Summary Germination ability and airborne counts of Scots pine (Pinus sylvestris L.) pollen were studied during the spring of 1993 at Turku in southern Finland (60°32′ N, 22°28′ E) and at Utsjoki in northern Finland (69°45′ N, 27°01′ E). Pollen was trapped from the beginning of May to the end of June in a high-volume air sampler. Germination tests were performed to determine the in vitro pollen viability of the trapped pollen. Airborne pine pollen counts were obtained from a continuously operating Burkard trap located near each high-volume sampler. When male flowering began, phenological observations were carried out on pollen grains collected in rotored samplers located in pine and spruce stands and open fields near Turku and Utsjoki. In southern Finland, the peak period of pine pollen production was short, lasting for only 3 days, but it accounted for about 80% of the total germinating pine pollen yield for the year. The peak count was on May 20, with over 2000 germi-nating pollen grains per cubic meter of air. Pollen germination rates of up to 70% were obtained during the week preceding the local pollen peak, and rates reached almost 90% on the peak day. Pollen viability remained at 45 to 65% for 1 week after the peak. There was no significant difference between the pollen counts for day and night, indicating that during the main pollen season, the pollen source was close to Turku. Before the local pollen peak, the counts of living pine pollen were low, indicat-ing that pine pollen transported over long distances was of little ecological importance in 1993 in the Turku area. In northern Finland, the first pollen grains were caught on July 4, and the peak day was July 13. However, no viable pollen was observed during this period, indicating that there was little gene drift from southern to northern Finland in 1993.
Keywords: flowering phenology, in vitro pollen germination, long-range gene transfer, Pinus sylvestris, seed set.
Introduction
Long-range gene transfer in Scots pine (Pinus sylvestris L.) is assumed to occur because pollen grains are often trapped before the local pollination period occurs (see Finnish Pollen Bulletin, Vols. 1--18). It has been suggested that long-range gene transfer results in improved adaptability and tolerance to a changing environment. However, little is known about the
viability of pollen that has been transported over long dis-tances.
Long-range pollen transport is controlled by large-scale, three-dimensional atmospheric motions, which are often diffi-cult to predict. Although air mass trajectories can be calculated retroactively (see e.g., Hjelmroos 1991, Comtois 1995) to determine the most probable source of pollen grains, pollen from distant sources cannot be distinguished from local pollen (Ho 1992). The large-scale phenological observations required to demonstrate that pollination is the result of pollen from distant sources are almost impossible to carry out over large areas. Also, the resuspension of old sedimented pollen is diffi-cult to exclude.
Although several staining techniques have been developed to evaluate viability of pollen grains (e.g., 2,3,5-triphenylen-tetrazolium chloride (Cook and Stanley 1960) for testing oxi-dative enzyme activity, and inorganic acids (e.g., 14.4% H2SO4) for testing cell wall permeability), the best method for
testing viability is to germinate pollen grains in a liquid me-dium or on an agar meme-dium containing sucrose (Stanley and Linskens 1974). A pollen grain is considered to have germi-nated when the length of the pollen tube exceeds half of the diameter of the pollen grain (Ho 1992). Because the germina-tion rate on agar medium is highly correlated with pollen germination in vivo (Jett et al. 1993), the test can be used to predict the total seed yield per cone and the number of first-year aborted ovules. However, the correlation between in vitro germination ability and seed setting ability is not always high (Jensen 1964).
The aims of the present study were (1) to evaluate the possibility of combining information on germination rates, Burkard pollen counts and local flowering to determine the origin of airborne pollen, and (2) to study the probability of long-range gene transfer in Scots pines.
Materials and methods
The viability of pine pollen was determined at Turku in south-ern Finland (60°32′ N, 22°28′ E) and at Utsjoki in northern Finland (69°45′ N, 27°01′ E). Pollen was caught from the beginning of May to the end of June in a high-volume air sampler (Hi-Vol SA2000, Graseby Andersen, GMW Inc.,
Vil-Viability and seasonal distribution patterns of Scots pine pollen in
Finland
P. PULKKINEN
1and A. RANTIO-LEHTIMÄKI
21 Foundation for Forest Tree Breeding, Taimitarhantie 34, FIN-76850 Naarajärvi, Finland 2 University of Turku, Department of Biology, FIN-20500 Turku, Finland
Received March 11, 1994
Tree Physiology 15, 515--518
lage of Cleves, OH, USA) with an intake volume of approxi-mately one cubic meter per min. The fiberglass filters (20.3 × 25.4 cm, Whatman) were changed twice a day, at 0800 and 2000 h. Particles were washed from each fiberglass filter onto a Nuclepore Membrafil filter with buffer solution. Each Nucle-pore filter was cut in half and placed on a filter paper over a convex glass placed in a sterile petri dish containing liquid germination medium (Brewbaker and Kwack 1963). The sam-ples were incubated at room temperature in the dark for 4 days and then soaked in 0.1% (w/v) aniline blue in water, which stained the cytoplasm and germ tube blue. By this means, the pollen grains could easily be distinguished from other airborne material. The Nuclepore filter was then cleared with a drop of immersion oil, and pollen grains were classified as either dead (i.e., neither stained nor germinated) or alive (i.e., stained pollen with a germ tube). A total of 400 grains were counted per sample; the percentages of living pollen were calculated separately for each day and night sample.
Airborne pine pollen counts per cubic meter of air were obtained from continuously operating Burkard traps (Hirst 1952; Burkard Manufacturing Co., Rickmansworth, Herts, U.K.) located at Turku and Utsjoki. The traps were situated about 15 m above the ground, corresponding roughly to the location of the female flowers of a pine tree. The numbers of germinating pollen grains per cubic meter were calculated on the basis of the samples caught in the Burkard traps.
Observations on the onset of local male flowering were carried out with rotorod samplers (Perkins and Leighton 1957) in three Scots pine stands, three Norway spruce (Picea abies (L.) Karst.) stands and an open field close to Turku in southern Finland, and in four pine stands and an open field close to the Kevo Research Station in northern Finland. The sites were selected as the places where the local flowering was expected to start first because of favorable growth conditions. Three sampling locations were selected for each sampling site. The rotorod samplers were run for 5 to 10 min at each location. Observations were made once a day per location. All the pollen grains on the sampling tapes were counted with the aid of a microscope and converted to pollen counts per cubic meter (Edmonds 1972).
Results and discussion
The germination rate of airborne pollen was characterized by four phases (Figures 1 and 2). First, a low pollen count with a low germination rate (in Turku, the germination rate was below 30% until the night of May 13); second, a moderate pollen count with a germination rate below 70%; third, a high pollen count with a high germination rate up to 90%; and fourth, a moderate pollen count with a decreasing germination rate. Although the first phase was affected by long-range pollen transport, the count of living pine pollen grains before the peak was low, indicating that long-range transported pine pollen was of little ecological importance in the Turku area in 1993. Probably because of the exceptionally early spring in Turku in 1993, flowering started on the same day in Turku as in the more southerly Estonia (Maret Saar, personal communication),
which is one of the most likely sources of long-range pollen in Finland. The onset of the second phase during May 13--14 coincided with the onset of local flowering, but some of the pollen in the second phase may have originated from distant sources. During the peak days, local pollen predominated. The decrease in germination rates after the peak period was prob-ably caused by pollen released during peak flowering remain-ing airborne for a longer time or the resuspension of pollen from surfaces where it had sedimented, or both.
In the pine stands close to Turku, the first pollen grains were Figure 1. Germination rates of pine pollen for day and night (0800--2000, 2000--0800 h) in southern Finland during the pollen flight of 1993.
Figure 2. Mean daily counts per cubic meter of air for total and germinating pine pollen grains during the 1993 pine pollen period in southern Finland. Data obtained from a Burkard trap in Turku.
observed on May 13, and a day later, some pine pollen was observed in the spruce stands and the open fields (Figure 3). Before May 19, the predominating wind direction varied from south to south-west, and the mean daytime temperature varied between 20 and 25 °C. The peak period for pine pollen in Turku lasted only 3 days (Figure 2). The peak count was on May 20, with over 2000 germinating grains per cubic meter. The wind direction during the peak period was south-west to south-east, and the mean daytime temperature varied from 25 to 29 °C; the mean temperature sum on May 19 was 246 degree days.
Although the germinating pollen count was higher during the day (average 199.7 grains m−3) than at night (average 115.8
grains m−3, Figure 1), the difference was not significant (P =
0.424) (cf. Stanley and Linskens 1974). We conclude, there-fore, that the pollen source during the main pollen season was close to the traps.
The difference between the start of local male flowering in the Turku area and in central Finland (a distance of 250 km) was only about 5 days (Pulkkinen and Rantio-Lehtimäki, un-published observations), and the amount of germinating pollen was high in the Turku area during these 5 days. The finding that flowering in Scots pine is synchronized over a large area supports the possibility of long-range pollination. Female flowers become receptive before the pollen cones on the same tree shed their pollen (Sarvas 1962).
In northernmost Finland, the first pollen grains were ob-served on July 3, and the peak occurred on July 13, with 230 grains m−3 (Figure 4). However, the germination rate
through-out the study was zero probably because the weather in June was exceptionally cold and wet. As a result, pollen release was delayed in 1993 compared with the mean pollen season for the area (Pessi and Pulkkinen 1994). Prevailing winds were from the Arctic Sea.
In the nearby pine stands close to Utsjoki, the number of pine pollen grains detected by the rotorod samplers was low, and the pollination period only lasted from July 12 to 15 (Figure 4). Thus, we conclude that there was no gene transfer from the south to Utsjoki.
Acknowledgments
We thank Irma Kemppainen and Marja-Leena Lahtinen for taking care of the high-volume sampler in Turku, and Kirsti Vuorisalo, Jaakko Heino and Jussi Heino for assistance in Utsjoki.
References
Brewbaker, J.L. and B.H. Kwack. 1963. The essential role of calcium ions in pollen germination and pollen tube growth. Am. J. Bot. 50:859--865.
Comtois, P. 1995. Pollen dispersal and long distance transport: the case of thermophilic pollen in subarctic Canada. Aerobiologia. In press.
Cook, S.A. and R.G. Stanley. 1960. Tetrazolium chloride as an indica-tor of pine pollen germinability. Silvae Genet. 9:134--136. Edmonds, R.L. 1972. Collection efficiency of rotorod samplers for
sampling fungus spores in the atmosphere. Plant Dis. Rep. 56:704--708.
Hirst, J.M. 1952. An automatic volumetric spore trap. Ann. Appl. Biol. 36:257--265.
Hjelmroos, M. 1991. Evidence of long-distance transport of Betula
pollen. Grana 30:215--228.
Ho, R.H. 1992. Quantifying the effect of viability of contaminant pollen. In Challenges in Pollen Dispersal and Pollen
Contamina-tion. Eds. F. Di-Giovanni and D. Joyce. Canadian Climate Center, Environ. Can., Egbert, Ontario, Rep. CCAD-92-008, pp 47--50. Jensen, C.J. 1964. Pollen storage under vacuum. Danish Atomic
En-ergy Comm., Kongelige Veterin, Landbohjskole Årsskrift, pp 133--146.
Figure 3. Pine pollen counts from rotorod samples around the Turku area during the 1993 pine pollen period. Sampling sites (n = 9) were
located in pine and spruce stands and in open fields.
Figure 4. Mean daily counts of pine pollen per cubic meter of air from Burkard trap samples at Kevo, northern Finland, and rotorod pollen counts from the area around Kevo, near Utsjoki.
Jett, J.B., D.L. Bramlett, J.E. Webber and U. Eriksson. 1993. Pollen collection, storage and testing. In Advances in Pollen Management.
Eds. D.L. Bramlett, G.R. Askew, T.D. Blush, F.E. Bridgwater and J.B. Jett. USDA Forest Service, Washington, D.C., pp 41--46. Perkins, W.A. and P.A. Leighton. 1957. The rotorod sampler. 2nd
Semiannual Rep. No. CML 186, Aerosol Lab., Stanford University, Stanford, CA, USA, pp 1--60.
Pessi, A-M. and P. Pulkkinen. 1994. Temporal and spatial variation of airborne Scots pine pollen. Grana 33:151--157.
Sarvas, R. 1962. Investigations on the flowering and seed crop of
Pinus sylvestris. Comm. Inst. For. Fenn. 53:7--198.
Stanley, R.G. and H.F. Linskens. 1974. Pollen: biology, biochemistry and management. Springer-Verlag, New York, 307 p.
The Finnish Pollen Bulletin. 1976--1993. Vols. 1--18. University of Turku, Department of Biology, Turku, Finland.
