ARTICLE IN PRESS

B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x

available at www.sciencedirect.com

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

Mushroom picking does not impair future harvests – results

of a long-term study in Switzerland

Simon Eglia,*, Martina Petera, Christoph Buserb, Werner Stahelb, François Ayera
a
Research Department Landscape, Swiss Federal Research Institute WSL, Zuercherstrasse 111, CH-8903 Birmensdorf, Switzerland
b
Department of Mathematics, Swiss Federal Institute of Technology, CH-8092 Zurich, Switzerland

A R T I C L E I N F O A B S T R A C T

Article history: Forest fungi not only have important functions within the forest ecosystem, but picking
Received 10 August 2005 their fruit bodies is also a popular past time, as well as a source of income in many devel-
Received in revised form oping and developed countries. The expansion of commercial harvesting in many parts of
27 October 2005 the world has led to widespread concern about overharvesting and possible damage to fun-
Accepted 31 October 2005 gal resources. In 1975, we started a field research project to investigate the effects of mush-
room picking on fruit body occurrence. The three treatments applied were the harvesting
techniques picking and cutting, and the concomitant trampling of the forest floor. The
Keywords: results reveal that, contrary to expectations, long-term and systematic harvesting reduces
Mushroom picking neither the future yields of fruit bodies nor the species richness of wild forest fungi, irre-
Trampling spective of whether the harvesting technique was picking or cutting. Forest floor trampling
Fungi does, however, reduce fruit body numbers, but our data show no evidence that trampling
Species richness damaged the soil mycelia in the studied time period.
Conservation
2005 Elsevier Ltd. All rights reserved.

1. Introduction at about US $ 1.67 billion (Watling, 1997). The ‘‘soil expecta-

tion value’’ for forest fungi (e.g. US dollars/ha/year) is on cer-
Forest fungi perform important functions within the forest tain forest sites as high as that for timber (Alexander et al.,
ecosystem. Saprobic species decompose organic matter and 2002).
ectomycorrhizal species enhance nutrient acquisition, im- According to a recently published FAO study, 2166 edible
prove stress tolerance and the pathogen resistance of their species are known worldwide and 470 species have useful
host trees (Smith and Read, 1997). Picking wild forest mush- medicinal properties (Boa, 2004). Harvesting pressure has in-
rooms is a popular pastime and recreational activity. In cer- creased in many parts of the world (Boa, 2004), and fungal
tain regions of the world, mushroom harvests are also species diversity is claimed to have decreased over the past
commercially important, especially for rural communities in decades (Arnolds, 1991; Wang and Hall, 2004). This has led
developing countries, and in some developed countries as to widespread concern about overharvesting and possible
well (Boa, 2004). In Eastern Europe the export of forest fungi damage to fungal resources. Several countries or regions have
has emerged as an important income source (Peric and Ping- introduced legal restrictions on the harvesting of edible fungi
uli, 2001). In the Pacific Northwest of the United States chant- in natural habitats because they fear that the removal of fruit
erelles have spawned a large commercial harvesting industry bodies from the forest, often before spore dispersal, might im-
over the last two decades (Pilz et al., 2003). The value of total pair their reproduction. Spores are important for the survival,
production of chanterelles on the world market is estimated migration, and distribution of genetic variability and for

* Corresponding author: Tel.: +41 1 7392271; fax: +41 1 7392215.

E-mail address: [email protected] (S. Egli).
0006-3207/$ - see front matter
2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biocon.2005.10.042
 ARTICLE IN PRESS
2 B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x

bringing together compatible mating types for sexual repro- 2.3. Sampling
duction (Dix and Webster, 1995).
Since the 1970s, 19 of the 26 cantons in Switzerland All fruit bodies of the epigeous macromycetes of soil-inhabit-
have introduced weight limits or closed seasons. This has ing species were identified and counted at weekly intervals
caused some controversy, since there is no scientific evi- from May to December (weeks 21–52). Thirty-nine species
dence regarding the effectiveness of such restrictions. We that form large quantities of very small fruit bodies (e.g. Myc-
therefore started a research project in 1975 to investigate ena sp., Strobilurus sp., Marasmius sp.) were excluded to avoid
the impact of harvesting mushrooms on subsequent fruit- counting difficulties. Moreover, 12 taxonomically critical spe-
ing. The study was carried out in the fungus reserve Cha- cies were also excluded to avoid possible irregularities due to
néaz, located in a typical mixed forest on the Swiss unclear identification. When first recorded, the fruit bodies
Central Plateau. In a preliminary analysis, no significant ef- were marked with methylene blue on the cap to avoid double
fects of harvesting were detected on 15 species that met the counting. In the picking and cutting treatments, only edible
minimum requirements for a statistical analysis (Egli et al., fungi were harvested.
1990). Consequently, we decided to extend the study to ob- Chanéaz. The survey was started in 1977, and continued
tain results for more species over a longer period. A second until 2003. Between 1980 and 1983, only the edible fungi were
study was started in a subalpine pure Norway spruce forest recorded. These four incomplete years were excluded from
(Moosboden) to obtain additional data from another preva- the analyses. A total of 436 species were included and
lent Swiss forest type and popular area for mushroom har- 97,700 fruit bodies counted (edible: 103 species/53,863 fruit
vesting. Whereas at Chanéaz we studied the effects of bodies; non-edible: 333 species/43,837 fruit bodies).
harvesting and harvesting techniques (picking/cutting), the Moosboden. The survey was started in 1990 and ended in
Moosboden study focused on the effects of the concomitant 2000. A total of 250 species were included and 50,222 fruit
trampling of the forest floor. This focus was chosen because bodies counted (edible: 51 species/10,173 fruit bodies; non-
an earlier experiment had shown a strong negative effect of edible: 199 species/40,049 fruit bodies).
trampling on the fruit body production of a colony of the A total of 582 species were recorded at the two sites, with
Yellow Foot Chanterelle, Cantharellus lutescens Fr. (Egli and 146 species common to both sites. Collections of all the spe-
Ayer, 1997). cies recorded are deposited in the mycoherbarium of the
Swiss Federal Research Institute WSL.
2. Materials and methods
2.4. Analyses
2.1. Study sites
We tested how the treatments did affect species richness and
The study was carried out in two fungus reserves in south- the total number of fruit bodies produced at the levels of spe-
western Switzerland. cies, families and edible and non-edible species. The numbers
The first, ‘‘Chanéaz’’ (74 ha), established in 1975 in a dom- of fruit bodies produced per year were log transformed to re-
inant forest type of the Swiss Central Plateau at 600 m a.s.l., is duce the influence of plots with large numbers of fruit bodies.
a mixed old-growth forest with deciduous and coniferous tree To avoid problems with zero observations, we added the value
species of different ages (mainly Fagus silvatica L., Quercus ro- 1 to each year sum before calculating the log. The same trans-
bur L., Picea abies (L.) Karst., Abies alba Mill., Pinus silvestris L., formation was used for the number of species after examin-
Pinus strobus L., and Larix decidua Mill.). ing the residuals.
The second, ‘‘Moosboden’’ (3 ha), established in 1990 at We applied two statistical models: 1. To assess the im-
1250 m a.s.l., is a pure, uniform Norway spruce forest (Picea pact of the different treatments by ANOVA, we calculated
abies (L.) Karst), reforested 110 years ago. means over all years since the production of fruit bodies
varied greatly from year to year. 2. To evaluate whether
2.2. Experimental design the progression over time of a parameter differed with
treatment, we applied a repeated measures ANOVA and
Chanéaz. Five 300 m2 blocks were divided into three plots of tested the interaction of treatment and time using Green-
10 m · 10 m with the treatments ‘‘harvesting by picking’’, house–GeisserÕs correction. In both models and datasets
‘‘harvesting by cutting’’, and ‘‘control’’. we included the blocks as a blockfactor. In Chanéaz there
Moosboden. Fourteen 13 · 13 m blocks were divided into was one treatment factor with three steps (control, picking,
four 6.5 · 6.5 m plots with the randomly distributed treat- cutting) and we calculated two contrasts (control vs. mean
ments ‘‘picking with trampling’’, ‘‘picking without trampling’’, of picking and cutting; picking vs. cutting). In Moosboden
‘‘no picking with trampling’’, ‘‘no picking without trampling’’. we tested two treatment factors (harvesting, trampling)
The treatment ‘‘with trampling’’ corresponds to normal walk- and their interaction. Since the interactions were not signif-
ing associated with mushroom harvesting, mimicking that of icant (p-values clearly above 0.05, only in one case close to
a mushroom picker. The ‘‘without trampling’’ plots were pro- this limit), the main effect model was applied. Model
vided with catwalks to avoid soil contact while picking or assumptions were checked using a graphic residual analy-
counting the fruit bodies. The installations were made in sis. We used a quantile–quantile plot to verify normality
April 1990, before starting the experiment. (normal plot), the Tukey–Anscombe plot to test for homo-
All observation plots were surrounded by fences to avoid scedasticity and lack of fit, and a leverage plot for finding
disturbance by mushroom pickers. dangerous leverage points. In general the assumptions held,
 ARTICLE IN PRESS
B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x 3

even if there were some outliers. Exclusions of outliers did The concomitant trampling of the forest floor, however,
not change the results noticeably. The analyses were per- significantly reduced the number of fruit bodies produced (Ta-
formed with the statistical software R (R Development Core ble 1 and Fig. 2(b)). Based on the statistical analysis it is likely
Team, 2004) and the SPSS software (version 12, see http:// that trampling reduces fruit body production to about 70% of
www.spss.com). that on untrampled areas. The mean number of fruiting spe-
cies per year was also significantly lower in trampled plots
3. Results than in non-trampled ones. Surprisingly, however, the total
number of species that fruited over the decade of sampling
The present data show that harvesting does not adversely af- was about the same in the trampled as in the non-trampled
fect the production of fruit bodies (Table 1 and Figs. 1, 2(a)). plots (195 and 189, respectively).
Edible fungi, which were selectively harvested, did not de-
crease relative to unharvested non-edible ones with respect 4. Discussion
to either the abundance of fruit bodies or species richness.
No different trends were detected, even over a period of 29 Fruit body and fruiting species numbers were unaffected in
years, in the harvested and non-harvested sites, irrespective our study areas when they were systematically harvested
of whether the harvesting technique was picking or cutting over a period of 29 years. Irregular field observations in other
(Fig. 1). These findings applied for all the fungal species as areas also suggest that the impact of harvesting may well be
well as for single species and families. negligible (Jahn and Jahn, 1986; Jansen and van Dobben, 1987;

Table 1 – ANOVA table of the effects of harvesting on fruit body production

a
Treatment Mean over time Treatment · time

Subject Group Exp(coef)b Conf. int.c p-Valued p-Value

Chanéaz (1977–2003)
Non-harvesting/harvesting (Reference: non-harvesting) (All treatments)
No. of fruit bodies All species 1.03 0.71–1.48 0.878 0.437
Edible e 1.11 0.78–1.58 0.513 0.261
Non-edible 0.90 0.65–1.26 0.501 0.679
Species richness All species 0.97 0.81–1.17 0.746 0.373
Edible 0.99 0.86–1.15 0.887 0.138
Non-edible 0.99 0.82–1.19 0.876 0.847

Picking/cutting (Reference: picking)

No. of fruit bodies All species 0.81 0.53–1.23 0.279
Edible 0.82 0.54–1.22 0.268
Non-edible 0.87 0.59–1.28 0.427
Species richness All species 0.94 0.76–1.16 0.503
Edible 0.99 0.84–1.17 0.936
Non-edible 0.89 0.72–1.11 0.267

Moosboden (1990–2000)
Non-harvesting/harvesting (Reference: non-harvesting)
No. of fruit bodies All species 1.13 0.85–1.51 0.380 0.343
Edible 1.16 0.81–1.67 0.413 0.660
Non-edible 1.09 0.79–1.52 0.589 0.702
Species richness All species 1.01 0.91–1.13 0.818 0.395
Edible 0.97 0.84–1.17 0.657 0.829
Non-edible 1.03 0.94–1.14 0.514 0.910

Non-trampling/trampling (Reference: non-trampling)

No. of fruit bodies All species 0.72 0.54–0.96 0.028* 0.419
Edible 0.64 0.45–0.92 0.019* 0.064
Non-edible 0.78 0.56–1.08 0.135 0.392
Species richness All species 0.86 0.77–0.96 0.006** 0.377
Edible 0.87 0.75–1.00 0.050* 0.119
Non-edible 0.86 0.78–0.95 0.004** 0.478
a
Treatment · time interactions of repeated measures compare the progression over time of the respective parameter among the different
treatments. At Chanéaz, all three treatments (control, picking, cutting) were compared.
b
Exp (coef) = exponent of the regression coefficient. This value is a measure of the treatment effect. It can be directly interpreted as a factor
value showing the expected changes in the numbers of fruit bodies or species.
c
Conf. int. = confidence interval 2.5%/97.5% of the exp (coef).
d
Asterisk indicates p-value < 0.05, double asterisk indicates p-value < 0.01.
e
Only edible species were harvested in harvested plots.
 ARTICLE IN PRESS
4 B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x

40

e ies
35

Number of spec
30
25
20
15
10
5
3
Log (number fruit bodies)

2.5

1.5

0.5

1993
1987

1990

1992

1995

1998

2001
1985
1986

1988
1989

1991

1994

1996
1997

1999
2000

2002
2003
1977

1984
1978
1979

Fig. 1 – Fungal species richness and transformed number of fungal fruit bodies produced in Chanéaz from 1977 to 2003. Fruit
bodies of all macromycetes were counted weekly in control plots (circles) and in plots where fruit bodies were harvested by
picking (squares) or by cutting (triangles). The values at the top are means of the numbers of fungal species observed per year
and those at the bottom are means of log10 transformed annual sums of fruit bodies (n = 5; bars show s.e.m.). Data from 1980
to 1983 are missing.

Arnolds, 1991). Moreover a 13-year study by the Oregon Myco- Fr. (Egli and Ayer, 1997), where a researcher imitated a
logical Society in the Mount Hood region (Oregon, USA) re- mushroom picker and harvested fruit bodies twice a week
vealed no statistical evidence that picking suppresses the for 12 years. As a consequence the fungus ceased forming
fruiting of the Golden Chanterelle, Cantharellus formosus Cor- fruit bodies, whereas it regularly fruited in the control plots,
ner 1966 (Norvell, 1995; Pilz et al., 2003). which were provided with catwalks to avoid soil contact
Harvesting the fruit bodies entails removing the spores. while harvesting. The treatments were changed twice, once
Theoretically, we would expect removing all fruit bodies after 6 and once after 11 years. In both cases the fruit bodies
and thus the possibility of sexual renewal by spores to lead appeared again the following year in quantities similar to
to a degeneration of these fungi over time. Fungal species those before the treatments.
vary in how important spores are for their reproduction Analysis of the interaction between the treatments tram-
(Fioré-Donno and Martin, 2001). Some species repeatedly re- pling and harvesting indicates that, if anything, the combi-
cruit new colonies from spores, whereas others propagate nation of harvesting/non-trampling seems to be the most
predominantly by vegetative spread. Laccaria amethystina appropriate precondition for producing a maximum number
Cke., for example, produces colonies and fruit bodies from of fruit bodies. Although trampling of the forest floor re-
spores each year (Fioré-Donno and Martin, 2001), but we duces the number of fruit bodies, this seems of minor
did not detect a negative impact of harvesting on this spe- importance compared to other factors influencing fruit body
cies in either study site. It is possible that adequate numbers formation, as suggested by the large annual variability in
of spores entered from the neighbouring areas, or that the fruit body production (Figs. 1 and 2). Good or poor mush-
fruit bodies in the plots released enough spores during the room years seem to be determined mainly by climatic condi-
weekly harvesting intervals. Nevertheless, the present exper- tions (Agerer, 1985; Kasparavicius, 2001; Straatsma et al.,
imental design realistically simulates strong harvesting 2001). Air pollution, such as nitrogen deposition in forests,
pressure. appears also to affect fungal species diversity, as demon-
On the second study site we showed that trampling of the strated in an adjacent study plot at Moosboden (Peter
forest floor associated with mushroom harvesting reduces et al., 2001). In this study the input of nitrogen proved to
the number of fruit bodies and fruiting species observed have an immediate and negative impact not only on the
per year, but not the number of species that fruited over fruit body production but also on the below-ground struc-
the decade of sampling. This means that, in spite of tram- tures of ectomycorrhizal fungi.
pling, the mycelia of all the species we sampled seem to per- Although fungi are difficult to study in the soil, we need to
sist in the soil, but simply fruited less often and in smaller understand their in situ ecology better if scientists are to pro-
numbers. We therefore hypothesize that the pre-fruit body vide conservationists and policy makers with clear criteria for
primordia formed at the soil surface might be mechanically evaluating measures to protect the biological diversity of forest
destroyed by walking on the forest floor, but that the myce- fungi and to maintain sustainable harvests. Our study shows
lium is not permanently damaged. This is supported by the no evidence that harvesting fruit bodies harms the diversity
results of an earlier trampling experiment in a plot with a of fungi residing in forest soils. From the anthropocentric
colony of the Yellow Foot Chanterelle, Cantharellus lutescens perspective, however, both fungal species richness and the
 ARTICLE IN PRESS
B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x 5

precaution to conserve forest fungi. In addition, we should

18 a
not underestimate the importance of the psychological
16
effects of constraints, for instance in increasing public recog-
14
12 nition of forest fungi as a precious natural resource in our for-
10 ests worthy of protection.
8
Number of species
i

6
4

18 b Acknowledgements
16
14 We thank Ch. Fillistorf, H. Bugnon, J.-M. Carrard, R. Dougoud
12 for field assistance, D. Pilz, S. Dingwall, M. Fenaroli, B. Senn-
10
Irlet, I. Brunner and P. Duelli for discussion and critical read-
8
6
ing of the manuscript. We are also grateful to two anonymous
4 reviewers for their helpful comments.
1991
1990

1998
1992
1993
1994
1995

1997

2000
1996

1999

R E F E R E N C E S
4.5
a
4
3.5
3.
Agerer, R., 1985. Zur Oekologie der Mykorrhizapilze. Bibl. Mycol.
97, Cramer, Vaduz.
3 Alexander, S.J., Pilz, D., Weber, N., Brown, E., Rockwell, V.A., 2002.
Log (number fruit bodies)

2.5 Mushrooms, trees, and money: value estimates of commercial

2 mushrooms and timber in the Pacific Northwest. Environ.
Manage. 30, 129–141.
1.5 Arnolds, E., 1991. Decline of ectomycorrhizal fungi in Europe.
5 Agric. Ecosyst. Environ. 35, 209–244.
b
4.5 Boa, E., 2004. Wild edible fungi. A global overview of their use and
4 importance to people. Non-Wood Forest Products 17. FAO,
3.5 Rome.
3
Dix, N.J., Webster, J., 1995. Fungal Ecology. Chapman & Hall,
London.
2.5
Egli, S., Ayer, F., 1997. Est-il possible dÕaméliorer la production de
2
champignons comestibles en forêt? LÕexemple de la réserve
1.5 mycologique de La Chanéaz en Suisse. Rev. For. Fr., Numéro
1991

2000
1993

1998
1992

1995

1997
1990

1994

1996

1999

special, pp. 235–243.

Egli, S., Ayer, F., Chatelain, F., 1990. Der Einfluss des Pilzsammelns
auf die Pilzflora. Mycol. Helv. 3, 417–428.
Fig. 2 – Fungal species richness and transformed number of
Fioré-Donno, A.-M., Martin, F., 2001. Populations of ectomycor-
fungal fruit bodies produced in Moosboden from 1990 to rhizal Laccaria amethystina and Xerocomus spp. show contrast-
2000. Fruit bodies of all macromycetes were counted weekly ing colonization patterns in a mixed forest. New Phytol. 152,
(a) in plots with (circles) or without (squares) harvesting of 533–542.
fruit bodies, and (b) in plots with (circles) or without Jahn, H., Jahn, M.A., 1986. Konstanz und Fluktuation der Pilzveg-
(squares) forest floor trampling. The values in the top figure etation in Norra Warleda (Uppland). Beobachtungen auf einem
are means of the numbers of fungal species and those in the schwedischen Bauernhof. Westfälische Pilzbriefe 10/11, 352–
378.
bottom one are means of log10 transformed annual sums of
Jansen, E., van Dobben, H.F., 1987. Is decline of Cantharellus
fruit bodies (n = 14; bars show s.e.m.; within the 14 blocks cibarius in the Netherlands due to Air Pollution? Ambio 16,
the two plots with the same treatments were averaged 211–213.
before calculating the mean and standard error). Kasparavicius, J., 2001. Effect of trampling on fructification of
Cantharellus cibarius and Boletus edulis in Scots pine forests.
Biologija 3, 21–23.
Norvell, L., 1995. Loving the chanterelle to death? The 10 year
yield of edible fungi are impaired by the trampling of the forest chanterelle project. McIlvanea 12, 6–25.
floor. It is the fruit bodies of the fungi that we see and use. Peric, B., Pinguli, E., 2001. South Eastern Europe mush-
Fewer are formed after trampling of the forest floor, but they rooms menaced by global demand. The Bulletin 10/4,
reappear when the sites are left to recover. 16–17.
Our results raise questions about the usefulness of har- Peter, M., Ayer, F., Egli, S., 2001. Nitrogen addition in a Norway
spruce stand altered macromycete sporocarp production and
vesting restrictions. We do not, however, know how many
below-ground ectomycorrhizal species composition. New
spores are needed to ensure the survival of fungal species. Phytol. 149, 311–325.
We are also not yet in a position to prove our hypothesis that Pilz, D., Norvell, L., Danell, E., Molina, R., 2003. Ecology and
trampling of the forest floor does not harm the mycelia in the Management of Commercially Harvested Chanterelle
soil. We therefore suggest maintaining closed seasons as a Mushrooms. Gen. Tech. Rep. PNW-GTR-576. Portland OR:
 ARTICLE IN PRESS
6 B I O L O G I C A L C O N S E RVAT I O N x x x ( 2 0 0 5 ) x x x –x x x

Department of Agriculture, Forest Service, Pacific Northwest Straatsma, G., Ayer, F., Egli, S., 2001. Species richness, abundance,
Research Station, pp. 1–83. and phenology of fungal fruit bodies over 21 years in a Swiss
R Development Core Team, 2004. R: A language and environment forest plot. Mycol. Res. 105, 515–523.
for statistical computing. In: R Foundation for Statistical Wang, Y., Hall, I.R., 2004. Edible ectomycorrhizal mush-
Computing, Vienna, Austria. ISBN 3-900051-07-0. Available rooms: challenges and achievements. Can. J. Bot. 82,
from: http://www.R-project.org. 1063–1073.
Smith, S.E., Read, D.J., 1997. Mycorrhizal Symbiosis. Academic Watling, R., 1997. The business of fructification. Nature 385, 299–
Press, London. 300.