ACTA PHYTOGEOGRAPHICA SUECICA 74 EDIDIT
SVENSKA V AXTGEOGRAFISKA SALLSKAPET
lngvar Backeus
Aboveground production and growth dynamics
of vascular bog plants in
Central Sweden
ACTA PHYTOGEOGRAPHICA SUECICA 74 EDIDIT
SVENSKA V AXTGEOGRAFISKA SALLSKAPET
lngvar Backeus
Aboveground production and growth dynamics
of vascular bog plants in
Central Sweden
Almqvist & Wiksell International, Stockholm UPPSALA 1 985
Doctoral thesis at Uppsala University 1 985 ISBN 9 1 -72 10-074-5 (paperback)
ISBN 9 1 -72 1 0-474-0 (cloth) ISSN 0084-5914
Backeus, I. 1985: A boveground production and growth dynamics of vascular bog plants i n Cen tral Sweden. Acta Phytogeogr. Suec . 74, 98 pp. ISBN 9 1 -7210-074-5; ISBN 9 1 -7210-740-0.
Aboveground primary production and biomass of the field layer plants were studied on an om brotrophic mire in the central Swedish uplands. The study was made on the population level, and results on the population ecology of certain species were also obtained. The study included
A ndromeda polifolia, Betula nana, Calluna vulgaris, Drosera anglica, D. rotundifolia, Empet rum nigrum s.str . , Eriophorum vaginatum, Rhynchospora alba, Rubus chamaemorus, Scheuchzeria palustris, Trichophorum caespitosum, Vaccinium microcarpum, V. oxycoccos
and V. uliginosum. Production and biomass per unit area of each species and of modules (leaves, inflorescences, etc .) of certain species were determined from figures on mean individual weight and mean density. Aboveground biomass of vascular plants was estimated at 2700 kg· ha·1 on hummocks, 682 kg· ha·1 in 'lawns' and 3 10-390 kg· ha·1 in two kinds of 'carpets'
(Cuspidatetum dusenietosum and C. tenelletosum Fransson, respectively). Total aboveground production of vascular plants was 830, 6 1 0, 360 and 340 kg · ha·1 • year·1, respectively.
Growth in some species was followed throughout the growing season through repeated har vesting. Seasonal variation in weight of individual leaves was similarly followed in evergreen species. Length growth of shoots of six species and length growth of leaves in two monocots were followed through direct measurements. Length growth rate of B. nana and Calluna shoots and of Scheuchzeria leaves was shown to be closely dependent on temperature, while growth of E. vaginatum leaves was not.
Two peaks in production were found: ( 1 ) during shoot formation in June and (2) in August when perennial leaves were becoming winter-hardened and wood increment in Calluna and A n dromeda (remarkably late) took place. Seasonal changes in biomass were comparatively small because of evergreenness in the dominant species.
Survivorship of leaves of certain species was studied. Flowering was poor and seedlings absent in most species. Instead plants were propagated vegetatively, and different means for such pro pagation are discussed. The interactions between the field and bottom layer plants are also dis cussed, notably how the former avoid being overgrown by mosses .
Ingvar Backeus, Institute of Ecological Botany, Box 559, S-751 22 Uppsala, Sweden
© Ingvar Backeus 1 985
Svenska Vaxtgeografiska Sallskapet Box 559, 75 1 22 Uppsala
Editor: Erik Sjogren
Technical editor: Gunnel Sjors Phototypesetting:
Textgruppen i Uppsala AB Printed in Sweden 1 985 by Borgstroms Tryckeri AB , Motala
Contents
The study area 7
Topography and geology 7
Climate and weather 7
Temperature 7, Precipitation 1 1
Vegetation 1 2
Description o f the sampling areas 1 5
Phenological development 1 7
Methods o f collecting and presenting the phenological data 1 7
Results and discussion 1 7
Production and dynamics of individual species 20
Methods 20 Andromeda polifolia 2 1 Betula nana 26 Calluna vulgaris 28 Carex limosa 33 Carex pauciflora 33 Drosera anglica 33 Drosera rotundifolia 34 Empetrum nigrum 35 Eriophorum vaginatum 39 Rhynchospora alba 48 Rubus chamaemorus 5 1 Scheuchzeria palustris 56 Trichophorum caespitosum 60 Vaccinium microcarpum 62 Vaccinium oxycoccos 63 Vaccinium uliginosum 64
Field layer density, biomass and production 68
Discussion on methods 68
Density 72
Mean total aboveground biomass and production 72
The seasonal course of the total aboveground production and changes
in the total aboveground biomass 77
The seasonal course of production 77, Seasonal changes in biomass 77
Variations between years in production 80
The dependence on environmental variables of length growth in stems
4 lngvar Backeus
The bog environment and the behaviour of plants Rate of production
Flowering and reproduction Vegetative propagation Moss overgrowth
Grime's C-, S- and R-selection Age structure of modules Interdependence of ramets Concluding remarks References 84 84 85 86 87 88 89 90 9 1 92
Introduction
Works in production ecology, e.g. within the Inter national Biological Programme, have often been ecosystem-oriented . Great efforts have been made to obtain figures on the total biomass and produc tion in various ecosystems. The work input in this kind of investigation is considerable also when only moderate resolution is accepted and time-consum ing harvesting, sorting and weighing are necessary. It is also possible to make production ecological studies on the population level . By working out suit able techniques for all or the more important spe cies , figures on total production can be obtained also by such an approach, which is attempted in this treatise .
The aim is somewhat different than in an ecosystem-oriented study, even though populations of organisms are often used as a basis also in the lat ter type of investigation. Other variables are measured and information on density, age struc ture, growth rhythm and growth rate is obtained .
In contrast to animals, higher plants do not have a definite size. The size instead varies with the en vironment, both the abiotic and the biotic. The indi vidual ' modules' of the plant, leaves , flowers , etc . usually vary much less (cf. Harper 1 978) . The study of these metapopulations (White 1 979) refines the production ecological methodology and increases our knowledge of the survival potential of the indi vidual .
Population ecologists usually work with only one or a few , similar or contrasting, species. Examples that will be discussed later are found in Fetcher & Shaver ( 1 983), Flower-Ellis ( 1 97 1 ) , Karlsson ( 1 982), Noble et al . ( 1 979), Robertson & Woolhouse ( 1 984a,b), Sarukhan & Harper (1973) and Schmid ( 1 984) . The response of these species to environ mental factors in different environments is studied . It has been less common to study the response to en vironmental factors of all the different populations in a community. Evidently the work on each species cannot be very intense in such a study and has to fo cus on the main points, but nevertheless it might be
possible to draw more general conclusions about the response of plants to the particular environment that was studied .
The aim of my investigation has been to find out the distribution of production in time and space (within a plant and within the site) of all field layer plants (including Betula nana) on an ombrotrophic mire . The amount of work and numerous methodo logical problems made it necessary to exclude the bottom layer and the rhizosphere as well as decom position at this stage. This limitation of course has caused gaps in my results that will have to be filled before a reasonably good understanding of the pro duction on the bog can be obtained .
The ombrotrophic mire is an extreme environ ment . There is no input of nutrients except through precipitation . There are several reasons for this choice of study object, besides the obvious fact that ombrotrophic mires constitute important ecosys tems in Scandinavia and are interesting per se:
The vegetation is poor in species and all species are fairly well distributed within the community. Such a simple system makes it easier to work out methods, and the low number of species makes it possible to study all species .
Few environmental factors vary within one site.
The water level varies in space in an obvious and easily understandable way and lack of water is rare for the field layer plants . The chemistry of the sub strate is very uniform. In time, temperature is an im portant variable factor and in space (and time) the overgrowth by mosses.
The ombrotrophic bog is an unusually well-de fined ecosystem and this was also a reason for my choice. Within a restricted area this ecosystem re curs from site to site with very little variation. In a larger area, e.g. Europe or Holarctis , the variation is certainly greater but still moderate and usually successive. Comparisons with investigations from other places are therefore easy to make.
As a mire ecologist I should also explain why I have chosen to study production. The mire plants Acta Phytogeogr. Suec. 74
6 lngvar Backeus
form their substrate themselves through production and decomposition. Information on total produc tion and decomposition must therefore be essential for a better understanding of the mechanisms be hind the rise of the mire surface and its differentia tion into hydromorphological structures, of the re lations between these structures and of their se quence in time . Production and production proces ses are therefore central problems in mire ecology. Here it must again be emphasized that my investiga tion is still in its beginning . The aboveground parts of the vascular plants contribute very little to the peat formation . Peat is mainly formed by mosses and below ground parts of vascular plants and a lot of work therefore remains to be done .
Important constituents of mire ecology are also supply and transport of mineral nutrients (see e.g. Malmer & Nihlgard 1 980) and translocations of or ganic nutrients . Neither of these aspects are treated here.
I have had the opportunity to carry out the field work on a bog that was previously thoroughly inves tigated ecologically, the ' Special Area' of the Skatt losberg Stormosse (Sjors 1 948) . Sjors described in detail the vegetation and the distribution of plants and mineral elements in this area. This made it poss ible for me to concentrate from the beginning on production studies without lengthy data collections concerning the vegetation and environment.
The field work was carried out in 1 980, 1 98 1 and 1 982, thus giving figures from three successive growing periods.
Nomenclature for vascular plants follows Moore ( 1 982) , except for Empetrum nigrum v. hermaphro ditum, which is here treated as a species (E. her maphroditum Hagerup) and Scirpus cespitosus, which is here called Trichophorum caespitosum (L.) Hartm . Nomenclature for bryophytes follows Cor ley et al. ( 1 98 1 ) and Grolle ( 1 976) and for lichens Santesson ( 1 984) .
The study area
The bog Skattlosberg Stormosse is located in the southwestern part of the province of Dalarna (Kop parberg County) in central Sweden (Fig. 1 ) . The physical conditions of Bergslagen (i.e. S Dalarna, N Vastmanland and E Varmland) in general and espe cially of the Skattlosberg Stormosse were dealt with by Sj ors ( 1 948: 1 6-32 and 68- 1 04 in Swedish; 277-278 and 282-284 in English) . Also the vegeta tion and flora of the Skattlosberg Stormosse were described in detail by Sjors (op . cit .: 105- 1 7 1 in Swedish; 284-286 in English) . For further details, reference should be made to Sjors' s publication.
Topography and geology
The Skattlosberg Stormosse is situated at an eleva tion of 265 to 285 m a.s.l. The geology of the area was described by Magnusson & Lundqvist ( 1 933). The mire rests on glacio-lacustrine sand and sandy ablation till.
The SkattlOsberg Stormosse is a large, 450 ha, mire complex (Sjors 1 948) . It consists of a number
Fig. 1 . Map of Central Sweden showing the lo cation of the Skattlosberg Stormosse.
of excentric bogs separated by fen soaks . The soaks, except the southern ones, run towards a central big soak sloping east and partly consisting of large flarks . A map of the mire was presented by Sjors ( 1 948) .
Climate and weather
TemperatureIn June 1 98 1 I set up a temperature screen with a thermohygrograph in the middle of the Special Area (ea. 280 m a.s.l.). Data were collected during 198 1 and 1 982, but with several breaks . In some periods maximum and minimum temperatures were also re corded from thermometers.
The aim of these measurements was to achieve a picture of the temperature conditions during the years of data collecting and to determine the local temperature climate.
The data had to be completed through a series of adjustments and interpolations . First the minimum (n = 35) and maximum (n =52) temperatures read
\ \ ' ' I I "',,
',-,�:1
\
Dalarna (/ Viirmland \ \ \ I ' \ \ '-,... )( �--, .... '"'';' ' ' I '-' I I ... � .... _, ',, I I ': \ ... ..'- I : vastmanland I I, / I / ; ... --- __l-/5\
,,_) -,' -' I Uppland8 Ingvar Backeus
from the thermometers were compared to minimum and maximum temperatures read from the graph on the same day. The mean deviation was used to cor rect the minimum temperature values given by the graph on days when thermometer readings were not available.
Then minimum (n = 1 85) and maximum (n = 1 78)
temperatures from the thermometers (when avail able) or from the graph and temperature values for 0700 (n = 1 80) , 1 300 (n = 1 82) and 1 900 (n = 1 83)
hours read from the graph (corrected twice a week by means of a mercury thermometer) and daily mean temperatures (calculated according to SMHI 1 966) (n = 1 57) were compared with values from the
meteorological station at SUilldalen, ea. 30 km SE of the Skattlosberg Stormosse (59° 57' N, 14 o 57' E ;
2 1 0 m a.s.l. ; SMHI, unpubl .). The mean differences were calculated (Table 1 a) . SUilldalen was chosen because it is the nearest station with a fairly long
se-Table la. Mean differences (X) with standard deviations of the samples (s) and standard errors of the means (S .E.) between tem perature at Stalldalen and at the Skattlosberg Stormosse. (X is positive when the Stalldalen value is higher .)
n X S.E. 0700h 1 80 -0. 5 2 . 2 0.2 1 300h 1 82 0.3 1 . 3 0. 1 1 900h 1 83 0.3 1 . 5 0. 1 max 1 78 0.3 1 . 3 0. 1 m in 1 85 1 . 1 1 . 8 0. 1 daily mean 1 57 0.2 0 . 8 0. 1
Table l b . Distribution of deviations between recorded
tempera-tures on the Skattlosberg Stormosse and temperature values from the same days using the mean difference between the bog and
Stalldalen (Table la). Negative deviations mean that the recorded
temperature was lower than the calculated.
deviation maximum minimum daily mean
-7 0 1 0 -6 1 2 0 -5 0 1 0 -4 3 3 0 -3 4 9 1 -2 1 0 1 8 4 - 1 27 35 38 0 72 55 73 + 1 48 3 1 35 + 2 1 0 1 7 4 + 3 2 8 2 + 4 0 4 0 + 5 0 2 0 + 6 0 0 0 + 7 0 1 0 + 8 0 0
ries of data. It is situated ea. 70 m lower than the screen on the bog . All hours are given as GMT + 1 h.
The difference in the daily means at the two sta tions is small, as are the differences in maximum temperatures and temperatures at 1 300 and 1 900 hours . The temperature at 0700 hours deviates and is-unlike all other temperatures-higher on the bog than at SUilldalen. The readings at the meteoro logical stations are often made up to 1 5 minutes be fore the full hour (E . Schmacke, in litt.). Tempera ture rises rapidly in the mornings and the deviations can be explained in this way.
There is also a notable difference in minimum temperatures between the stations . The minima on the bog are lower on average, certainly an effect of its situation in a depression. This effect depends largely on the weather and the deviations from the mean are therefore pronounced (Table 1 b).
After thus having compared data from the Skatt losberg Stormosse and SUilldalen, missing data from the bog were calculated from the SUilldalen values by adding the mean differences between the stations . In that way a complete table for the bog was constructed which was used for estimating the number of frost nights (Table 2), length of growing season (Table 3a) and the sum of effective tempera tures (Fig . 4) in the years 1 980- 1 982.
Monthly mean temperatures (which differ from the monthly means of the daily mean temperatures) for the months April to October were then calculat ed for the years 1 980- 1 982. These temperatures were compared with the corresponding values from Stalldalen and were found to be 0.2°C lower .
The mean temperatures at Stalldalen can be con sidered normal for the area and elevation. Sj ors ( 1 948) found (preliminarily) that the mean tempera ture in Bergslagen (the central Swedish uplands) decreases ea. 0.65 ° C per 1 00 m elevation . The dif ference between Stalldalen and the bog is thus some what smaller than expected .
The means of the monthly mean temperatures at Stalldalen over the years 1 967- 1 982 (the station was not in use before 1 967) were calculated from the yearbooks of the Swedish Meteorological and Hydrological Institute for the years 1 967- 198 1 (SMHI 1 968- 1 982) and from their unpublished lists for 1 982 (Fig . 2) . When the differences in monthly means between the stations are added to
Production and growth dynamics of vascular bog plants 9 STALLDALEN (210m) [16] 725mm mm 80 40 20
Fig . 2. Climate diagram in the sense of Waiter & Lieth (1 960) for SHilldalen, ea. 30 km SE of the Skattlosberg Stormosse.
these figures the means of the monthly mean tempe ratures of the Skattlosberg Stormosse are obtained. It was found to be + l5°C in July and approxi mately -6. 5°C in January.
The extrapolation made the temperature data somewhat uncertain. Most recordings on the bog were made during the April - October period and other months were therefore excluded from the calculations. Tables la and 1 b give information on the variation in the material. The rather poor corre lation concerning the minimum temperatures is un fortunate, as it makes it difficult to estimate the number of frost nights for periods when records from the bog are not available . As for the daily
-i CD 3 30 "0 CD
�
c: ii) 20 a 10 b c d e -10 AM..JJASO 1980 AMJ ..JASO 1981 AM..JJASO 1982Fig . 3. Means of the five highest maximum temperatures per month (a) , monthly means of daily maximum tempe rature (b), mean monthly temperatures (c) , monthly means of daily minimum temperature (d) and means of the five lowest minimum temperatures per month (e) on the Skattlosberg Stormosse in 1980- 1 982. Only the months April to October are included . From own record ings and calculations based on figures from the meteoro logical station at SUilldalen . See further in text.
mean temperature, on the other hand , the correla tion is much closer . This is to be expected, as this temperature in itself is the (weighted) mean of three or four recordings .
The monthly mean temperatures and the monthly means of maximum and minimum temperatures for 1 980- 1 982 are given in Fig . 3. The same figure also shows the monthly mean of the five highest maxi mum temperatures and of the five lowest minimum temperatures . It was not considered meaningful to depict the absolute maxima or minima, since the in dividual values are often calculated from the Stall dalen record .
The estimated number of frost nights is given in
1 0 Ingvar Backeus
Table 2 . Number of frost nights with minimum temperatures
be-low different temperature limits in I 980- I 982. Numbers in
pa-renthesis refer to values based on the records at Stalldalen.
Growing season defined according to Perttu et al. ( I 978a) with
threshold value + 5 ac.
temp. month limit I 980 I 98 I I 982 April 0 (24) (28) 1 0( + I 6) - I (24) (25) 8( + I 2) -2 (2 I ) (22) 7( + 1 0) -3 ( 1 6) (2 I ) 6 ( + 1 0) May 0 (I5) (9) 8 ( + 3) - I ( 1 3) (8) 4 ( + 2) -2 ( 1 0) (7) 4 ( + I ) -3 (8) (5) I ( + I) June 0 0 0 I I - I 0 0 I I -2 0 0 6 -3 0 0 4 July No frost August 0 0 2 0 - I 0 0 0 September 0 (2) 5 (5) - I ( I ) 3 (3) -2 0 2 (2) -3 0 2 ( I ) October 0 ( 1 9) 6( + 7) (5) - I ( 1 6) 5( + 6) (3) -2 ( I 2) 4( + 4) (2) -3 (I I ) 3 ( + 3) ( I ) growing 0 (26) (4) I 8 season: - I (24) (3) I 5 spring -2 (20) (3) 9 -3 ( 1 5) (2) 5 growing 0 (6) 8 ( 1 0) season: - 1 (2) 3 (6) autumn -2 ( 1 ) 2 (4) -3 ( I ) 2 (2)
Table 2. It should be noted that an individual frost night deduced from the Stalldalen record is uncer tain, especially if the calculated temperature is -2°C or higher. Light and more severe frosts have therefore been separated in the table. Judging from Table la, an extrapolated value of -3 °C is likely to correspond to a real value below zero .
According to Sjors ( 1 948 : 23) , spring frosts are re markably rare in Bergslagen after the middle of May. This holds true for 1 98 1 , when the last severe frost occurred around May 1 0. There was an un usual snowfall on June 1 2 this year but without frost. In 1 980 there were at least two frost nights in the second half of May. The year 1 982 was extreme. After some very hot days in early June, there was a long period with an unparalleled number of
sum-Table 3a. Start, end and duration in days of the growing season in 1 980- 1 982. The start and end are defined as the four first and
last consecutive days with a mean temperature on or above the
threshold value (according to Perttu et al . 1 978a).
threshold value oc 0 3 5 6 1 0 1 980 start 4.IV 1 2.IV 1 3 . IV 3 . V 1 5 . V
end 20.X I 3 .X I 2.X 9.X 1 0. IX duration 200 1 85 I 83 I 60 1 1 9 I 98 I start 3 1 .111 9.IV 6.V 8.V 1 0.V end 3 1 .X 14.X I O.X 6.X I . X duration 2 I 4 1 89 1 58 I 52 I 45 I 982 start 23 . I ll I 5 . I V 5 . V 1 2.V 25 .V
end 28.XI 1 3 .XI 2.XI 7.X 2 l . IX
duration 25 I 2 1 3 I 82 I 49 1 20
Table 3b. Mean start, end and duration of the growing season for the period I 967- I 982. The start and end are defined graphic ally (according to Langlet I 935).
threshold value ac 3 5 6 1 0
I 967- I 982 start 1 8 . IV 28 . IV 2.V 22.V
end ea. 20.X 8.X 2.X 8 . IX
duration ea. 1 85 1 63 1 53 1 09
mer frosts. From June 7 to June 23 frost occurred eleven times . On six occasions the frost temperature lasted four hours or more.
The autumn frosts of 1 980 were few and light un til October . The first one probably occurred on Sep tember 1 . In 1 9 8 1 the first light frost occurred on August 25 . In the unusually mild autumn of 1 982 the first frost probably occurred around September 6. Growing season: The start, end and duration of the growing season or period is given in Table 3. Sev eral threshold values are used. in the table, as the choice is rather arbitrary. There are different ways of defining start and end of the growing season (cf. Tuhkanen 1 980: 1 3) . Supan ( 1 887) , Hamberg ( 1 922) , Langlet ( 1 935) and others defined these dates graphically from the curve of the monthly means . What these authors determined was the mean length of the growing season. Perttu et al. ( 1 978a) defined the growing season as the time when the daily mean temperature continuously exceeds the threshold value . 'Continuously' means four consecutive days counting from the first day (in spring) to the last (in autumn) . This definition can be advocated especially when the start and end of the growing season of a single year is to be defined , and i t has been used i n Table 3a.
pe-Production and growth dynamics of vascular bog plants 1 1 riod 1 967- 1 982 has been determined graphically (Langlet 1 935) from the corrected data from SHill dalen. The results are shown in Table 3b. The
-1100 figures are normal for the area according to Ang
strom ( 1 953). Perttu et al . ( 1 978b), using their own method, obtained a mean length of the growing sea son at Stalldalen which is two weeks longer. -1000 -900 -800 -700
�
-eooe
::I (I) Q)a
Ill a; c. -500 E Q) 1--400 -300 -200 -100 -0I am aware that the growing season has a restrict ed ecological value (cf. e.g. Hytteborn 1 975: 7). It is included here foremost as a means of comparing the temperature conditions of the Skattlosberg Stor mosse with other sites from which corresponding data are available.
The effective temperature sum (T) was calculated and is here, following i.a. Kolkki ( 1 966) and Sarvas ( 1 967), defined as the cumulative sum of daily mean temperatures above + 5 °C during the growing sea-son:
n
T+5°C =
m�l Um-
5)(n = number of days;
tm
= mean temperature of the m:th day) .Results are shown in Fig . 4, following the ap proach of Lindholm ( 1 980).
Precipitation
Precipitation data from the years 1 980- 1 982 are available from Stalldalen and from Fredriksberg ( 1 8 k m W o f the SkattlOsberg Stormosse; 60°08'N, 1 4°22'E; 300 m a.s .l. ; SMHI 1 982-83 and un publ.). The monthly precipitation at these stations is given in Table 4. Its percentage of the normal amount of precipitation is also given for Stalldalen (SMHI , unpubl.). No severe drought occurred dur ing these three years .
The former precipitation station at the village of Skattlosberg recorded a mean yearly precipitation of 726 mm for the period 1 92 1 - 1950 (Bergsten 1 954). The rain gauge was placed ea. 3 km E of my sampling areas and at a higher level . The precipita tion in this hilly landscape varies considerably from place to place due to elevation and exposure (cf. Sj ors 1 948 : 25-32). A more elevated area with Fig. 4. The progression of the temperature sum (T+5oc) in 1 980- 1 982. The beginning of each day is represented with a bar. + indicates frost nights (only between May 1 5 and Sept . 30). ( +) indicates weak frost (not below - 1 °C).
12 lngvar Backeus
Table 4. Precipitation at SUilldalen and Fredriksberg 1 980- 1982. OJo = percent of normal precipitation.
Jan . Feb . Mar. April
Stalldalen 1 980 mm 12 1 8 23 24 2 1 0 m a.s.l. OJo 23 47 75 57 1 98 1 mm 24 32 56 1 5 OJo 45 85 1 88 37 1 982 mm 32 3 1 62 6 1 OJo 6 1 82 207 1 45 normal mm 53 38 30 42 Fredriksberg 1 980 mm 12 19 31 24 300 m a.s.l. 1 98 1 mm 25 29 74 1 6 1 982 mm 34 35 78 59
some hills reaching about 470 m a.s.l. lies W of the Skattlosberg Stormosse. This area probably re ceives a precipitation exceeding 800 mm per year , as is suggested on the precipitation map published by Sjors ( 1 948 : 26) . The station at Skattlosberg was situated at an elevation of 330 m a.s.l. on a hill which reaches 366 m a.s.l. This hill occupies a rather small area and should have only a minor effect on the pre cipitation, especially considering that the clouds of ten lose some of their moisture content already over the hills W of the mire. (Precipitation is highest when winds are southwesterly.) The amount of pre cipitation in the sampling areas (at about 280 m a.s. l.) can thus be anticipated to be close to or some what lower than the amount at the former Skattlos berg precipitation station .
An index of humidity cannot be very accurately determined from the data available. ' Humidity' de fined as surplus precipitation (0. Tamm 1 959) gives a value close to or somewhat lower than 400 mm, which is rather high for central Swedish conditions , but typical for the area.
Vegetation
The ombrotrophic bog vegetation of the Skattlosberg Stormosse was described in detail by Sjors ( 1 948: 1 07-1 07-1 6). Here I will attempt to put Sjors's descriptions into the context of later works on bog vegetation.
Sjors followed the principle of separation between ombrotrophic (bog) and minerotrophic (fen) communi ties which was first suggested by Du Rietz in 1 933 and later further elaborated on a number of occasions (Du Rietz 1 949, 1 950a-c, 1 954) . Du Rietz ( 1 954) introduced the
names Ombrosphagnetea for the ombrotrophic mire ve
getation and Sphagno-Drepanocladetea for the
minero-May June July Aug . Sept. Oct. Nov . Dec. year
34 77 67 1 52 59 135 44 37 6 1 74 1 1 4 40 84 92 1 2 1 66 57 686 1 63 47 90 1 2 1 1 8 1 98 95 95 1 27 55 42 39 1 23 1 1 5 49 744 1 8 1 65 45 5 1 1 84 1 7 1 82 103 57 65 7 1 89 62 1 1 4 74 778 8 1 77 76 1 1 7 92 1 7 1 1 24 1 07 70 85 93 76 67 67 60 725 1 7 1 50 1 1 4 6 1 1 62 82 70 833 1 95 68 3 1 52 1 25 1 23 48 847 50 46 88 101 53 1 1 8 77 8 13
trophic mire vegetation, i.e. he treated these entities as classes. Malmer ( 1 968) pointed out that Du Rietz's classi fication is very good "from the point of view of habitat ecology' ' but ''as to phytosociology it is not so well found ed" (transl.).
Nevertheless, due to its obvious ecological merits the system is now in general use in Sweden and in several places elsewhere. It should be considered primarily as an ecological classification rather than a sociological one and it is symptomatic that Du Rietz's syntaxonomical names are rarely used .
An entirely different system has been developed in Cen tral and Western Europe, where both ombrotrophic and minerotrophic vegetation are found in the same classes , the Oxycocco-Sphagnetea and Scheuchzerio-Caricetea nigrae (e.g. Westhoff & den Held 1 969, Neuhausl 1 972)
and sometimes also in the Vaccinio-Piceetea. Malmer has
adopted the Central-European syntax system, at least for S Sweden.
Sjors ( 1 948) pointed out the transitional position of wooded bogs between moist dwarf shrub-conifer forest on mineral soil and treeless hummocks. In Finland (fol lowing Cajander 1 91 3) all wet forests have been classified as mires. In central Europe, on the other hand, it has been common to emphasize the connections with forest types and, as a consequence, to put wooded bogs and treeless bog hummocks in different classes , the Vaccinio-Piceetea
and Oxycocco-Sphagnetea, respectively. Nevertheless,
the phytosociological uniting of wooded and woodless bog areas has got a footing also in Continental phyto sociology through Neuhausl ( 1 972; cf. Malmer 1 968). Neuhausl's classification was, at least in this respect, ac cepted by Dierl3en ( 1 977). The problem is further discus sed by Dier13en & Dierl3en ( 1 982).
My investigation is only concerned with ombrotrophic
vegetation . The class Ombrosphagnetea was divided by
Du Rietz ( 1 949) into four regional types with relevance to Sweden south of the Norrland terrain. He treated them
as subformations (i.e. subclasses sensu Braun-Blanquet) .
The subformations were further divided into alliances
characterized by the presence or absence of Pin us sylves
Production and growth dynamics of vascular bog plants 1 3 subformation . The following alliances are of interest in
the present area:
( 1 ) The Parvifolion alliance (Du Rietz in Waldheim 1 944, Du Rietz 1 949; cf. Du Rietz 1 950a) of the pine bogs. A variant of this alliance without Ledum palustre charac terizes the southwest Swedish pine bogs.
(2) The Eufuscion alliance (Du Rietz in Waldheim 1 944, Du Rietz 1 949; cf. Du Rietz & Nannfeldt 1 925 , Du Rietz 1 950c) of the open bog plains in eastern south Swe den with Sphagnum fuscum hummocks and S. balticum and S. cuspidatum in the upper and lower parts of the hol lows, respectively.
{3) The Rubello-fuscion alliance {Du Rietz in Wald heim 1 944, Du Rietz 1 949; cf. von Post & Sernander 1 910, Du Rietz 1 950b, Backeus 1 972) of the open bog plains in the central parts of south Sweden, where Sphagnum ru
bellum and, in some areas , S. magellanicum have super
seded the species of the former alliance from the lower parts of the hummocks and the higher parts of the hol lows .
The Skattlosberg Stormosse, being situated north of the border of the north Swedish uplands (the 'Norrland terrain'), does not conform fully to any of these vegeta tion types. The wooded bog areas belong to the Parvifo
lion. Ledum palustre is rare but is abundant in similar ve
getation a short distance to the east. The bog further devi ates from the typical Parvifolion in the presence of the
northern species Betula nana. The community was called
the Pinus - Vaccinium bog association by Sjors { 1 948 : 107- 109). Fransson ( 1 972: 35-36) put similar vegeta tion into the association Vaccinietum uliginosi.
In treeless bog vegetation there are gradual changes around the border of the north Swedish uplands as de scribed by Sjors ( 1 948), Fransson ( 1 972) and Backeus ( 1 984) . The Sphagnum rubellum zone of the hummocks of the Rubella-juscion vanishes towards the north and S.
majus occurs together with S. cuspidatum in wet hollows .
Cetraria delisei also occurs .
The treeless bog areas on the SkattlOsberg Stormosse are intermediate between the Eufuscion and the Rubello fuscion with the addition of the northern features j ust
mentioned (cf. Sjors 1 948: 1 1 1 and Malmer 1 962: 148). Sphagnum rubellum here forms only a narrow zone along the border between hummocks and hollows . S. magellani cum occurs both on hummocks and in hollows but is never dominant. Trichophorum caespitosum is an important constituent here as in the Rubello-fuscion . (The species does not occur in the Eufuscion .)
The open bog areas on the SkattlOsberg Stormosse were divided by Sjors into one hummock and three hollow communities, corresponding to the lawns (upper parts), carpets (lower parts) and mud-bottoms (without sphagna) .
Hummocks: The hummock vegetation was called by Sj ors ( 1 948) the Calluna - Cladonia - Sphagnum fuscum as sociation. Similar vegetation was described by Du Rietz (1 950b ,c) as the association Calluneto-fuscetum , by Mal mer (1 962) under the names Eriophorum vaginatum
-Sphagnum rubellum small association and Eriophorum
vaginatum - Cladonia small association and by Fransson
( 1 972) under the name Calluno-fuscetum.
In the Special Area Calluna vulgaris practically always
dominates the hummock community. Empetrum nigrum
dominates locally, mostly on high hummocks. Otherwise the latter species usually occurs as a subordinate species and in places it is entirely lacking. A ndromeda polifolia,
Eriophorum vaginatum and Rubus chamaemorus are pre
sent more or less everywhere. Vaccinium uliginosum has an uneven distribution and mainly occurs near the scatter ed pines . V. microcarpum, V. oxycoccos and plants inter mediate between them occur frequently. Trichophorum
caespitosum is mostly lacking on higher hummocks but
is common in lower parts. Betula nana occurs sporadically and usually in patches. Drosera rotundifolia is common but absent from the lichen-dominated facies of the com munity. The bottom layer is usually dominated by Sphag
num fuscum (S. fuscum facies Fransson 1 972) but locally
by Cladina spp . (Cladonia facies Fransson) or liverworts
(esp . Mylia anomala; liverwort facies Fransson). Lawns and carpets: The lawn vegetation of the Skattlos berg Stormosse was described by Sjors ( 1 948) as the Erio phorum vagina tu m - Scirpus caespitosus ( = Tricho
phorum caespitosum) Sphagnum rubellum balticum -cuspidatum association. The lawn communities are de limited against hummock communities by the absence (except a few colonisers) of Calluna vulgaris (cf. Du Rietz 1 949). In Malmer's (1 962) scheme of comparison with other authors Sjors's limit between lawns and hummocks is drawn higher than the Calluna limit, indicating that the Trichophorum caespitosum variant of the Eriophorum vaginatum - Sphagnum rubellum small association would be a lawn community sensu Sjors . The opinions of the exact position of the limit between hummocks and lawns are thus deviating, but the difference is probably some what exaggerated in Malmer's scheme (Sjors, pers. comm .).
The bog carpet vegetation constituted Sjors' s ( 1 948) Scheuchzeria Rhynchospora alba Carex limosa -Sphagnum cuspidatum - Dusenii { = majus) association .
Sjors used characters from the field layer when delimiting the bog carpet communities from the bog lawn communi ties . His differential species occurring in the carpets were Scheuchzeria palustris, Rhynchospora alba, Carex limosa and Drosera anglica.
Du Rietz included all bog hollow vegetation in his asso ciation Cuspidate/urn (Du Rietz 1 949). He described a progressive upper hollow stage of the Rubello-fuscion as the subass. Magellanico-cuspidatetum (Du Rietz 1 949) and of the Eufuscion as the subass. Baltico-cuspidatetum {Du Rietz 1 950b) . He further described a regressive upper hollow stage as Tenello-cuspidatetum (Du Rietz 1 949, 1 950a) in both alliances. The bog carpets of all the al liances were his subass . Eucuspidatetum (Du Rietz 1 949,
1 950c; cf. Du Rietz 1 950b). Contrary to Sj ors's delimita tions, the composition of the bottom layer determined the
1 4 Ingvar Backeus
boundaries between Du Rietz's subassociations of the
Cuspidatetum (Du Rietz 1 949) .
Fransson (1 972) described similar lawn vegetation from SW Varmland under the name Rubello-tenelletum. Car pets and mud-bottoms together constituted another asso ciation which he called the Cuspidatetum. Within this he distinguished three subassociations, two of which were carpet commumtles: Cuspidatetum dusenietosum (Sphagnum dusenii = S. majus) and Cuspidatetum tenel
letosum . The former, as described by Fransson, has a very
characteristic physiognomy with dense carpets of Sphag num cuspidatum and S. majus and a field layer with
Scheuchzeria palustris, which is often the only field layer species . Carex limosa occurs in places. Cuspidatetum te nelletosum is dominated by Rhynchospora alba in the field layer . A ndromeda polifolia, Vaccinium oxycoccos
and Drosera anglica appear more regularly here than in the preceding subassociation as well as Eriophorum vagi natum. In the bottom layer Sphagnum cuspidatum and S. tenellum are constants (Fransson, op . cit.).
Malmer ( 1 962) put bog lawn vegetation into the Tri
chophorum caespitosum variant of the Eriophorum
vagi-natum - Sphagnum mage/lanicum small association (con sidered progressive) and the Eriophorum vaginatum -Sphagnum tenellum small association (considered regres sive). Bog carpet vegetation is found in the Drosera ang lica variant of his Eriophorum vaginatum - Sphagnum magellanicum small association and in the Eriophorum vaginatum - Sphagnum cuspidatum small association (both considered progressive).
In the lawns of the Skattlbsberg Stormosse Eriophorum
vaginatum is the most common species and omnipresent.
Vaccinium oxycoccos and Andromeda polifolia are also
common, as is Trichophorum caespitosum which , how ever , is often lacking in the wetter parts of the community. More or less typical Vaccinium microcarpum occurs to some extent. In the bottom layer Sphagnum balticum, S. cuspidatum, S. majus, S. tenellum and, locally, S. rubel lum alternate as dominants.
Most carpets of the SkattlOsberg Stormosse belong to typical Cuspidatetum dusenietosum (Fransson). C. tenel letosum (Fransson) vegetation was not described from this locality by Sj ors ( 1 948) and Fransson assumes this community to be southern. Nevertheless, it does exist on
Table 5. Vegetation analyses from Cuspidatetum dusenietosum and C . tenelletosum (Fransson 1 972) on the Skattlosberg Stormosse, just east of the Special Area. Cover degrees ace. to the scale of Hult-Sernander-Du Rietz. Square size 1 /4 m2• Locations of squares not randomized. Vegetation intermediate between the subassociations was not analysed.
Square B 10 is from the edge of a shallow pool in the extension of the lasiocarpa soak and thus deviating.
FIELD LAYER Andromeda polifolia Calluna vulgaris
Pinus sylvestris, seedling Vaccinium oxycoccos Drosera anglica Drosera rotundifolia Rubus chamaemorus Carex limosa Eriophorum vaginatum Rhynchospora alba Scheuchzeria palustris Trichophorum caespitosum BOTTOM LAYER Sphagnum balticum Sphagnum cuspidatum Sphagnum magellanicum Sphagnum majus Sphagnum papillosum Sphagnum rubellum Sphagnum tenellum Drepanocladus fluitans Cephalozia spp . Gymnocolea inflata + Cladopodiella fluitans Cladonia squamosa Cuspidatetum dusenietosum A 2 3 4 5 6 7 8 9 1 0 2 2 2 2 2 2 3 5 5 5 3 3 2 2 3 3 4 4 4 2 Cuspidatetum tenelletosum B 1 2 3 4 5 6 7 3 2 2 2 1 1 1 2 1 3 3 2 2 2 2 2 - - 1 2 2 4 3 4 3 2 1 5 3 3 2 1 1 2 4 2 3 5 2 2 2 2 - 1 8 9 1 0 2 1 3 1 2 5 5
Production and growth dynamics of vascular bog plants 1 5 hummocks
I
I
lawns I Cusp. tenellet. Cusp. duseniet. a ;:j t: a Cl) ;:j o. ....�-2
I
a2 §
0 ... ...c Cl! O.c: 0 .... ·-... Cl! 01) r.I.l ;:..I
I a 2 a 0 ;:j ...c Vl o.o 0 .-::: ...c 0. (.) Vl .... Cl) � � II
II
Fig. 5 . Distribution of field layer plant species in different habitats in the Special Area on the Skattlosberg Stormosse.
the SkattlOsberg Stormosse in a hollow outside the Special Area, near its W border. This hollow is to a very minor extent influenced by originally minerogeneous, here very diluted water from the extension of the lasiocarpa soak, which can be seen in the presence of scattered , non flowering Carex pauciflora along its borders. Sphagnum papillosum also occurs to some extent.
Vegetation analyses from Cuspidatetum tenelletosum
and from C. dusenietosum are presented in Table 5. C. tenelletosum combines features from carpets and lawns and the limit against the lawn vegetation is sometimes dif ficult to define. In places even scattered Calluna is grow ing intermingled with typical carpet species (cf. square B9 in Table 5). For tables on the other bog communities, see Sjors ( 1 948) .
Mud-bottoms: The vegetation of the mud-bottoms was
called the Scheuchzeria - Rhynchospora alba - Carex li mosa mud-bottom association by Sjors ( 1 948). Malmer
( 1 962) described similar vegetation in his Eriophorum va
ginatum - Cladopodiella fluitans small association.
Fransson ( 1 972) treated the bog mud-bottom vegetation as a subassociation : Cuspidatetum zygogonietosum. The mud-bottom community is characterized by the absence of sphagna or other mosses . The same field layer species are found as in the carpets. Mud-bottoms are not further discussed in this treatise.
In my study I have followed Sjors's classification (with the addition of Cuspidatetum tenelletosum according to Fransson) . In so doing a direct adoption of his vegetation map of the Special Area has been possible.
For convenience, the distribution of the vascular plants in the bog communities of the Special Area is depicted in Fig. 5.
Description of the sampling areas
Sjors ( 1 948) selected a ' Special Area' in the NW part of the mire, 300x400 m large. A small esker runs along the western border of this area. Most of the Special Area is an ombrotrophic bog sloping
excent-rically eastwards , i . e . from the esker. In two places there·are springs at the edge of the esker. The south ern springs considerably influence the vegetation in the Special Area. Their water flows in a narrow soak (called the 'lasiocarpa soak' by Sj ors) out into the mire . The soak is widened there into a series of flarks and flark pools. Here the influence of the mi nerogeneous water is very weak, although quite dis tinctive.
There is also a more northern soak (the 'Scirpus ( = Trichophorum) soak ' ; Sj ors , op . cit .) with a
weak penetration of minerogeneous water . During periods of drought its flow of water ceases almost completely (Sj ors 1 948) . Between the two soaks a bog area occurs with hummocks and hollows of lawn type extended in a N-S direction. My non-de structive sampling of hummocks and lawns was concentrated to this area. Even here, a sporadic in fluence of the minerogeneous water of the lasio carpa soak can be deduced from the sparse occur rence of non-flowering Carex pauciflora (see Sj ors 1 948: map 1 4) . The sampling area thus is probably not strictly ombrotrophic in its entirety. It is as sumed, however, that the very minor influx of mine rogeneous water did not influence the results signifi cantly. The minerogeneous influence is more ob vious in the carpets and pools along the extended lasiocarpa soak . Sphagnum papillosum is common here and Menyanthes trijoliata occurs sporadically. Trichophorum caespitosum is considerably more luxuriant here than in the purely ombrotrophic areas . These carpets were therefore excluded from my investigation. Instead , two carpet hollows just outside the northern part of the eastern border of the Special Area were chosen for non-destructive sampling, one with Scheuchzeria dominance (
1 6 Ingvar Backeus
pidatetum dusenietosum Fransson) and the other with Rhynchospora dominance (C. tenelletosum Fransson) .
In the outer parts of the Scirpus soak the minero geneous influence is also quite weak . Its presence is indicated by richer occurrence of Carex pauciflora and the minerotrophic Sphagnum fa/lax. Also the Scirpus soak was excluded from the sampling .
For a more comprehensive description of the Special Area, see Sj ors ( 1 948 : 77- 1 04) .
The vegetation of the Special Area is evidently stable . Sj ors's map of this area was established in 1 944- 1 945 and my investigations were carried out in 1 980- 1 982. Within this period hardly any
detect-able changes have occurred within the pools, mud bottoms, carpets and lawns . The wooded bog areas, on the other hand , have expanded somewhat over previously treeless hummock areas . This is evident from Sj ors 's map and from several of his photos . Destructive sampling in hummocks and lawns was made east of the Special Area, SE of Puukko lam ( = Brittas hal on the new topographic map) . It also slopes eastwards . Hummocks and lawn hollows dominate and are orientated in the contour direc tion . In the southern and eastern parts carpets are also common.
Destructive sampling in carpets was made close to the non-destructive sampling .
Phenological development
Phenological data were collected in order to make it easier to choose suitable times for harvesting (cf. Persson 1 975a) . As harvesting however had to start already during the first summer, the phenological results were only gradually incorporated into the harvesting schedule .
The phenological information is , of course , also of interest as such and is important in the discussion on the dynamics of production in the different spe cies .
Methods of collecting and presenting
the phenological data
Data on phenological development were collected through general visual inspections at intervals of 5-7 days during the summer and less frequently during the spring and autumn. No exact measure ments were attempted.
The results are presented in qualitative phenologi cal spectra (cf. Dierschke 1 972) . Each studied prop erty is presented separately as I find such a presenta tion more easily interpretable than traditional dia grams. Similar diagrams were published i.a. by Perttula ( 1 949), Falinska ( 1 972) and Persson ( 1 975a).
A property often presented in phenological dia grams is flower colour. This has been omitted here, since it is not important in a production study. It could be noted , though, that all conspicuous flowers on bogs in south and central Sweden are white or mauve.
Separate diagrams for the years 1 98 1 and 1 982 are presented because ( 1 ) the 1 980 material is incom plete, (2) the period is too short for an average to be meaningful and (3) the average of years with early development and other years with late development will give the false impression of prolonged phases with an early and gradual beginning and a late and gradual termination . (I f averages are used the
aver-age yearly length in days of each phase should also be calculated .)
The kind of information that can be obtained from the spectra is evident from the key to the signs . Only a few comments are needed :
A shoot was considered to flower when the first flower bud had burst. In most species on the bog the great majority of shoots did not flower at all. This was the case in A ndromeda polifolia, Betula nana, Drosera anglica, D. rotundifolia, Eriophorum vagi natum, Rubus chamaemorus, Scheuchzeria palust ris and Vaccinium spp . It is therefore important to note that the flowering and fruiting symbols are used when more than half of the fertile shoots had reached the stage in question . The proportion be tween flowering and non-flowering shoots will be discussed in later chapters . In 1 982 no fruits of R . chamaemorus and V. myrtillus reached a mature stage, mainly due to severe night frosts , and Scheuchzeria did not flower at all .
Results and discussion
The results are presented in Fig. 6. They will be dis cussed together with other data in later chapters . Only a few points will b e taken u p here.
Although May was much colder in 1 982 than in 1 98 1 (Fig. 3), the onset of growth seemed to be very little later in 1 982 . After the cold and frosty month of June in 1 982 the cessation of growth was hardly later than in 1 98 1 . A marked effect of the summer frosts was that Andromeda polifolia, Rubus chamaemorus, Vaccinium microcarpum and V. oxycoccos formed new vegetative shoots in July to replace damaged shoots.
Few species showed any activity before the end of May. The exceptions were Eriophorum vaginatum, Scheuchzeria palustris, Betula nana and Trichopho rum caespitosum . The two first-mentioned species have a long period of active growth and their leaves
1 8 I ngvar Backeus 1 981 Andromeda polifolia 1 982 1 98 1 Betula nana 1 982 1 98 1 Call una vulgaris 1 982 1 98 1 Drosera anglica 1 982 1 98 1 Drosera rotundifolia 1 982 1 98 1 Empetrum nigrum 1 982 1 98 1 Eriophorum vaginatum 1 982 1 98 1 Rhynchospora alba 1 982 1 981 Rubus chamaemorus 1 982 1 981 Scheuchzeria palustris 1 982 1 98 1 Trichophorum caespitosum 1 982 1 98 1 Vaccinium micro'carpum 1 982 1 98 1 Vaccinium myrtillus 1 982 1 98 1 Vaccinium oxycoccos 1 982 1 98 1 Vaccinium uliginosum 1 982 1 98 1 Vaccinium vitis-idaea 1 982 M A Y
I
JUN EI
J U LYI
A U GI
S E PI
O CT •••••:!:!!j,,, ,,,,,,,,, ,,,,,,,,,, z,,,, ,, , , , , ,, , , ,, , ,, I I r r I nmP,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,r.l\�ll.}w�.t.�a >i11a 1 bilibbilol•?IJ}ii}j)\j'iii)� :o: l t � I I I I I M M M M •O!O!f l";lzzz I I I I z z ,,,,,,,,,,,,,,,,,,,,4 ... I I I I I I I I
i!MMMMMMM,,Af#Z Ill llllllll 1111 Ill 11
I I ••••••••'yAP ,, ,,,, z z z z z zz z z ,,,,.., r I I MM! ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, I I I T I I I I IWiMIMIMIMIMIMI I I I I I I I I I I -m, � .. ,,,,,,,,,.__ !!!!!\-·-·-·-·-·-· I I I I I �,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, I I l I I I I I I I I T I I I nu,,,,,,,,,,, lllllllllllllllllllllllihhiliiiiiil !f!"!w •••••••• il I I I I I I I I I --��,,,,,,,,,,,,,,,,,�
--=---=R''''""''''''""'''''""'''''""''''''"''''''"'''''''""''''''"'''''''""''''''""'''''Iu:::
I I I I I 2 3 4 5 , , , ..._ , , , , "' " " , '- ' "' � • • • • n n • • • • • • • • • • • • u • u u u n n • - · - · - • - • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • l l l l l l l l l l l l l l ,/1Production and growth dynamics of vascular bog plants 1 9 successively senesce and die throughout the growing
season. In these species it was therefore considered desirable to have an early harvest in May I June in addition to a late harvest in August .
Dwarf shrubs were harvested in the middle or later part of August. When the material had been processed it was found that an even later sampling
had been desirable in evergreen shrubs to include late-season weight increase in the leaves . In some cases also wood increment was found to take place very late in the season.
Herbs and Rhynchospora alba were sampled in late July or early August , before senescense became pronounced .
Fig . 6. Phenological spectra of field layer plants in the Special Area on the SkattlOsberg Stormosse. (1) Development of current shoots. (2) Green foliage present. (3) Foliage withering or withered (in autumn). (4) Flowering (at least one flower open in flowering shoots). (5) Fruits/seeds ripening or ripe (up to shedding or overripeness). A species is con
sidered to be in a certain phase when more than half of the individuals (in 4 and 5: More than half of the fertile individu als) are in this phase. The horizontal bars denote initial and final stages (i.e. when 1 0-50 OJo of the individuals are
within the phase in question) . Data collecting ended on Sept. 24, 1 982.
Production and dynamics of individual species
Methods
The yearly aboveground production of each field layer species (except Pin us sylvestris, Carex limosa and C. pauci
j
lora)
was determined . Other studies on certain species , including repeated sampling over the year and direct measurement of length growth will be presented under the species in question. For the assessment of the yearly aboveground produc tion (called 'the main sampling' below), the method put forward by T. Traczyk ( 1 967a and b) was used. The procedure is similar to the method commonly used for the estimation of tree and shrub layer pro duction . It implies two steps: the determination of the density (D) of each species and of the average current year production (G) of an individual of each species . The total production (P) of all(
n)
species is thenn
p = � Gi Di i = 1
Newbould ( 1 967) suggested the same method under the name the individual plant method . The same or similar methods for estimating biomass and produc tion have been used in Poland by Aulak ( 1 970) , Plewczynska ( 1 970), Moszynska ( 1 970, 1 973), H .Traczyk ( 1 97 1 ), Puszkar et al . ( 1 972), T. Traczyk et al . ( 1 973) and T. & H. Traczyk ( 1 977), in Czecho slovakia by Brechtl & Kubicek (1 968), Kubicek & Brechtl ( 1 970) and Kubicek & Jurko (1975), in Ger many by Eber ( 1 97 1 ), in the Sovietic Far East by An dreev et al . ( 1 972) , on Greenland by Lewis & Calla ghan ( 1 97 1 ) and in Canada by Reader & Stew art ( 1 972) and Stewart & Reader (1972). A theoretical interest in the correlation between biomass and den sity has also arisen with the establishment of the '3/2 power law' (White & Harper 1 970 and several later authors) .
Density: Plants were counted in quadrats , the size of which varied, depending on the species , from -l 0x 1 0 cm to 200x200 cm (see Table 6) . The analyses
A cta Phytogeogr. Suec. 74
in hummock and lawn vegetation were made along straight lines , running more or less perpendicular to the contour lines, in the Special Area north of the extension of the lasiocarpa soak . There were five such lines traversing hummocks and lawns and in addition a few shorter lines in lawns. Points falling on the limit between a hummock and a lawn were disregarded . The quadrats of different sizes were nested and placed at every second metre. In the larg est squares (200x200 cm) the vegetation was not everywhere homogeneous . The analysed area was in such cases extended perpendicularly to the baseline until approximately 4 m2 of hummock or lawn ve getation were included . The intention was to keep the ends of the lines at fixed points , but in order to avoid errors due to trampling, the lines had to be moved slightly before the harvest in 1 98 1 .
Carpet analyses were made in an area immediate ly to the east of the Special Area. Cuspidate turn du senietosum was analysed in a carpet with more or less pure stands of Scheuchzeria. Thirteen quadrats of 50x50 cm were analysed . In C. tenelletosum 30 squares were analysed, their size being 1 0x 1 0 cm for Rhynchospora alba and 25x25 cm for other species . It cannot be claimed that the carpet analyses give a true picture of the average density of the species in these communities . To obtain this an extensive and time-consuming sampling in the small carpets scat tered over the bog would have been necessary. The obj ective has been restricted to give examples of the production in the two carpet communities .
Individual weight: Specimens of each species were sampled at regular intervals along straight lines on hummocks and in lawns east of the Special Area (eastern part of Slaktmossen). The number of col lected specimens varied between species from 25 to
1 00 according to degree of variation in plant size . The plants were taken to a refrigerator on the same day and moved to a freezer not later than the next day. They were later fractioned, dried at + 85 ° C for 48 hours (dwarf shrubs) or 24 hours (others) . After
Production and growth dynamics of vascular bog plants 2 1 Table 6 . Quadrat sizes i n cm i n the density measurements.
1 980 1 98 1 1 982
hummocks lawns hummocks lawns hummocks lawns
Andromeda polifolia 25 x 25 25 x 25 25 x 25 25 x 25 25 x 25 25 x 25 Betula nana 200 x 200 200 x 200 200 x 200 200 x 200 200 x 200 200 x 200 Calluna vulgaris 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 Carex pauciflora 10 x 10 1 0 X 10 Drosera rotundifolia 50 X 50 50 X 50 50 x 50 50 x 50 50 x 50 50 x 50 Empetrum nigrum 25 X 25 25 x 25 25 x 25 25 x 25 25 x 25 25 x 25 Eriophorum vaginatum 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 1 0 x 1 0 lO x 1 0 1 0 x 1 0 Ditto, flowering 200 x 200 200 x 200 200 x 200 200 x 200 200 X 200 200 x 200 Rubus chamaemorus 25 x 25 50 x 50 25 x 25 50 x 50 25 x 25 50 x 50
Ditto, with fruits 200 x 200 200 X 200 200 x 200 200 x 200 200 x 200 200 x 200
Trichophorum caespitosum 1 0 x 1 0 1 0 x 1 0 20 x 1 0 20 x 1 0 20 x 1 0 20 x 1 0
Vaccinium microcarpum 1 0 x 1 0 1 0 x 1 0 20 x 1 0 20 x 1 0 20 x 1 0 20 x 1 0
V. oxycoccos 25 x 25 25 x 25 20 x 10 20 x 1 0 20 x 1 0 20 x 1 0
V . uliginosum 25 x 25 25 x 25 50 x 50 50 x 50 50 x 50 50 x 50
Carpets: Cuspidatetum dusenietosum: All species 50 x 50 cm . C . tenelletosum : Rhynchospora alba 10 x 10 cm . All other species
25 X 25 .
cooling in an exsiccator the plants were weighed in dividually with an accuracy of 0. 1 mg.
In carpets only species not occurring in lawns or hummocks were sampled. Therefore a full picture of the field layer production of the carpets has not been obtained.
What is an individual? Plants often differ from ani mals in not havin� distinct individuals (cf. Williams
1 964) . Seedlings of other species than Eriophorum vaginatum and Drosera rotundifolia (and probably D. anglica) seem to be rare on bogs . Most of the pro pagation must therefore be considered to be vegeta tive and genetic individuals (genets) are usually im possible to distinguish . For practical reasons the ramets obtained when the plants are cut at ground level or at the first adventitious root are here called 'individuals ' . Strictly, what has been studied is not populations of plants but metapopulations from an unknown number of genets (White 1 979) . In order to avoid genetic bias it was considered important to spread the sampling over a rather wide area.
A few species form mats . In such cases no indi viduals of any kind could be distinguished. Instead , current shoots or other units ('plant units' sensu Williams 1 964) were considered . It has not been possible-nor desired-to treat all species in the same way during the sampling and procedures that were followed will be presented under each species . Some changes in the procedures were made from year to year in order to improve the methods and
make the sampling easier . They affect biomass de terminations only. The production measurements, which have been considered more essential in this in vestigation, were not affected.
Throughout this work attached dead is main tained as a separate category and not included in biomass . 'C' means 'current year ' , 'C + 1 ' means
'previous year' etc.
Andromeda polifolia
Material and methodsAt the main sampling current shoots of Andromeda were counted in 1 980 and 1 98 1 but individuals in 1 982. Individuals were harvested on hummocks and in lawns, cut at the moss surface irt 1 98 1 , at the first adventitious root in 1 980 and 1 982, but always so that at least the whole current shoot was included. On the collected individuals leaves were frac tioned into generations (C, C + 1 , C + 2, C + 3).
Current and older stems were separated in 1 980; in 1 98 1 and 1 982 also C + 1 and C + 2 stems were kept as separate fractions . Attached dead also constitut ed a separate fraction.
Each fraction was weighed and the number of leaves of each generation and of current shoots were counted. The lengths of C ( 1 98 1 - 1 982 only) and C + 1 ( 1 982 only) stems were measured . In 1 982 at tached dead in current shoots was weighed separate-Acta Phytogeogr. Suec. 74
22 Ingvar Backeus
ly. Shoot generations are easily separated by means of the remaining bud scales .
The production per individual was calculated as the sum of the following:
(a) C stem weight
(b) weight increase in C + 1 stems: (C + 1 stem weight · length1
-- C stem weight · length-1) • - C + 1 stem length
As estimates of shoot length were not available for all years, the existing estimates from 1 98 1 and 1 982 were used for calculations over all three years. Esti mates for the weight increase in three-year-old stems and older are not available .
(c) weight of C leaves at the end of the season: weight of one C + 1 leaf · number of C leaves
There was a considerable weight increase in C leaves after harvest . As is shown below, in autumn they attained the same weight as the C + 1 leaves. The C + 1 leaf weight times the number of C leaves therefore gives an estimate of current leaf produc tion during the whole growing season.
The figures from the 1 980 and 1 98 1 countings were divided by the average number of current shoots per individual to obtain the number of indi viduals per unit area.
Besides the main sampling, 25 individuals from hummocks were harvested in 1 98 1 twice a month with the purpose of studying weight changes in leaves and stems. This material was transported to a freezer within three or four hours . Later the plants were fractioned and then immediately dried at 8 5 ° C for 36 hours . The plants were fractioned into C leaves , C + I leaves, C stems and C + 1 stems. Older leaves and stems were too few to give meaningful es timates . Stems and individual leaves were weighed and measured after drying with an accuracy of 0.01 mg and 0. 1 mm respectively. To avoid errors due to chance fluctuations of leaf size in the samples, the weight to length ratio was used in the calculations (cf. Flower-Ellis 1 975). Leaf weight is linearly re lated to length according to Flower-Ellis ( 1 973).
In 1 982 the length growth in current stems and leaves was followed by direct measurement with ver nier calipers . This experiment was not very success ful because of the summer frosts which substantially
damaged the shoots and therefore no results are pre sented .
Results and discussion
The length growth of current shoots in 1 98 1 and 1 982 commenced in the last week of May and ended in late June (Fig. 6) . Lindholm ( 1 982) reports a longer period, from early May to late June . No bark formation takes place in the upper parts of the shoots during their first summer (Segerstedt 1 894) . The limit between the two parts of the shoot were easily visible because of a sharp change in colour. Empetrum nigrum shows a similar situation, see further under that species .
Weight increase in current stems continued dur ing the whole growing season (Fig . 8) . No weight in crease was found in the C + 1 stems in the earlier part of the growing season but a remarkable increase in the weight of these stems occurred in August and September, thus indicating that this was the time when wood increment took place.
The Andromeda shoots on the bog were usually short . 75 OJo of the shoots were shorter than 1 3 or 17 mm in lawns and on hummocks respectively. The plants therefore ran the risk of being overgrown by Sphagnum . It was not uncommon to see Andro meda shoots with only the upper parts of a couple of leaves showing above the moss surface . Some times, however, considerably longer shoots (run ners) are formed (cf. Warming 1 908) . They origi nate from a bud on a buried stem and grow oblique ly upwards until they reach the surface, then gradu ally bend to a vertical position and form green leaves . The tallest measured runner from a lawn was 8 cm and from a hummock 1 5 cm . Subterranean runners that grew for more than one year before reaching the surface and forming leaves were very seldom seen (cf. Keso 1 908). With these runners the plant is both propagated vegetatively and escapes the rise of the bog surface. The shoots of the next summer will again be of normal length . According to Keso ( 1 908) such shoots will live (in Hame, Fin land) for 5 to 1 1 years before they die. Serebryakov ( 1 962) reports (on Sphagnum bogs in the Moscow region) 6 to 8 years . Rosswall et al . ( 1 975) on the other hand, report up to 30-year-old shoots in some microhabitats on the Stordalen mire in north Swe den. Due to the repeated forming of runners the
to-Production and growth dynamics of vascular bog plants 23 r-(I) Ol
;o.4
(I) ce· ;?; '0 � s:::: 2. 0.3 ... CD ::s cc ;. 3�
0.2 3 3 � 0.1 J J A s 0 Time (month)Fig . 7. A ndromeda polifolia. Changes in mean weight
(± 1 S.E.) per unit length of C ( • ) and C + 1 <•> leaves
on hummocks during 1 98 1 .
tal length o f a subterranean stem can be consider able. Keso ( 1 908) reports up to 2 1 4 cm, Metsavainio ( 1 93 1 ) up to 1 . 5 m. The old stems lie horizontally due to the compaction of the peat .
Leaves overwinter. From the material of the re peated leaf sampling in 1 98 1 (Table 7) it can be seen that there was no significant leaf mortality until July in the second summer (similar results in Malmer & Nihlgard 1 980) . Only few leaves lived during a third season. Estimates from the Stordalen mire by Flower-Ellis ( 1 980a) based on a considerably larger material gave a mortality of not more than 5 OJo dur ing the interval between their appearance and that of the next year's C leaves . About 60 % survived their second winter at Stordalen.
The current leaves of 1 982 reached their full Table 7. Andromeda polifolia. Number of leaves per shoot in
1 98 1 . Hummocks only. Means of 25 shoots.
Average
Generation Period number OJo survival
c 1 5 . VII-22.X 3 . 8 1 00
C + 1 1 6.VI- 1 5 .VI I 3 . 5 90
1 5.VII-22.X gradually de- 90 to 50
creasing from 3 . 5 to 1 .7 C + 2 1 6.VI-22.X 0.4 1 0 (/) ... (I) 3 0.4 :E (I) <0' ::r ... '0 � § 0.3 ;:;· CD ::s cc ... ::r 3 cc 0.2 3 3� 0.1 J A s 0 Time (month)
Fig. 8. A ndromeda polifolia. Changes in mean weight ( ± 1 S.E.) per unit length of C ( • ) and C + 1 <•> stems on hummocks during 1 98 1 .
length in the first half of June but there was a con tinuous increase in dry weight during the whole growing season (Fig. 7) . In autumn the C leaves had attained the same weight as the C + 1 leaves had at the beginning of the season. Similar results were ob tained by Flower-Ellis ( 1 975). The C + I leaves lost weight notably in June and July. This indicates a dry matter allocation to other tissue, presumably cur rent shoots (cf. Johansson 1 974) . According to Jo hansson translocation to the roots , on the other hand, mainly takes place before the current leaves are formed . The surviving leaves then gradually re sumed their former weight. Flower-Ellis ( 1 975) found a marked decrease in dry weight in September in leaves of the second season (C + 1 ) . My curve deviates in this respect .
Leaves are xeromorphic but the degree of xero morphy is variable and usually more accentuated in hollows . Simonis ( 1 948) found that pot-cultured specimens were more xeromorphic in wet cultures than in dry. It is notable that some shoots have broader and much less xeromorphic leaves . This is also so in leaves attacked by Exobasidium karstenii and E. sundstroemii (N annfeldt 1 98 1 ; the latter spe cies not seen on the bog).
