Exp Appl Acarol

DOI 10.1007/s10493-009-9326-3

Abundance and distribution of Microdispus lambi (Acari:

Microdispidae) in Spanish mushroom crops

Marı́a-Jesús Navarro • Francisco-José Gea •

L. Adriana Escudero-Colomar

Received: 12 March 2009 / Accepted: 28 September 2009

Ó Springer Science+Business Media B.V. 2009

Abstract The myceliophagous mite Microdispus lambi has become a veritable plague
since 1996, when it was first observed in Spanish mushroom crops, and is now causing
substantial economic losses, particulary in spring and summer. This study looks at seasonal
variation of the pest, its distribution on commercial farms and the population development
during the crop cycle of the common white mushroom, Agaricus bisporus. Over a period of
18 months, 24 consecutive mushroom crop cycles were monitored and a total of 24 spawn
and 960 substrate samples were analysed. We found that it is usually the substrates in the
growing rooms that are infested, most commonly the compost. In many cases, the pest can
be detected when the first ‘flush’—i.e., mushroom growth surge, with weekly periodicity—
is harvested, although damage does not become evident until the third flush. Mites were
detected at the back of the mushroom growing room and, to a lesser extent, near the access
door.

Keywords Agaricus bisporus
Mushroom pest
Myceliophagous mite

Population dynamics

Introduction

The mushroom mite, Microdispus lambi (Krczal), was described when working on
mushroom crops in New Zealand (Krczal 1964). This small, specialist mite damages
mushrooms by feeding on the mycelia, and it can develop and reproduce only on Agaricus
species (Clift and Toffolon 1981a; Gao and Zou 2001). It is also found in New South

M.-J. Navarro
F.-J. Gea (&)

Centro de Investigación, Experimentación y Servicios (CIES) del Champiñón, c/Peñicas s/n,
16220 Quintanar del Rey, Cuenca, Spain
e-mail: fjgea.cies@dipucuenca.es

L. A. Escudero-Colomar
IRTA Mas Badia Agricultural Experimental Station, Crop Protection, La Tallada D’Empordà,
17134 Girona, Spain

123
 Exp Appl Acarol

Wales, Australia, where it has been responsible for sizeable economic losses (Clift and
Toffolon 1981a, b), and in Shangai Province, Eastern China, where it was considered the
most serious pest in the region (Gao et al. 1986).
In Spain, the first signs of damage by this mite were observed in the spring and summer
of 1996 in commercial farms in Castilla-La Mancha, S.E. Spain (Ferragut et al. 1997) and
afterwards, in La Rioja, N.E. Spain, in 1997. The pest is now widely entrenched in both
these important mushroom growing regions. During spring and summer, mite populations
lead to the slow disappearance of the mycelium and substantial yield losses. When density
is low, the mite may easily go unnoticed by growers but not for long because it rapidly
increases in numbers and moves to the surface of the growing substrate and then onto the
mushroom cap. When this stage is reached, yield losses may be extremely high, sometimes
leaving farmers with no mushrooms to harvest at all. During autumn and winter, mite
populations decrease and although individuals may be detected from the third flush
onwards; at this stage they cause less damage than in other seasons.
An understanding of how M. lambi infests and spreads in a mushroom culture is
essential in order to develop effective control measures. In Australia, M. lambi was found
to be phoretic on sciarid and phorid flies (Clift and Larsson 1987), so an effective control
of the Diptera would contribute to the reduction of mite populations. In Shanghai, con-
taminated mushroom spawn was a major source of mite infestation (Wu and Ma 1988; Wu
and Zhang 1993a) but it was found that the problem could be eradicated by freezing the
spawn at -10°C for 24 h (Wu and Zhang 1993a, b).
Since little or nothing was known about the population dynamics of M. lambi in Spain,
we set out to analyse the spawn and growing substrates (compost and casing layer) as
possible sources of contamination. We also studied the seasonal dynamics of mite popu-
lations and how they spread inside the growing rooms during the crop cycle. The findings
will help establish where and how mites initiate colonization of the mushroom crop.

Materials and methods

Twenty-four consecutive crop cycles of the common white mushroom, Agaricus bisporus
Lange (Imbach), were monitored over a period of 18 months, from March 1998 to August
1999 (four crop cycles per season), in 24 mushroom facilities. Each crop was located in a
growing room (35 9 2.5 9 2 m), containing 500 mushroom growing units (60 9
40 9 20 cm), stacked at three heights. All growing rooms had a door for access at the front
and an opening for ventilation at the rear. Each crop was entirely grown in a single room
and finished after 70 days. During this time, depending on the crop stage, relative humidity
inside the room varied between 100 and 82% and temperature between 22 and 16°C.
Each mushroom growing unit contained two parts: the compost and the casing layer. All
units with pasteurised compost were seeded with mycelium of A. bisporus (spawn) and
kept inside the growing rooms for 20 days, at which point a casing layer of soil and black
peat was applied to the top of the growing unit. Fifteen days later the first flush began to
appear. Mushrooms were harvested daily during five flushes (35 days approx.).

Possible sources of infestation

Possible sources of infestation of the mites are the compost, the spawn and the casing
layer. Each was examined separately before each production cycle for presence of mites.

123
Exp Appl Acarol

For compost, six 20 g samples were collected at randomly chosen points from each
pasteurisation chamber, immediately after opening the chamber. The spawn samples were
collected in the composting plant before being fed into the seeder. Samples of casing
layer were collected from each farm before being applied to the compost. Twenty-four
spawn samples were analysed after being incubated in test tubes for 21 days at 25°C, with
four repetitions per sample. The tubes were then examined by binocular microscope to
check for the presence of mites. To extract the mites from the 960 samples of selected
compost and casing layer, each sample was submitted, in duplicate, to an extraction
process using Berlese-Tullgren funnels. The mites were collected in a solution composed
of 60% ethanol and 10% glycerine and placed in Petri dishes, where they were identified
and counted, following the method proposed by Solomon (1945). Because of the high
numbers of mites found in the samples, five counts were made for each one in order to
obtain the mean.

Spatial distribution of Microdispus lambi in the farm

A sampling calendar was established for each of the 24 cycles studied and samples were
collected at six points in time: before sowing (filling: day 0), after incubation (approx. day
20), after the primordia had formed in the upper surface of the growing unit (induction, day
30 approx.), and after harvesting the first flush (F1, day 41 approx.), third flush (F3; day 56
approx.) and fifth flush (F5; day 70 approx.). Samples were taken from three growing units
at intermediate height, in three locations per room: near the access door (door), in the
middle of the room and near the rear ventilation opening at the back. Compost and casing
layer samples were taken separately from each growing unit. The compost samples were
taken from the top 5 cm, which had been shown to contain most mushroom pests (San
Antonio et al. 1974). A total of 36 samples were taken from each crop cycle and submitted,
in duplicate, to the extraction and counting method described above. A data logger
(Tinytalk, Gemini Data Loggers, Chichester, UK) recorded ambient and substrate tem-
peratures during each cultivation cycle.

Statistical analysis

A non-parametric Kruskal–Wallis test was used and the Mann–Whitney U test was
employed to re-check the Kruskal–Wallis results. The spatial distribution inside the
growing room during the crop stages was drawn as a tri-dimensional graph. Data was
analyzed using the SAS program (SAS Institute 2002–2003).

Results

Study of the infestation sources

Microdispus lambi was not detected in any spawn samples, nor was there any sign of this
mite in the casing layer or in the recently pasteurised compost samples before filling
(Table 1). In this last source, however, two other mite species were detected, Bakerdania
mesembrinae (Canestrini) (Siteroptidae) and Histiostoma spp. (Histiostomatidae).

123
 Exp Appl Acarol

Table 1 Mites detected in spawn, compost and casing layer before filling
Material n Microdispus lambi Bakerdania mesembrinae Histiostoma sp.

Spawn 24 – – –
Pasteurised compost 72 – ? (1)a ? (1)
Casing layer 24 – – –
n number of samples analysed in each case
a
In parentheses: the number of samples in which individuals of the species were detected

Distribution of Microdispus lambi

The abundance of M. lambi, the total number detected in the 36 samples collected during
one complete crop, was extremely variable. In some cases M. lambi populations reached
very high levels of 8,500–21,200 mites per cycle, but most cycles showed levels of
infestation of \2,200 mites per cycle. Figure 1 shows the mean number of mites detected
in each stage of a standard mushroom cycle during each season. One can see that in spring,
summer and autumn the population dynamics were similar, but in the winter of 1999
almost no adult mites were detected in the growing rooms. Statistical analysis indicates a
highly significant difference between the winter and the other seasons (v2 = 67.87, df = 3,
P \ 0.0001).
Importantly, however, the level of infestation of M. lambi was related to the time at
which the mites were detected in each growing room. The earlier the detection, the greater
the likelihood of a major subsequent infestation. In cycles where mites were detected
during the initial stages of spawn running, rapid proliferation was encouraged, reaching
levels of 100 mites at F1, 2,000 at F3 and 14,000 at F5. In other cycles, where M. lambi
was detected for the first time after harvest, population growth was slower, with levels of
250 at F3 and 3,700 mites at F5. In the cycles where mites were first detected during the
final stages of the crop, the number of mites found was very low, (200 mites at F5). The
temperature inside the growing units varied depending on the stage of the crop cycle
(Fig. 2). In the first 16–18 days before the application of the casing layer, the temperature

3,5
Abundance of M. lambi

2,5

1,5

0,5

0
Filling Incubation Induction F1 F3 F5

Crop cycle

Fig. 1 Seasonal abundance (log) of Microdispus lambi during a standard crop cycle in 1998 and 1999:
(diamond) spring 1998; (square) summer 1998; (triangle) autumn 1998; (9) winter 1999; (circle) spring
1999; (-) summer 1999

123
Exp Appl Acarol

29

27

Temperature (°C)
25

23

21

19
Incubation Casing Induction Harvest
17

15
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
Crop cycle (days)

Fig. 2 Temperature registered inside the growing units and the determination of time of infestation
(square) detection of mites; (circle) infestation time

rose from 22 to 26–28°C; in the following 8 days, before induction began, the temperature
fell to 20–22°C; lastly, during the 6 days until F1 harvest began and during all the flushes,
the substrates remained at about 16–18°C.
Distribution of the mites inside the room and their development during the crop cycle is
illustrated in Fig. 3 which shows the mean numbers of M. lambi detected at each stage and
location in the growing room. Analysis of compost and casing layer samples at the initial
stages (incubation and induction) indicates the absence of M. lambi in the rooms. However,
from this time onwards, the percentage of growing crops in which the pest was detected
gradually increased, as did the number of individuals per sampling. At F1, the pest was

Fig. 3 Population development of Microdispus lambi during the culture cycle in each location on the farm.
Light shaded\260 mites; medium shaded\547 mites; dark shaded\834 mites; heavily dark shaded\1,121
mites; full dark shaded \1,408 mites

123
 Exp Appl Acarol

significantly more present at the back (on average 21.23 mites per sample) than at other
sites (v2 = 7.61, df = 2, P \ 0.02); the back, near the opening for ventilation, is where the
infestation was generally first detected. After F3 the mean number of mites detected at the
back reached 317.35 individuals, showing a highly significant statistical difference
(v2 = 32.69, df = 2, P \ 0.0001) from all previous stages. Moreover, at F3 the mite also
began to be detected at the door (on average 11.91 mites). At F5 mites were present
throughout the installation: on average, 1,407.81 mites at the back and 142.52 near the
entrance (door). Statistical analysis indicated significant differences (v2 = 30.09, df = 2,
P \ 0.0001) between F5 and all the preceding stages. There were more mites at the back at
any moment during the cycle, than at either of the other two locations, door or middle
(v2 = 38.92, df = 2, P \ 0.0001). To summarize, the M. lambi population was first
detected at the back, increased as the cycle progressed, finally colonising the whole farm,
but numbers were always highest at the back of the room.
Another objective of the study was to improve our knowledge of the infestation dynamics
in the compost and casing layer. After F1, mites were detected principally in the compost (on
average 14.66 mites per compost sample), although in very few crop cycles they were also
present in the casing layer (on average 0.36 mites per sample). This 40-fold difference
between the two substrates was highly significant (v2 = 35.45, df = 1, P \ 0.0001). The
relative difference between compost and casing layer remained throughout the crop cycle,
although it became smaller: at F5 the number of mites found in the compost exceeded the
number in the casing layer by only 39 (on average 912.96 vs. 265.16 mites per sample).

Discussion

Microdispus lambi is present on mushroom farms in Castilla-La Mancha (Spain)

throughout the year, although the incidence declines markedly over the winter, probably
due to the cooler temperatures. These results are different from the findings of Clift and
Toffolon (1981b), who detected no seasonal variation in the incidence of M. lambi in
Australian mushroom farms. The bioclimatic classification of this mushroom farm area of
Spain is known as Mediterranean Pluviseasonal Oceanic Upper Mesomediterranean Low
Dry, with average and absolute minimum temperatures for the coldest month of the year of
-0.6 and -22.8°C, respectively. In contrast, Sydney belongs to the bioclimatic classifi-
cation known as Temperate Hyperocenic Low Termotemperate Low Humid, with average
and absolute minimum temperatures for the coldest month of 12.1 and 0.0°C, respectively
(Rivas-Martı́nez and Rivas-Saenz 1996–2009). Perhaps these large differences in tem-
perature parameters explain the different population development of the pest in the Spanish
and Australian studies.
Unlike observations made in China (Wu and Zhang 1993a), spawn cannot be considered
a common M. lambi infestation pathway for mushroom farms in Spain, because no mites
were found in our 24 spawn samples. Nor were they found in our 72 compost samples, or
in any of the 24 samples of casing layer analyzed, so they may not be considered a source
of contamination.
The level of infestation was related to the temperature inside the growing unit. In cycles
where mites were detected during the spawn running—a phase characterised by high
germination temperatures of 24–28°C, which is very close to the optimum temperature
required for the mite’s population growth (Gao and Zou 2001)—the level of infestation
was much higher than when mites were first detected at harvest, when the temperature in
the growing units was between 16–18°C.

123
Exp Appl Acarol

Microdispus lambi in A. bisporus culture was occasionally detected earlier than the first
flush (F1), although yield losses were not noticed until the third flush (F3). Damage was
therefore limited because at that point more than 70% of the production had already been
harvested. Clift and Toffolon (1981b) recorded similar findings in Australian mushroom
houses, although in their survey, which used pitfall traps 5 cm sunk into the casing layer,
M. lambi was not detected before the third flush and rarely swarmed at the end of cropping.
In our study, M. lambi only swarmed twice, both times at the fifth flush (F5). In most cases
the mites could only be recognised upon close inspection, when they perched in exposed
positions on the mushroom surface, rearing up on their hind legs with their other three pairs
extended in order to attach to a passing vector. In our study, mean counts of M. lambi in
compost and casing layer indicated that initial infestation was in the compost and later
moved to the casing layer. This behaviour could imply programmed dispersal, as described
by Gao and Zou (2001): after females mated, only some of them crawled directly down
into the compost, to feed and reproduce. The remainder crawled to the top of the culture
and swarmed on the surface of the substrate even though inside the compost food was
plentiful and the temperature favourable.
Mean generation time of M. lambi is 17–23 days at 18°C or 10 days at 28°C (Clift and
Toffolon 1981a; Gao and Zou 2001). On this basis, if initial infestation had occured at
spawn running, mites would have swarmed at the end of the cycle (Clift 1987). However,
because the initial infestation occurred after the application of the casing layer, the first
generation was detected during the first flush, the second during the third flush, and the
third generation during the fifth flush, and no swarming was detected. This explains why
the compost is the first substrate to manifest the pest: during the first few days after the
application of the casing layer, only the compost contains a food source for the mites, i.e.,
the mushroom mycelium.
The distribution of the pest inside the growing rooms (higher concentrations towards the
back and the door, and lower in the middle) suggests that the ventilation opening and the
entrance door provide the mites with access to the growing cycle, whereas the fact that the pest
was not detected in any of the samples of compost, spawns or casing layer before filling,
indicates that the air and/or some flies that are common in the growing room, may act as
vectors.
The application of the Council Directive 91/14/CEE relating to the commercialization of
insecticides has resulted in the withdrawal of more than two-thirds of the existing active
ingredients in the EU, leaving very few compounds to be used in mushroom crops. For this
reason, farm hygiene must be regarded as the most important tool to control pests and diseases
in this crop, and the situation with M. lambi is no exception. This first study of M. lambi in
Spain suggests that the potential of this pest to damage A. bisporus would be substantially
reduced if infestation could be prevented until after the induction stage has begun. However,
more studies are being carried out to establish with certainty the role that these vectors play as
an infection source and as a focus for the dispersion of the pest inside the growing rooms.

Acknowledgments Funding for the research was provided by the Ministry of Agriculture, Fisheries and
Food of Spain, I.N.I.A.-Proyects SC98-011-C3 and RTA01-017-C3.

References

Clift AD (1987) A computer model to simulate the occurrence and increase in numbers of several insect and
mite pests of cultivated mushrooms in Australia. In: Wuest PJ, Royse DJ, Beelman RB (eds) Culti-
vating edible fungi. Elsevier, Amsterdam, pp 371–378

123
 Exp Appl Acarol

Clift AD, Larsson SF (1987) Phoretic dispersal of Brennandania lambi (Kraczal) (Acari: Tarsonemida:
Pygmephoridae) by mushroom flies (Diptera: Sciaridae and Phoridae) in New South Wales, Australia.
Exp Appl Acarol 3:11–20
Clift AD, Toffolon RB (1981a) Biology, fungal host preferences and economic significance of two pyg-
mephorid mites (Acarina: Pygmephoroidae) in cultivated mushrooms, New South Wales, Australia.
Mushroom Sci 11:245–253
Clift AD, Toffolon RB (1981b) Insects and mites associated with mushroom cultivation on three com-
mercial farms near Sydney N.S.W., Australia. Mushroom Sci 11:537–549
Ferragut F, Gea FJ, Garcı́a-Morrás JA (1997) El ácaro del champiñón Brennandania lambi (Krczal) (Acari:
Pygmephoriodea): introducción en España, importancia económica y separación de especies afines.
Bol San Veg Plagas 23(2):301–311
Gao JR, Zou P (2001) Biology, life table and host specificity of the mushroom pest, Brennandania lambi
(Acari: Pygmephoroidea). Exp Appl Acarol 25:187–202
Gao JR, Zou P, Ma EP (1986) Two new records of mushroom pygmephorid mites from China (Acari:
Pygmephoridae). Acta Agric Shanghai 2(3):27–32
Krczal H (1964) Pygmephorus lambi, eine neue Pyemotide aus Champignonkulturen. Zool Anz Bd 172,
Heft 4
Rivas-Martı́nez S, Rivas-Saenz S (1996–2009) Worldwide Bioclimatic Classification System. Phytosocio-
logical Research Center, Spain. http://www.globalbioclimatics.org
San Antonio JP, Civerolo EL, Lawson RH (1974) Diagnosis of pests, diseases, and other problems of the
cultivated mushroom, Agaricus bisporus. Mushroom Sci 9:109–123
SAS Institute (2002–2003) The SAS System Version 9.1 for Windows. SAS Institute, Cary, NC
Solomon ME (1945) Tyroglyphid mites in stored products. Methods for the study of population density. Ann
Appl Biol 32:71–75
Wu J-F, Ma E-P (1988) Studies on biological characters of Brennandania lambi (Krczal) the most harmful
mite for mushroom production in Shangai region. Acta Agric Shangai 4(3):41–46
Wu J, Zhang Z-Q (1993a) Host feeding, damage and control of the mushroom pest, Brennandania lambi
(Acari: Pygmephoroidea) in China. Exp Appl Acarol 17:233–240
Wu J, Zhang Z-Q (1993b) Control of Brennandania lambi (Acari: Pygmephoroidea) by freezing: evaluation
of its efficacy and effects on mushroom growth and yield. Exp Appl Acarol 17:531–540

123