Mushroom fruiting and climate change
 *Microbial Evolution Research Group and
 ^{†}Centre for Ecological and Evolutionary Synthesis, Department of Biology, University of Oslo, P.O. Box 1066 Blindern, NO0316 Oslo, Norway; and
 ^{‡}Department of Botany, Natural History Museum, University of Oslo, P.O. Box 1172 Blindern, NO0318 Oslo, Norway
See allHide authors and affiliations
Edited by Hans R. Herren, Millennium Institute, Arlington, VA, and approved January 22, 2008 (received for review September 23, 2007)
Article Figures & SI
Figures
Data supplements
Kauserud et al. 10.1073/pnas.0709037105.
Supporting Information
Files in this Data Supplement:
SI Table 1
SI Table 2
SI Text
SI Table 3
SI Table 4
SI Figure 4
SI Figure 4Fig. 4. The same figure as Fig. 3, but here also including partial residuals.
Table 1. Listing of all species included
Species
n
Feeding mode
Initial mean fruiting day
Regression coefficient
Standard error
P
Amanita fulva
448
ECM
230.4
0.30
0.05
<0.001
Amanita muscaria
632
ECM
255.2
0.02
0.07
>0.1
Amanita porphyria
414
ECM
242.6
0.14
0.08
0.07
Amanita regalis
338
ECM
219.7
0.43
0.08
<0.001
Amanita rubescens
544
ECM
236.4
0.24
0.08
<0.001
Amanita vaginata
315
ECM
240.7
0.12
0.08
0.10
Amanita virosa
313
ECM
239.0
0.21
0.06
<0.001
Boletus badius
401
ECM
254.5
0.05
0.05
>0.1
Boletus edulis
590
ECM
238.7
0.20
0.06
<0.001
Boletus subtomentosus
555
ECM
236.8
0.26
0.09
<0.001
Cantharellus cibarius
944
ECM
239.1
0.18
0.09
0.03
Chaliciporus piperatus
507
ECM
238.6
0.25
0.11
0.01
Chroogomphus rutilus
313
ECM
251.6
0.05
0.10
>0.1
Clitopilus prunulus
604
Sapro
248.0
0.07
0.05
>0.1
Cortinarius armillatus
310
ECM
240.5
0.16
0.07
0.02
Cortinarius croceus
314
ECM
250.0
0.43
0.12
<0.001
Cortinarius rubellus
382
ECM
237.6
0.13
0.12
>0.1
Cortinarius semisanguineus
303
ECM
261.5
0.02
0.09
>0.1
Craterellus sinuosus
277
ECM
248.4
0.22
0.18
>0.1
Craterellus tubaeformis
523
ECM
263.5
0.02
0.05
>0.1
Cystoderma amianthinum
450
Sapro
267.3
0.12
0.06
0.06
Cystoderma carcharias
271
Sapro
278.2
0.08
0.09
>0.1
Galerina marginata
342
Sapro
262.5
0.17
0.11
>0.1
Gomphidius glutinosus
456
ECM
225.5
0.29
0.11
<0.001
Gomphidius roseus
262
ECM
238.3
0.20
0.08
0.01
Gymnopilus sapineus
359
Sapro
259.4
0.03
0.10
>0.1
Gymnopus dryophilus
452
Sapro
207.6
0.63
0.12
<0.001
Hydnum repandum
657
ECM
249.9
0.21
0.08
<0.001
Hydnum rufescens
556
ECM
254.7
0.18
0.09
0.02
Hygrocybe conica
430
Sapro
245.7
0.24
0.09
0.006
Hygrocybe pratensis
397
Sapro
255.8
0.18
0.06
<0.001
Hygrocybe punicea
285
Sapro
259.5
0.17
0.05
<0.001
Hygrophoropsis aurantiaca
320
Sapro
265.8
0.05
0.07
>0.1
Hygrophorus olivaceoalbus
348
ECM
258.3
0.01
0.08
>0.1
Hypholoma capnoides
469
Sapro
269.4
0.04
0.11
>0.1
Hypholoma fasciculare
293
Sapro
228.1
0.71
0.20
<0.001
Hypholoma lateritium
263
Sapro
261.4
0.31
0.10
<0.001
Inocybe geophylla
410
ECM
262.1
0.05
0.07
>0.1
Laccaria amethystina
435
ECM
264.5
0.03
0.05
>0.1
Laccaria laccata
549
ECM
237.6
0.42
0.11
<0.001
Lactarius aurantiacus
294
ECM
253.3
0.26
0.12
0.02
Lactarius deliciosus
227
ECM
243.8
0.27
0.09
<0.001
Lactarius deterrimus
564
ECM
225.5
0.17
0.14
>0.1
Lactarius glyciosmus
487
ECM
255.9
0.10
0.09
>0.1
Lactarius helvus
283
ECM
245.9
0.11
0.10
>0.1
Lactarius necator
444
ECM
256.2
0.06
0.07
>0.1
Lactarius rufus
603
ECM
236.7
0.31
0.07
<0.001
Lactarius theiogalus
292
ECM
237.2
0.22
0.14
0.08
Lactarius torminosus
631
ECM
249.5
0.24
0.10
0.02
Lactarius trivialis
539
ECM
243.6
0.17
0.07
0.008
Lactarius vietus
509
ECM
249.6
0.07
0.07
>0.1
Leccinum scabrum
610
ECM
238.5
0.33
0.10
<0.001
Leccinum versipelle
491
ECM
237.3
0.52
0.08
<0.001
Marasmius oreades
226
Sapro
213.8
0.55
0.22
0.006
Megacollybia platyphylla
375
Sapro
217.4
0.50
0.16
<0.001
Micromphale perforans
423
Sapro
219.4
0.64
0.13
<0.001
Mycena epipterygia
339
Sapro
278.7
0.13
0.08
>0.1
Mycena galericulata
429
Sapro
252.1
0.18
0.12
>0.1
Mycena pura
396
Sapro
233.9
0.34
0.15
0.01
Paxillus involutus
612
ECM
254.2
0.26
0.07
<0.001
Pholiota mutabilis
459
Sapro
215.1
0.69
0.12
<0.001
Pleurocybella porrigens
295
Sapro
258.5
0.18
0.08
0.01
Pluteus atricapillus
326
Sapro
238.6
0.19
0.10
0.02
Rhodocollybia butyracea
447
Sapro
279.5
0.18
0.07
0.008
Rozites caperata
486
ECM
242.3
0.23
0.07
<0.001
Russula aeruginea
626
ECM
239.2
0.46
0.09
<0.001
Russula claroflava
397
ECM
227.5
0.38
0.09
<0.001
Russula decolorans
423
ECM
227.3
0.29
0.10
<0.001
Russula delica
280
ECM
244.4
0.18
0.11
>0.1
Russula emetica
367
ECM
228.0
0.27
0.09
0.002
Russula foetens
286
ECM
234.3
0.41
0.11
<0.001
Russula queletii
314
ECM
243.1
0.19
0.09
0.02
Russula vesca
316
ECM
230.4
0.38
0.10
<0.001
Stropharia semiglobata
262
Sapro
227.3
0.60
0.17
<0.001
Suillus bovinus
326
ECM
254.7
0.06
0.09
>0.1
Suillus luteus
395
ECM
249.1
0.17
0.09
0.05
Suillus variegatus
379
ECM
244.0
0.14
0.10
>0.1
Tapinella atrotomentosa
328
Sapro
239.8
0.14
0.12
>0.1
Tricholoma album
312
ECM
260.1
0.13
0.07
0.04
Tricholoma fulvum
413
ECM
259.2
0.08
0.05
>0.1
Tricholoma saponaceum
442
ECM
260.8
0.13
0.09
>0.1
Tricholoma vaccinum
331
ECM
272.7
0.03
0.05
>0.1
Tricholomopsis rutilans
453
Sapro
239.9
0.23
0.08
0.004
n indicates the number of records for each species (outliers, n = 60, excluded), and Feeding mode shows whether the fungi are growing in symbiosis with plants (ECM) or growing on dead organic matter (Sapro). Initial mean fruiting day is based on records from the period 19401959. The regression coefficient (change in fruiting day per year) was calculated for each species separately by using GAMs with linear time effects and the geographic effects accounted for by smooth functions of longitude and latitude (thinplate regression spline with maximally 11 degrees of freedom). Standard errors and P values for regression coefficients were calculated by using bootstrap tests for each species (see Materials and Methods).
Table 2. Performance of alternative models
Predictor variables
R^{2}_{ADJ}
CV
None

24.97
Sp
15.8
23.04
Sp + f(Long,Lat)
19.3
22.67
Sp + f(Long,Lat) + Year
21.4
22.43
Sp + f(Long,Lat) + g(Year)
22.2
22.41*
Sp + f(Long,Lat) + g_{i}(Year  Feeding mode 12)
22.3
22.42
Sp + f(Long,Lat) + g_{i}(Year  Region 15)
23
22.43
Sp + f(Long,Lat) + g_{i}(Year  Initial Fruiting Day 13)
22.9
22.32
Sp + f(Long,Lat) + h(Year, Initial Fruiting Day)
23.2
22.28*
Sp + f(Long,Lat) + T_{AUG t}
21.9
22.31
Sp + f(Long,Lat) + T_{NOV t1} + T_{AUG t}
23.4
22.08
Sp + f(Long,Lat) + T_{NOV t1} + T_{MAY t} + T_{AUG t}
24.2
21.97
Sp + f(Long,Lat) + T_{NOV t1} + T_{MAY t} + T_{AUG t} + P_{JUL t1} + P_{NOV t}
24.9
21.87
Sp + f(Long,Lat) + T_{NOV t1} + T_{FEB t} + T_{MAY t} + T_{AUG t} + T_{OCT t} + P_{JUL t1} + P_{NOV t}
25.3
21.86
Sp + f(Long,Lat) + T_{JUN t1} + T_{NOV t1} + T_{FEB t} + T_{MAY t} + T_{AUG t} + T_{OCT t} + P_{JUL t1} + P_{JUN t} + P_{OCT t} + P_{NOV t}
25.8
21.85
Sp + f(Long,Lat) + T_{JUN t1} + T_{NOV t1} + T_{FEB t} + T_{MAY t} + T_{JUN t} + T_{AUG t} + T_{OCT t} + P_{JUL t1} + P_{SEP t1} + P_{JUN t} + P_{OCT t} + P_{NOV t}
26.1
21.83
Sp + f(Long,Lat) + T_{NOV t1} + T_{FEB t} + g_{1}(T_{MAY t}) + T_{JUN t} + T_{AUG t} + T_{OCT t} + P_{JUL t1} + P_{JUN t} + P_{OCT t} + g_{2}(P_{NOV t})
26.5
21.75*
Sp + f(Long,Lat) + T_{NOV t1} + T_{FEB t} + g_{1}(T_{MAY t}) + T_{JUN t} + T_{AUG t} + T_{OCT t} + P_{JUL t1} + P_{JUN t} + P_{OCT t} + g_{2}(P_{NOV t}) ........
......... + g_{3}(Year)
27.1
21.78
......... + h(Year, Initial Fruiting Day)
28.1
21.62
......... + T_{FEB t}*Initial Fruiting Day + T_{OCT t}*Initial Fruiting Day
26.8
21.73
......... + T_{FEB t}*Initial Fruiting Day + T_{OCT t}*Initial Fruiting Day + g_{3}(Year)
27.3
21.76
......... + T_{FEB t}*Initial Fruiting Day + T_{OCT t}*Initial Fruiting Day + h(Year, Initial Fruiting Day)
28.1
21.64
R^{2}_{ADJ}: Proportion of variance explained (%) adjusted for the numbers of degrees of freedom. CV: genuine crossvalidation error = square root of mean squared prediction errors, determined by leaving data for one year out at a time, thus accounting for withinyear spatial autocorrelation. Models with lower CV have higher outofsample predictive power.
Sp: species (categorical variable). Long: longitude (ºE). Lat: Latitude (ºN). T: temperature (ºC anomaly). P: precipitation (% of normal). T and P refer to regional, monthly means for either the same (t) or the preceding (t  1) year as fungal fruiting.
f: thinplate regression spline with maximally 11 degrees of freedom (12 knots).
g [g_{i}]: natural cubic splines with maximally 3 degrees of freedom (4 knots).
h: tensorproduct smooth function constructed from linear combinations of terms that are cubic regression spline basis functions of Year and Initial Fruiting Day, respectively, each with maximally 3 df (4 knots).
* Models shown in Figures 13.
Table 3. Linear changes in temperature and precipitation in the periods 19402006 and 19802006 in Norway
19402006
19802006
Month
Temperature
Precipitation
Temperature
Precipitation
Jan.
3.9 ± 1.1 *
54 ± 15 *
4.2 ± 1.8 *
25 ± 27
Feb.
3.0 ± 1.2 *
45 ± 19 *
2.4 ± 1.7
59 ± 36
Mar.
1.5 ± 0.9
25 ± 16
0.3 ± 1.3
3 ± 26
Apr.
1.0 ± 0.5
4 ± 14
1.7 ± 0.8 *
22 ± 19
May
0.7 ± 0.5
26 ± 14
0.4 ± 0.7
31 ± 21
June
0.4 ± 0.5
1 ± 13
0.3 ± 0.9
28 ± 21
July
0.4 ± 0.5
1 ± 11
1.4 ± 0.7
5 ± 20
Aug.
1.0 ± 0.5
3 ± 12
2.5 ± 0.9 *
3 ± 15
Sep.
0.6 ± 0.5
3 ± 10
2.2 ± 0.8 *
15 ± 17
Oct.
0.2 ± 0.6
14 ± 10
0.6 ± 1.1
22 ± 15
Nov.
0.3 ± 0.7
21 ± 12
2.2 ± 1.3
11 ± 22
Dec.
0.1 ± 1.0
20 ± 12
2.8 ± 1.7
11 ± 20
Temperature and precipitation indices are average anomalies (in °C and %, respectively) across five regions of Norway. Estimates of changes across each year range ± standard errors are shown. *, P < 0.05.
Table 4. Linear changes in temperature (°C) and precipitation (% anomalies) in the periods 19402006 and 19802006 in southeastern Norway, the region in which 47% of the mushroom data were collected
19402006
19802006
Month
Temperature
Precipitation
Temperature
Precipitation
January
5.0 ± 1.4*
51 ± 18*
4.9 ± 2.3*
37 ± 28
February
3.5 ± 1.5*
25 ± 23
3.8 ± 2.2
32 ± 41
March
1.6 ± 1.1
56 ± 17*
0.5 ± 1.6
47 ± 27
April
0.8 ± 0.6
38 ± 21
1.6 ± 0.8
59 ± 36
May
0.6 ± 0.5
27 ± 20
0.6 ± 0.8
17 ± 36
June
0.2 ± 0.6
4 ± 18
0.2 ± 1.0
3 ± 32
July
0.5 ± 0.6
13 ± 16
1.3 ± 0.9
10 ± 27
August
1.0 ± 0.6
15 ± 21
2.4 ± 1.0*
13 ± 30
September
0.6 ± 0.6
11 ± 14
2.5 ± 0.9*
29 ± 20
October
0.2 ± 0.7
33 ± 19
0.5 ± 1.2
7 ± 32
November
0.5 ± 0.9
12 ± 18
2.6 ± 1.5
22 ± 36
December
0.0 ± 1.2
8 ± 18
2.9 ± 2.0
2 ± 25
Estimates of changes across each year range ± standard errors are shown. *, P < 0.05.
SI Text
Output from Residual Analysis.
Model formula: Y_{ijkl} = m + b1_{i} + b2_{ij} + b3_{ijk} + e _{ijkl}
where Y_{ijkl} is residual fruiting day (residuals from model shown in Fig. 1B, accounting for effects of species, location, and temporal trends) for observation l in municipality k of species j in year i.
m represents the overall mean (not significantly different from zero).
b1 represents normal distributed random variability between years. The standard deviation (SD) of b1 was estimated to be 8.9 (approximate 95% confidence interval, c.i. = 7.4, 10.9).
b2 represents normal distributed random variability between species nested in years. The SD of b2 was estimated to be 3.4 (c.i. = 2.9, 4.0).
b3 represents normal distributed random variability between municipalities nested in species nested in year. The inclusion of this term accounted for spatial autocorrelation and provided better fit (lower AIC) than a model with Gaussian spatial autocorrelation structure depending on geographical coordinates. The SD of b3 was estimated to be 14.6 (c.i. = 14.3, 14.9).
e represents heteroscedastic, independent error. The error SD was modeled to follow an exponential function of time (t_{i} = 066 years): e_{(}t_{i)} = a e^{bt}. This error structure provided significantly better fit than one assuming homogeneous error variance (c^{2}_{1} = 291.3, P < 0.0001). The coefficient a, representing withinspecies withinmunicipality variability in year 1940, was estimated to be 19.5 (c.i. = 18.8, 20.2). The coefficient b was estimated to be 0.0056 (c.i. = 0.0063, 0.0050), indicating that, after 60 years, the SD of fruiting time of a given species at a given place was reduced by 29% (c.i. = 26%, 31%) compared to the initial SD.