Initiation of A. bisporus strains in axenic and nonaxenic flaskculture. These A. bisporus strains were used: commercialstrain A15, UK wild strain 97-3, and the strain B430. Tworeplicate axenic and nonaxenic flasks were prepared witheach strain, arranged in a randomized block design.

Larger-scale nonaxenic culture. A. bisporus strains were cul-tured with polypropylene trays (60 3 40 3 18 [deep] cm)in a controlled-environment room, with cultural conditionsas described in Noble et al (1999). Each tray contained 9

623NOBLE ET AL: MUSHROOM PRIMORDIA

kg of the above composted substrate colonized with A. bis-porus strains after inoculation with 1% w/w rye grain spawnas mentioned above. The substrate was covered with a 28mm deep casing layer, which consisted of a 4:1 v/v mixtureof black peat and CaCO3 (Noble et al 1999).

Initiation of A. bisporus strains in larger-scale nonaxenic cul-ture. These A. bisporus strains were used: commercialwhite hybrid strains 2100 (Amycel Ltd, Burton, Staffordshi-re, United Kingdom) and A15, UK wild strains 96-4, 97-2,97-3, JFB II, JFB III (HRI Culture Collection), AMA4 (de-posited as ARP174, Sylvan Spawn Laboratory Inc.) andAmerican wild strains ARP116, ARP130, ARP155 (SylvanSpawn Laboratory Inc.) and the strain B430. StrainsARP130 and ARP174 are var. burnettii and var. bisporus re-spectively. Four replicate trays of each strain were prepared;two of the trays were used for recording the population ofprimordia and sporophores, while the casing layer in theother two trays was not disturbed and only the number ofsporophores was recorded. Each tray was divided into a 33 5 grid, each square measuring 10 3 10 cm. At each re-cording date, the casing layer in a randomly selected squarewas removed carefully so as not to disturb adjacent squares.The full depth of casing layer then was examined to recordthe number of primordia. After recording, all Stage 5 spo-rophores (Hammond and Nichols 1976) were removed andthe disturbed casing layer sample was replaced over thecompost in the square on the tray. Individual squares in thegrid were examined once. Primordia and sporophores wererecorded at 2 d intervals for 26 d after the first primordiawere visible. The numbers of primordia and sporophoreswere recorded by examination of destructively sampled cas-ing layer and recording the number of: (i) live primordia13 mm diameter (small primordia) (Wood 1976, Flegg1979); (ii) live primordia .3 mm diameter (large primor-dia); (iii) dead primordia $1 mm diameter; (iv) sporo-phores that had reached Stage 5 (Hammond and Nichols1976). A small number of primordia that were attached tothe base of sporophores were recorded in the above cate-gories. The numbers of primordia in the recording areaswere plotted against time. Estimates of primordia numberson days between the recording dates were obtained by cubicinterpolation between the observed points. Where this cu-bic interpolation gave negative estimates, values were resetto zero. A measure of the total production and survival ofprimordia was obtained by summing the daily counts (bothobserved and interpolated), giving an approximate area un-der the interpolated curve for each strain, referred to asthe integral number of primordia. These parameters ofgrowth curves were analyzed: (a) the area under the curve,i.e., the integral numbers of primordia in categories (i) to(iii) with time; (b) the peak value of primordial numbersin categories (i) to (iii).

Core (20 mm diameter) profiles of the casing layer alsowere examined with either a binocular microscope (310)or scanning electron microscope (SEM) (3353100) aftergold coating of frozen hydrated samples of casing material,with mycelium and primordia attached.

Two replicates of the 12 strains were arranged followinga latinized row-column alpha design. Each replicate was ar-

ranged in a single layer of trays (three trays wide by eighttrays deep), each strain being applied to a pair of adjacenttrays. Within each pair of trays, one was left undisturbed,while the casing of the other was sampled on occasions dur-ing crop development (these treatments applied at a sub-plot level). The design ensured that each strain only ap-peared at most once in each of four blocks of trays alongthe length of the experiment and that the precision of com-parisons between pairs of strains was as equal as possibleacross all pairs.

Data from this experiment first were analyzed using theREML approach to allow proper adjustment for differencesbetween rows and columns within replicates, but with littleevidence of any positional variation, a simpler analysis ofvariance was used.

Measurement of bacterial numbers in casing materials. In allthe experiments, bacteria were isolated from casing mate-rials on nutrient and Pseudomonas isolation agars (PIA)(Difco Laboratories, Detroit, Michigan) to determine thetotal bacterial populations as colony forming units per g dryweight casing material (cfu g21) and to estimate the pro-portion that were Pseudomonas spp. An estimate of the pro-portion of Pseudomonas spp. that were P. putida isolates wasobtained with these tests (Stanier et al 1966, Lelliott andStead 1987) on 30 cfus from the PIA isolation: (i) fluores-cence under UV light on Pseudomonas Agar F (PAF)(Merck Ltd., Poole, Dorset, United Kingdom); (ii) inabilityto hydrolyze gelatin; (iii) arginine dihydrolaze test positive.Results from these tests on Pseudomonas spp. isolates pre-viously were found to correspond to characterization by ri-botyping (Fermor et al 2000).

Bacterial and primordial populations showed evidence ofa mean-variance relationship. These populations respective-ly were subjected to a logarithmic or square root transfor-mation before analysis. All differences in the results sectionwere significant at P , 0.05, or if stated, P , 0.01 or 0.001.

RESULTS

Measurement of primordia. Both binocular light mi-croscopy and SEM observations showed that the ear-liest distinguishable initials were at least 0.5 mm indiameter. Mycelial knots or aggregates smaller thanthis consisted mainly of a disorientated mass of hy-phae (strains A15 and B430). These could not bedistinguished easily from the vegetative hyphae andhyphal cords (as observed by Umar and van Gri-ensven 1997). Visual observation of primordia 0.5mm or greater in diameter, in a larger casing materialsample, therefore was a more reliable method formeasuring the occurrence and amount of primordia.The development of primordia and sporophores wasslower in the microcosm tests than in the larger-scaleculture due to the lower temperature (16 C com-pared with 18 C) (Flegg 1979).

Primordia initiation on axenic casing materials. Thecasing material in microcosms in which initiation did

624 MYCOLOGIA

FIG. 2. Numbers of primordia $1 mm diameter (squareroot transformation) produced in different axenic and non-axenic casing materials. No primordia formed on vermicu-lite or on animal charcoal. LSDs are at P 5 0.05. LSD 1 (60df) is for comparing casing materials of the same sterilitytreatment; LSD 2 (159 df) is for comparing sterility treat-ments of the same casing material.

FIG. 3. Numbers of primordia $1 mm diameter (squareroot transformation), produced by different A. bisporusstrains on axenic and nonaxenic peat-based casing material.LSDs are at P 5 0.05. LSD 1 (25 df) is for comparing strainsof the same sterility treatment; LSD 2 (30 df) is for com-paring sterility treatments of the same strain

not occur became overgrown with mycelium i.e., stro-ma (Fletcher et al 1986), except in animal charcoal,where mycelial growth was poor. This might havebeen due to the high pH value of this casing material(TABLE I). After initiation on the casing materials,mycelial growth declined or stopped.

No primordia formed on axenic bark, coir, peat,rockwool, silica gel or vermiculite, or on animal char-coal (FIG. 2). Primordia formed on axenic and non-axenic charcoal (wood and activated), coal, ligniteand zeolite. Of the axenic materials, most primordiaformed on activated charcoal, and of the nonaxenicmaterials, more primordia formed on coir, activatedcharcoal and peat than on the other materials (P ,0.01). In axenic and nonaxenic microcosm treat-ments in which primordia were recorded, 12 pri-mordia developed into mature sporophores. Two pri-mordia formed in one of the six replicate micro-cosms of the nonaxenic vermiculite treatment, andone of these developed into a mature sporophore.

In the nonaxenic casing materials at the end of theexperiment, bacterial numbers were lowest in the lig-nite treatment (15 cfu g21) and greater in wood andanimal charcoal (3 3 1084 3 108 cfu g21) than inthe other materials (1057 3 107 cfu g21) (P , 0.01).There was no correlation between the numbers ofprimordia and the numbers (cfu g21) of total bacteriaor Pseudomonas spp. in different casing materials. Inpeat casing, 38% of the total bacterial populationwere estimated as being Pseudomonas spp. and 13%produced classical P. putida reactions, according tothe tests of Stanier et al (1966). In zeolite, bark, char-coal, coal and lignite casing materials, 1862% of thebacteria were estimated to be Pseudomonas spp. but

none of the isolated bacteria produced classical P.putida reactions. In coir and vermiculite casing ma-terials, none of the isolated bacteria grew on PIA andtherefore were unlikely to be Pseudomonas spp. Nobacteria were detected in the axenic microcosms atthe end of the experiment.

The electrical conductivities of activated charcoal,animal charcoal, coir and lignite (range 922 to 1486mS) were higher than those of the other casing ma-terials (range 58 to 452 mS). Coir and silica gel hadlower pH values and the charcoal sources had higherpH values than the other casing materials (TABLE I).However, there were no relationships between casingelectrical conductivity or pH and the initiation of pri-mordia.

Axenic and nonaxenic initiation of A. bisporusstrains. B430 was the only strain to produce pri-mordia on axenic peat-based casing material, buthere the number produced was less than on nonax-enic material (P , 0.001, FIGS. 1 and 3). The otherstrains produced only vegetative mycelium in axenicpeat-based casing material (FIGS. 1 and 3). The ru-dimentary primordia of B430 were similar to thoseshown by Elliott and Wood (1978). Between 35 ofthese primordia in each axenic or nonaxenic micro-cosm reached a diameter of 4 mm, but none of theprimordia formed differentiated tissues associatedwith developing sporophores (Wood 1979). For theother strains in nonaxenic microcosms, 12 primor-dia developed into mature sporophores. Strains A15and B430 produced more primordia than the otherstrains in nonaxenic casing, whereas strain ARP116produced the fewest (P , 0.01, FIG. 3).

Bacterial populations at the end of the experiment

625NOBLE ET AL: MUSHROOM PRIMORDIA

TABLE II. Numbers of primordia (square root and back-transformed values) and sporophores produced by A. bisporus strainsin axenic and nonaxenic flask cultures. Each value is the mean of two flasks

Strain Culture

Number of primordia $1 mm

Mean Square root Sporophores (stage 5)

A15

B430

97-3

axenicnonaxenicaxenicnonaxenicaxenicnonaxenic

016124

1780

48

(0.0)(12.6)(4.9)

(13.3)(0.0)(6.9)

070507

LSD (P 5 0.05, 6 df) (4.43) 4.7

(5 3 1052 3 106 cfu g21) were not affected by theA. bisporus strains in the nonaxenic microcosms. Ofthese populations, 60% were estimated to be Pseu-domonas spp. and 22% were estimated to be P. putida,according to the above tests. No bacteria were de-tected in the axenic microcosms at the end of theexperiment.

Initiation of A. bisporus strains in axenic and nonax-enic flask culture. As in the previous experiment,B430 was the only strain to produce primordia onaxenic peat-based casing material, and again none ofthese primordia grew beyond a diameter of 4 mmand the number produced was less than on nonax-enic material (TABLE II). The other two A. bisporusstrains produced only vegetative mycelium in axenicpeat-based casing material, but all three strains pro-duced mature sporophores in the nonaxenic flasks.

At the end of the experiment, bacterial popula-tions in the nonaxenic flasks were 6.6 3 1069.4 3106 cfu g21 and no bacteria were detected in the axe-nic flasks.

Larger-scale nonaxenic culture of A. bisporus strains.The first primordia were recorded 89 d (16 d forstrain ARP155) after the casing material was applied,and the first sporophores attained Stage 5 develop-ment (Hammond and Nichols 1976) after 1318 d(22 d for ARP155). A. bisporus (var. bisporus) strainshad 8098% bisporic basidia (50 observed basidia)and A. bisporus (var. burnettii) strains had 73100%tetrasporic basidia. This result agrees with Callac etal (1993), who observed 17100% bisporic basidiaand 024% tetrasporic basidia for var. bisporus and 08% bisporic and 54100% tetrasporic basidia for var.burnettii. Sporophores of all the wild strains hadbrown caps, with ARP116 and ARP130 being darkerthan JFB II, JFB III; the caps of 96-4, 97-2 and 97-3were pale brown. Strain B430 produced severely dis-torted sporophores with gill and spore tissue on thepileus surface, similar to rosecomb described by

Fletcher et al (1986), and split stipes (FIG. 4). How-ever, gill tissue was normal, producing viable spores.

Most primordia were formed close to the surfaceof the casing layer; very few primordia were formedin the lower 20 mm of the casing layer. Strains 2100,97-3 and ARP130 produced more small (13 mm)and total primordia than the other strains, both interms of the peak and integral numbers of primordia(TABLE III). Strains JFBII, ARP116, ARP155 andARP174 produced the smallest integral numbers ofprimordia and the first three of these strains also pro-duced the smallest peak numbers in primordia (TA-BLE III). Strain ARP130 produced the most large pri-mordia (.3 mm) and Strain ARP174 the fewest.Strain 97-3 produced the most dead primordia, andstrains 97-2, ARP155 and ARP174 the fewest (TABLEIII).

Differences in the timing of production of primor-dia by A. bisporus strains were observed. The peaknumbers of small, large and total primordia of strains2100, 96-4, 97-2, 97-3, ARP116, ARP130, B430, JFBIIand JFBIII occurred at or before the first flush ofsporophores. With strains A15, ARP155 and ARP174,the peak number of small, large and total primordiaoccurred after the first flush of sporophores. Thetiming of production of primordia and sporophoresof strains A15, B430, 97-3 and 96-4 is shown in FIG.5. With all strains, there was an increase in the num-bers of these categories of primordia before eachflush of sporophores, and there was a marked in-crease in the number of dead primordia betweendays 14 and 35 after the first recording of primordia(FIG. 5). The numbers of dead initials probably wereunderestimated because they discolored and decayedrapidly before being counted.

The integral number of small primordia producedby a strain closely correlated with the total integralnumber of primordia and integral number of deadprimordia it produced (r2 5 0.98 and 0.71; P ,0.001). Most strains produced 198240 mature spo-

626 MYCOLOGIA

FIG. 4. A. bisporus strain B430 in nonaxenic culture.

TABLE III. Peak and integral numbers of primordia and numbers of sporophores for different A. bisporus strains. Means ofpeak primordia numbers are back-transformed values. Numbers are per 0.1 m2 of tray area

Strain

Peak primordia

MeanSquare

root Sporophores

Integral numbers of primordia

13 mm .3 mm Dead Total

Sylvan A15Amycel 2100ARP116ARP130ARP155ARP174

367404149451168254

(19.1)(20.1)(12.2)(21.2)(12.9)(15.6)

14131612

16

409650801925465419751373

673496209

175523483

6927872822535818

545163622414665722451466

B430JFBIIJFBIII96-497-297-3

21526089

358240420

(14.7)(16.1)(9.4)

(18.8)(15.4)(20.5)

137

14161420

268233291990314126915838

368369724568429810

608639223625132

1360

365343352935432432517994

LSD (P 5 0.05, 11 df) (4.05) 2.9 2014.5 251.0 258.1 2101.4

rophores per tray. Strain 97-3 produced more, andstrains ARP130, ARP155 and JFBII produced fewermature sporophores per tray (TABLE III). There wereno correlations between the numbers of sporophoresand primordia for different strains or between num-

bers of primordia produced by the six A. bisporusstrains in microcosms and the same strains in larger-scale culture.

Bacterial populations in the casing material on thetrays were similar to those recorded on peat-based

627NOBLE ET AL: MUSHROOM PRIMORDIA

FIG. 5. Development of primordia and sporophores of four A. bisporus strains. Broken line: live primordia, solid line:dead primordia. The method used for curve fitting is described in the text. Days are counted from transfer to 18 C, 5 dafter applying casing material.

materials in the microcosms and flasks. They werenot affected by the A. bisporus strains (105107 cfug21), and 49% of the bacterial populations were es-timated to be Pseudomonas spp. and 20% were esti-mated to be P. putida, according to the above tests.

DISCUSSION

Primordia initiation on different casing materials.The results found here agree with Visscher (1979),who found no fruiting of A. bisporus on sterile peat,rockwool or organic wastes (or on sand, brown coal,glasswool, glass beads and chopped bricks), but ini-tiation occurred on these materials if they were treat-ed with an aqueous-filtered extract of normal casingmaterial. As here, Long and Jacobs (1974) and Woodand Blight (1983) found that initiation occurred onaxenic activated charcoal, although our experimentsalso showed that other carbonized materials (woodcharcoal, coal and lignite), as well as zeolite, couldstimulate initiation in axenic culture. The observa-tion made hereof mycelial growth declining orstopping after initiation on different casing materi-alsalso was made by Eger (1962) after the intro-duction of stimulatory bacteria into the casing layerof sterile cultures of A. bisporus.

The results here indicate that different casing ma-terials support different populations of bacteria, al-though these were not related to the numbers of ini-tials formed. The proportions of fluorescent Pseudo-monas spp. found in peat casing (1656%) corre-spond to those of Miller et al (1995), who found thatthey were 1441% of the total bacteria present. Hayesand Nair (1974) found that the proportion of Pseu-domonas spp. increased from 34% of the total bac-teria present at the time peat casing was applied to95% of the total bacteria at the end of a mushroom

crop. Although coir and vermiculite did not appearto support the growth of Pseudomonas spp., othergenera of bacteria, Bacillus and Alcaligenes, havebeen shown to stimulate initiation of A. bisporus(Park and Agnihotri 1969, Fermor et al 2000).

Previous attempts to isolate the putative inhibitorsto initiation from activated charcoal have been con-founded by the large number of compounds ad-sorbed and identified (Grove 1981, Wood and Blight1983). Of these, 1-octen-3-ol was found to inhibit pri-mordial formation in plate culture. Here, more pri-mordia were formed on axenic activated charcoalthan on other axenic materials; this might be due tothe relatively greater adsorption action of the mate-rial (Long and Jacobs 1974). Chemical analysis by gaschromatography-mass spectrometry of adsorbent,stimulatory and nonstimulatory casing materialsmight identify differences in compounds present. Ini-tiation on zeolite might provide the opportunity toselectively adsorb putative fruiting inhibitors, using arange of synthetic zeolites with different adsorptioncharacteristics (Eitel 1966).

Primordia initiation of different A. bisporus strains.Tschierpe (1983) and Fritsche and Sonnenberg(1988) recorded differences in the number of pri-mordia produced by different A. bisporus strains un-der similar cultural conditions, and Flegg (1979)found that this proportion could be altered by tem-perature. Here, an estimate of the proportion of pri-mordia that developed into sporophores could be ob-tained by dividing the peak number of total primor-dia (including primordia that developed into sporo-phores before the peak) by the numbers ofsporophores that developed (TABLE II, FIG. 5). Theseproportions are overestimates because primordia thatdeveloped between the peak value and Day 35 are

628 MYCOLOGIA

not included and the number of dead initials mightbe underestimates as previously mentioned. In moststrains, 2.76.1% of primordia developed into spo-rophores. In strains ARP116 and JFBIII, these figureswere higher (10.4 and 15.6%), and in strains ARP130and ARP155 they were lower (0.2 and 1.2%). Theproportions found here correspond to those of Flegg(1979), who found that 3.2% of primordia of thestrain D649 developed into sporophores at a crop-ping temperature of 16 C. He also suggested that thesporophores, which develop during a period of sev-eral weeks, arise from mycelial aggregates formedduring the period immediately after casing is appliedto the start of the first flush. This observation agreeswith the results in this work, although in some strains(97-3 and ARP174) significant numbers of new pri-mordia developed after the first flush of sporo-phores. Wannet et al (1999) found that all hyphalknots were formed before the first period of aggre-gate development on MEA. At each period of aggre-gate release or flush, about 25% of these knots de-veloped synchronously into aggregates.

Strain ARP116 did not produce primordia on anaxenic casing layer, although Wannet et al (1999)found that it produced primordia on axenic MEA. Innonaxenic culture, the production of primordia andsporophores of ARP116 followed a similar pattern tothat in other A. bisporus strains. The developmentalvariant strain B430 was capable of initiating rudimen-tary primordia on axenic peat-based casing materialbut, similar to the observations of Elliott and Wood(1978) and Wannet et al (1999), the primordia wereunable to grow beyond Stage 2 growth. Flegg andWood (1985) found that light and electron micro-scope sections of primordia produced by B430 onaxenic media showed them to be indistinguishablefrom normal primordia that were capable of furtherdifferentiation. As here, the rudimentary primordiaconsisted mainly of a disorientated mass of hyphae.The formation of normal primordia and the subse-quent numbers of sporophores of B430 in nonaxenicculture were similar to those of other A. bisporusstrains. This result contrasts with that of Elliott andWood (1978), who were unable to demonstrate con-clusively that B430 was capable of producing maturesporophores in vivo.

In flask culture, B430 primordia developed intomature sporophores using nonaxenic casing materialbut not with axenic casing material. This responseindicates that the mechanism for producing rudi-mentary primordia in axenic casing material differsfrom that in nonaxenic casing material, where pri-mordia capable of further differentiation are pro-duced. The lack of sporophores of B430 in nonax-enic microcosms might have been due to the limited

substrate weight (30 g compared with 400 g in flasks).Wood (1979) found that A. bitorquis strains producedhyphal aggregates on defined axenic media but, sim-ilar to B430, they were incapable of further differ-entiation. San Antonio and Peerally (1979) found norelationship between primordia formation of 22strains of A. bitorquis on axenic media and sporo-phore production in culture on peat casing and com-post. This report agrees with the results found herefor primordia formation and sporophore productionof A. bisporus strains on casing material and compostin nonaxenic culture. Rainey et al (1990) developedan in vitro Petri dish bioassay using A. bitorquis(strain W19) to determine the involvement of Pseu-domonas spp. in stimulating initiation. However, sub-sequent tests (Fermor et al 2000) showed that thestimulatory response of W19 in the bioassay did notcorrespond with in vivo results obtained with A. bis-porus in the microcosms used here.

This work supports the hypothesis (Flegg andWood 1985) that inhibitors prevent initiation of pri-mordia and that the putative inhibitors either can bemetabolized by the casing microbiota (P. putida) orbe adsorbed by certain casing materials. These pu-tative inhibitors do not prevent the formation of ru-dimentary primordia in strain B430 but prevent theformation of primordia capable of differentiation.Further work is needed to determine whether thenumbers of primordia and the proportion that de-velop into sporophores are related to the productionof putative inhibitors by the mycelium of different A.bisporus strains, or their metabolism or adsorbtion bythe casing layer.

ACKNOWLEDGMENT

This work was financed by the Department for Environ-ment, Food and Rural Affairs.

LITERATURE CITED

Allen PG. 1975. Casing variablesplastic tunnels. Mush-room J 37:2229.

Callac P, Billette C, Imbernon M, Kerrigan RW. 1993. Mor-phological, genetic, and interfertility analyses reveal anovel tetrasporic variety of Agaricus bisporus from theSonoran Desert of California. Mycologia 85:835851.

Eger G. 1961. Untersuchungen uber die Function der Deck-schicht bei der Fruchtkorperbildung des Kulturcham-pignons, Psalliota bispora Lange. Archiv fur Mikrobiol-ogie 39:31334.

. 1962. Untersuchungen uber die Funktion derDeckschicht bei der Fruchtkorperbildung des Kultur-champignons. Mushroom Sci 5:314320.

Eitel W. 1966. Silicate science, IV Hydrothermal silicate sys-tems. New York: Academic Press. 617 p.

629NOBLE ET AL: MUSHROOM PRIMORDIA

Elliott TJ. 1985. Spawn-making and spawns. In: Flegg PB,Spencer DM, Wood DA, eds. The biology of the culti-vated mushroom. Chichester, UK: John Wiley & Sons.p 131139.

, Wood DA. 1978. A developmental variant of Agar-icus bisporus. Trans British Mycological Soc 82:373381.

Fermor T, Lincoln S, Noble R, Dobrovin-Pennington A.2000. Microbiological properties of casing. In: van Gri-ensven LJLD, ed. Science and cultivation of edible fun-gi. Rotterdam: Balkema. p 447454.

Flegg PB. 1979. Effect of temperature on sporophore ini-tiation and development in Agaricus bisporus. Mush-room Sci 10:595602.

, Wood DA. 1985. Growth and fruiting. In: Flegg PB,Spencer DM, Wood DA, eds. The biology of the culti-vated mushroom. Chichester, UK: John Wiley & Sons.p 141177.

Fletcher JT, White PF, Gaze RH. 1986. Mushroomspestand disease control. Newcastle upon Tyne, UK: Inter-cept. 156 p.

Fritsche G, Sonnenberg ASM. 1988. Mushroom strains. In:van Griensven LJLD, ed. The cultivation of mush-rooms. Rustington, Sussex, UK: Darlington MushroomLaboratories. p 101123.

Grove JF. 1981. Volatile compounds from the mycelium ofthe mushroom Agaricus bisporus. Phytochem 20:20212022.

Hammond JBW, Nichols R. 1976. Carbohydrate metabolismin Agaricus bisporus (Lange) Sing.: changes in solublecarbohydrates during growth of mycelium and sporo-phore. J Gen Microbiol 93:309320.

Hayes WA. 1978. Progress in the development of an alter-native casing medium. Mushroom J 78:266271.

, Nair NG. 1974. Effects of volatile metabolic by-prod-ucts of mushroom mycelium on the ecology of the cas-ing layer. Mushroom Sci 9(I):259268.

, Randle PE, Last FT. 1969. The nature of the micro-bial stimulus affecting sporophore stimulation in Agar-icus bisporus (Lange) Sing. Ann Appl Biol 64:177187.

Lelliott RA, Stead DE. 1987. Methods for the diagnosis ofbacterial diseases of plants. Methods in plant patholo-gy. Vol. 2. Oxford, UK: Blackwell Scientific Publica-tions. 216 p.

Long PE, Jacobs L. 1974. Aseptic fruiting of the cultivatedmushroom Agaricus bisporus. Trans British MycologicalSoc 63:99107.

Miller N, Gillespie JB, Doyle OPE. 1995. The involvementof microbiological components of peat based casingmaterial in fructification of Agaricus bisporus. In: ElliottTJ, ed. Science and cultivation of edible fungi. Rotter-dam: Balkema. p 313321.

Noble R, Gaze RH. 1995. Properties of casing peat typesand additives and their influence on mushroom yieldand quality. In: Elliott TJ, ed. Science and cultivationof edible fungi. Rotterdam: Balkema. p 305312.

, , Willoughby N. 1998. A high yielding sub-strate for mushroom experiments: formula 3. Mush-room J 587:2728.

, Dobrovin-Pennington A, Evered CE, Mead A. 1999.Properties of peat-based casing soils and their influ-ence on the water relations and growth of the mush-room (Agaricus bisporus). Plant and Soil 207:113.

Park JY, Agnihotri VP. 1969. Bacterial metabolites triggersporophore formation in Agaricus bisporus. Nature(Lond.) 222:984.

Rainey PB. 1989. The involvement of Pseudomonas putidain sporophore initiation of the cultivated mushroomAgaricus bisporus. [PhD Dissertation]. Canterbury, NewZealand: University of Canterbury.

, Cole ALJ, Fermor TR, Wood DA. 1990. A modelsystem for examining involvement of bacteria in spo-rophore initiation of Agaricus bisporus. Mycological Res94:191195.

San Antonio JP, Peerally A. 1979. Growth, primordia for-mation, and fruiting of twenty-two strains of Agaricusbitorquis. Mushroom Sci 10:587594.

Stanier RY, Palleroni NJ, Doudoroff M. 1966. The aerobicPseudomonads: a taxonomic study. J Gen Microbiol 43:159271.

Thompson R, Welham SJ. 2000. REML analysis of mixedmodels. In: Payne RW, ed. The guide to GenStat, Part2: statistics. Oxford, UK: VSN International. p 413503.

Tschierpe HJ. 1983. Environmental factors and mushroomstrains. Mushroom J 132:41729.

Umar MH, van Griensven LJLD. 1997. Morphologisch on-derzoek: Levensfasen van een champignon. De Cham-ignoncultuur 41:4750.

Visscher HR. 1979. Fructification of Agaricus bisporus (Lge.)Imb. in relation to the relevant microflora in the casingsoil. Mushroom Sci 10(1):641664.

Wannet WJB, Aben EMJ, van der Drift C, van GriensvenLJLD, Vogels G, Op den Camp HJM. 1999. Trehalosephosphorylase activity and carbohydrate levels duringaxenic fruiting in three Agaricus bisporus strains. Cur-rent Microbiol 39:205210.

Wood DA. 1976. Primordium formation in axenic culturesof Agaricus bisporus (Lange) Sing. J Gen Microbiol 95:313323.

. 1979. Studies on primordium formation initiationin Agaricus bisporus and Agaricus bitorquis (syn) edulis.Mushroom Sci 10:565586.

, Blight M. 1983. Sporophore initiation in axenic cul-ture. Rep Glasshouse Crops Res Inst 1981:140.