This article was downloaded by: [178.18.24.55] On: 02 July 2015, At: 08:49 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG

Grana Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/sgra20

Trends of some airborne tree pollen in the Nordic countries and Austria, 1980—1993 a

b

b

Siegfried Jäger , Siwert Nilsson , Britt Berggren , Anna-Mari Pessi , Marjo a

Helander & Hallvard Ramfjord a

d

ENT Clinic, University of Vienna , Vienna, Austria

b

Palynological Laboratory, Swedish Museum of Natural History , S-104 05, Stockholm, Sweden c

Department of Biology , University of Turku , Aerobiology Unit, Turku, FIN-20014, Finland d

Department of Botany , University of Trondheim , Dragvoll, N-7055, Norway Published online: 01 Sep 2009.

To cite this article: Siegfried Jäger , Siwert Nilsson , Britt Berggren , Anna-Mari Pessi , Marjo Helander & Hallvard Ramfjord (1996) Trends of some airborne tree pollen in the Nordic countries and Austria, 1980—1993, Grana, 35:3, 171-178, DOI: 10.1080/00173139609429078 To link to this article: http://dx.doi.org/10.1080/00173139609429078

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Grana: 35: 171-178, 1996

Trends of some airborne tree pollen in the Nordic countries and Austria, 1980-1 993 A comparison between Stockholm, Trondheim, Turku and Vienna SIEGFRIED JAGER, SIWERT NILSSON, BRITT BERGGREN, ANNA-MAR1 P E W , MARJO HELANDER and HALLVARD RAMFJORD

Downloaded by [178.18.24.55] at 08:49 02 July 2015

JBger, S., Nilsson. S., Berggren, B., Pessi, A.-M., Helander, M. & Ramfjord, H. 1996. Trends of some airborne tree pollen in the Nordic countries and Austria, 1980-1993. A comparison between Stockholm, Trondheim, Turku and Vienna. - Grana 35: 171-178 ISSN 0017-3134. Long term trends in the occurrence of airborne pollen might help in elucidating the probability or unlikelihood of developments which have been predicted in theory such as global warming or forest decline. We have investigated and compared the trends over a 14 year period (1980-1993) of five selected common tree pollen types (Abiur, Corjliu, Betula, Piniu, and Ulnius) from three sites in North Europe (Stockholm, Trondheim, Turku) and one in Central Europe (Vienna) in terms of the start, peak, end, and duration of the season, peak values and annual totals, using data from the European pollen data bank. The existence of trends has been investigated by calculating correlation against years and the statistical significance of the regression lines has been examined at the 5% probability level (pt0.05). A consistent trend in the earlier commencement of the flowering season is observed in all cases, as well as an earlier incidence of peak days. A marked trend to an earlier end of the season was also found in the majority of cases. In most cases, the length of the season showed a trend towards prolongation although this was not significant. Both the peak values and the annual totals exhibit clear trends towards higher values. Since both the start of the season and peak days are primarily temperature-depcndant. the clear trends towards an earlier start and peak day in practically all the pollen types and sites examined supports the theory of global warming or, at least, a temporary rise in air temperature over the seasons covered by this investigation. The same conclusion can be drawn from the fact, that most of the trends of pollen frequency show an increase in the number of airborne pollen grains. It was not possible to comment on the possibility of forest decline, as there is insufficient knowledge on the influence of environmental factors on pollen production. Our material shows no trends which are distinct and consistent enough in terms of either decreasing or increasing annual totals or peak values to throw light on the question of forest decline. Sieg/ried Jciger, ENT Clinic, University of Vienna, Vienna, Atutria; Siwert Nilsson and Britt Berggren. Palynological Laboratory, Swedish hfiiseuni of Natural History, 9104 05 Stockliolni, Sweden; AnnaAfari Pessi and Afarjo Helander, Aerobiology Unit, Departnient of Biology, University of Turku, FIN20014 Turku, Finland; Hallvard Raiilfjord, Departnient of Botany, University of Trondlieini, N-7055 Dragroll, Norway. (Mamiscript accepted 4 Jiine 1996)

Since the European pollen data bank in Vienna contains a large amount of recent pollen data from many European sites, it is possible to make comparisons between different regions in terms of medium range trend analyses. In this study an attempt is made to detect trends for the start of the season, annual totals, peak days, time of maximum occurrence and annual totals for five tree pollen taxa at four localities, Trondheim, Stockholm, Turku and Vienna. The questions of potential forest decline and global warming influenced the choice of taxa: Alnirs, is a common and widespread genus in central and northern Europe, and shows a wide variation with regard to flowering time and pollen production. The same is true of Curyltrs and Bettila. For these three genera the start of the pollen season is strongly influenced by temperature and their annual productivity seems to undergo a fairly distinct biennial fluctuation. Ultiius 0 1996 Scandinavian University Press. ISSN 0017-3131

is known for its vulnerability to parasitic diseases causing death. Pitrirs is significant in forest decline. With the exception of Pitius, all these taxa are also of allergological interest. Previous studies (Jager 1989, 1990, Jager et al. 1991) have shown that, to some extent, significant trends are to be expected. MATERIAL AND METHODS The location of the sampling sites and their elevation are shown in Fig. 1. At every site sampling was by means of Burkard traps. The distribution and frequency of the taxa is summarized in TableI. Daily pollen counts for Alnus, Coryliis, Betufa. Ulniiu and Pinus for the period 1980-1993 from Stockholm, Trondheim, Turku and Vienna, were extracted from the European pollen data bank by means of the European Aeroallergen Network Server (EANS). Annual totals and peak amounts have been calculated. The peak Gram 35 ( I 996)

172

S. Jager et al.

Table I. Distribution andfrequency of taxa. ***common**usual*rare(*)occasional

Trondheim

Downloaded by [178.18.24.55] at 08:49 02 July 2015

TurLu Vienna

-

a I 1. a b b e m . Ic\cl;

59'21" 63' 26' N 60' 32' N 48' 14' N

-

a p. P

18'03'E 10" 26'E 22' 28' E 16" 22' E

20 m a s I. 58 rn a s I. 40 m a.s.1. I70 m a s 1.

2Omag 4mag ISmag 20mag

Alnus incana

*

U. l a d s U. minor

Trondheim

Turku

Vienna

**

** *** (*I **

**

-

-

-

-

*** **

*** *** (*I *(*I (*I t** ** (*I (*I

*

*** ***

(*I * *** t

-

-

*** *** (*I (*I

(*I ** *

t**

-

*** (*)

*** * (*I *** (*I -

(*I (*I +** ** t

* *

.according to Ehrendorfer (1973) & Hrimet-Ahti et al. (1989)

~ ground C

Fig. 1. Sampling sites.

day, i.e. the day with the highest value was also considered. In case of multiple peaks of equal height for the same taxon in the same year at a single station, the first occurrence was chosen as the peak day. The beginning of the pollen season (main pollen season) has been defined in different ways by different authors (Lejoly-Gabriel 1978, MuIIenders et al. 1972, Nilsson & Person 1981, Pathirane 1975). In this study, the onset of the pollen season was defined as the first day with a pollen amount exceeding 1% of the annual total, providing this was not followed by more than six consecutive days with zero values. The end of the season was defined as the number of the day upon which 95% of the annual total was reached. The duration of season is the difference in days between the start and the end of the season. In 1982, 1989 and 1990 sampling was started late in Trondheim. For this reason Aliiw, Corjlus and to some extent Ulnius are excluded from the Trondheim calculations for those years. Corylus occurs in very low quantities in Turku. It was not possible, therefore to determine the start of the season date for Corylus at that site in 1981, 1982, 1983, 1985 and 1988. In addition no Corylus pollen was trapped in Turku in 1988 which led to the elimination of this year from the calculation of the trends of the peak day for Corylus at that site. All other data sequences were complete and contained a full data sequence for 14 pars. A simple correlation of the pollen values against the year was made, following the correlation coefficient evaluation described by Henryson (1957). The statistical significance was calculated to an error level of less than 5% (Table I). The slope coefficient (b) of the regression equations was tested for its Significant difference from 0 in order to verify the trends. These calculations were performed using SAS statistical software (Table 11). The graphs have been constructed using a commercial graphic software package named FREE-LANCE.

RESULTS AND DISCUSSION

The trends selected include: (1) the beginning of the pollen season, (2) the time of the first peak day, (3) the end of the pollen season, ( 4 ) peak concentrations, (5) annual totals. Grana 35 (1996)

Stockholm

A. ghrtinosa A. viridis Corylio avellana C.coluriia Betula huinilis B. pendula B. pubescens B. nana Pinus cembra P. mugo P. nigra P. sylvestris U h u s glabra

Sampling sities:

Stockholm

Taxa.

Trends in general Corylus (Fig.2). - The annual totals rise slightly in Stockholm and Turku and significantly in Vienna, although the annual totals for Coryliu are numerically very low in Turku. No distinct trend is seen in the Trondheim material. The peak concentrations follow these trends, but are not statistically significant. In Trondheim a slight trend towards lower peak values is even observed. There is a distinct trend towards a n earlier start of the season, which is statistically significant in Stockholm and Trondheim (note that values for five of the fourteen years are missing due to the low amount of Corylus pollen in Turku and Trondheim). Peak days show a trend towards occumng earlier in the year at all the sites, this being significant in Stockholm and Turku. I n Vienna, this trend is less pronounced. In correspondence with the trends towards a n earlier start and peak day, the end of the season also shows a trend towards occurring earlier in the year. Again, in Stockholm and Turku this trend is significant. The duration of the season consistently shows a slight trend towards prolongation, although this is not statistically significant.

Abzirs (Fig.3). - The annual totals show n o mutual trend at all. A slight increase is observed in Vienna, while in Turku a faint trend decrease is detected. The peak concentration shows a slight downward trend in Turku but this is not significant. The onset of the season became significantly earlier during the period 1980 to 1993 in Stockholm, Trondheim and Turku, but this trend is not significant in Vienna. The peak day shows a similar trend towards the beginning of the year and this is statistically significant in Stockholm and Trondheim. The trend for the end of the season is towards occurring earlier in Trondheim, Stockholm and Vienna (the slope of the regression coefficient is very close t o being significant). I n Turku the trend is similar t o that of the other sites but very weak. A trend towards an earlier end of the season is significant in Stockholm. In the

Pollen trena!s in Scandinavia and Vienna

173

Table 11. Trends in airborne tree pollen, 1980-1993. Significant trends (pt0.05) p-value for slope coefficient different from 0 0 )the trend for the end of the Abiils season in Vienna was very close to significance (p=0.0502) Sampling stations raxa

4biw

Forjlus

Downloaded by [178.18.24.55] at 08:49 02 July 2015

3etula

total peak value start peak day end

Stockholm

Turku

*p=O.O266 *p =0.0449

*p =0.0126 *p = 0.0074 *p=0.0386

*p=0.0171

Vienna

0

*p=0.0453

total peak value start peak day end

*p= 0.0224 *p =0.0298 *p=0.0361

total peak value start peak day end total peak value start peak day end

'inw

Trondheim

*p=0.0333 *p=0.0035

*p =0.0058 *p =0.0267 *p=0.0203 *p=0.0161

total peak value start peak day end

Jordic countries a weak trend towards a lengthening of the :ason is detected, whereas in Vienna a trend towards a horter season is found, though this is not significant. Betirla (Fig.4). -No significant trends exist with the single xception of a trend to an earlier end of the season in Vienna. .nnual totals show increasing trends in Scandinavia but no *endat all in Vienna. Relatively distinct, but not significant .ends are observed in higher peak values at all sites. The art of the season shows uniform trends to earlier in the :ar, this being more pronounced in the Nordic countries llthough still not significant) than in Vienna. A weak trend wards earlier peak days is seen in Stockholm and Vienna, hile in Trondheim and Turku no trend at all is visible. The id of the season inclines slightly towards a later date in rondheim and, more distinctly, towards an earlier date in tockholm, Turku, and Vienna. The Betula season obviously LOWS a trend towards longer duration in the Nordic counies and towards shorter duration in Vienna. Ulmtrs (Fig.5). - The annual totals show no trend, or ily a slight trend,. towards lower values in Trondheim and .ockholm, and a trend towards higher values in Turku and ienna. Of these, the trend in Vienna is significant. The peak

*p=0.0687 *p =0.0442

days show exactly the same trends, with the exception, that the trend in Vienna is no longer significant. The onset of pollen dispersal reveals a general trend towards an earlier commencement of flowering, which is significant in Trondheim. The trends of the peak day show exactly the same pattern. There is no trend for the end of season in Trondheim, but there is a significant trend based on the correlation with a n earlier end to the season in Stockholm. However, this has no significance for the regression slope (p =0.0687). This is the only occasion where the two methods showed different results. Weak trends have been detected in the same direction in Turku and Vienna. No distinct trends with regard to the duration of season are visible, with the exception of Trondheim, where a trend towards a longer season is obvious, but not significant. Pinirs (Fig. 6). - The annual totals remain stable in Trondheim, there is a slight decrease in Stockholm, and a distinct increase in Turku and Vienna. The trend in Vienna is significant. With the exception that it is not significant, the peak concentration shows exactly the same pattern. Generally, a slight trend towards an earlier start to the pollen season is found. This is more pronounced in Vienna than elsewhere. Slight, but not significant trends for earlier peak Gram 35 (1996)

corylus

A

trends in start of s e m n

A

Alnus

trends in start of reason

(20,

120

....

rm ...

t

...

... .----

1 ..........-_ --...........-......-...... a

o-..o--

20

..-b

....

....

..

-I--**-&__-__

.....--o ......+. ...: .... ............... .......... -._. . . . . . . . . . . . . 0

...*

.

- 0

.

trends in peak dayr

im .

.

,B I

.-...................

Downloaded by [178.18.24.55] at 08:49 02 July 2015

Im

1 "e .....

................................................. -

.

y

*-.-~---. .. ...

.....

s

,c

trends in end of season

im

..................

....................

.

....

......

.............

sm

- -

. .~

_______ __a. ..............

.............. ....................- * - - - - - .....

103

..

p

...........

...+*=.EL-. 0

D

-

.

....;.;.-..-.-.-

. _. o

-T

~-

_....

...

m

E?i?zPJ*.. % :

P

u

.

.

m

~

u

yew

yew beom

. .

am

0

--.-

...........

tudl rm *Mm*

TEz3 EZ!? 2% .I%?..

Fig. 2. Corjlus. (A) Beginning of pollen season. (B) Peak days. (C) End of pollen season. (D) Peak concentrations. (E) Annual

Fig. 3. A~JUS. (A) Beginning of pollen season. (B) Peak days. (C) End of pollen season. (D) Peak concentrations. (E) Annual

totals.

totals.

Gram 35 ( 1996)

A

Betula kcnds in start of s e w n 1 0

I

trendsin peak days

Downloaded by [178.18.24.55] at 08:49 02 July 2015

10,

..a

........

...

..

4.m

.............

.

.*-.

.

.. ,.-.._I

,B

......

--•

........ 0

F.-

/--

E

l : : : r _____. l j - - l Kg

.......-*--.

.....>-~

--%-

!.--L--;--

...

--A

la

u

m

2-•

~~. ....

....... ....l................... "

O

I

p

y

Fig. 4. Berulu. (A). Beginning of pollen season. (B) Peak days. C) End of pollen season. (D) Peak concentrations. (E) Annual

Fig. 5. Ulntus. (A) Beginning of pollen season. (B) Peak days. (C) End of pollen season. (D) Peak concentrations. (E) Annual

otals.

totals. Gram 35 (1996)

S. Jager et al.

176

h,,

days are found in all four stations. A similar weak trend towards an earlier end of season is observed in Stockholm and Vienna, while a trend in the opposite direction is observed in Turku. The duration of the season shows no trend in Trondheim and Stockholm but a slight trend towards longer seasons in Turku and Vienna.

A

Pinus

trendsin start of selton

im I

1

...

.....

3 ...................

........ 0.

0

..

190,

trend3 in peak day

....

................

Downloaded by [178.18.24.55] at 08:49 02 July 2015

110

P

Y

MC f l 4

.

....P . . . . . . I

ye= 1MWn-1lm 8.-&

.

...........

Tuhl

-c -c ---*

U

D

.

.

.I

DiJJerettces betweeti the satitpliiig sites. - For Trondheim, the trends towards an earlier start of the seasons for Coryhrs, Alniis and U h t a as well as for earlier peak days of A ~ J C I S and Ultittis were significant. For Stockholm, the trends towards an earlier start, peak day, and end of the season for Altitrs and Coryhis proved to be significant, as well as a trend towards an earlier end of the season for Ultiiia. For Turku, significant trends have been detected for the start of the Altttis season, and for an earlier peak and end of the season for Coryltis. For Vienna, the trends towards higher totals of Coryhrs, Ultiiiis, and Pittiis, and an earlier end of the Betirla season were significant.

.B

- ........ ~

I

..... .... VLm 0

trends in end of season

C

.......

I

Items ewliiated. - Of the 120 trends examined, nineteen proved to be significant. With respect to annual totals,. a significant upward trend is observed for Corylia, Pinus, and Uhiits. The peak concentration shows no significant trend at all. The start of the pollen season shows a gradient towards earlier occurrence for the period 1980 to 1993 at all sites and for all the pollen types investigated, where the trends for A h i s in all the Nordic countdes, Coryltis in Stockholm and Trondheim, and Uliiiris in Trondheim were significant. Similarly, the peak day was significantly earlier for Aliius in Trondheim and Stockholm, for Coryltts in Stockholm and Turku, and for Ulimis in Trondheim. The end of the season showed significant trends towards an earlier occurrence for Alms in Stockholm, for Corylirs in Stockholm and Turku, for Bettila in Vienna, and for Ultinrs in Stockholm. The majority of the trends evaluated showed an increase in the length of the season, but none were significant.

Specific trends Begittttitlg of the pollen sensoti (Fig. 7). - The early springs and mild winters, especially for the years 1988 to 1990 and 1992 to 1993 probably affected the onset of the pollen season and the incidence of the peak days, but not the annual totals

I

1

1980-1993 0.5

. E

$ o

!

4c

8

-

T W i SaxkWm -c

-.m m-@