Assessing conservation values of forest stands based on specialised lichens and birds

Assessing conservation values of forest stands based on specialised lichens and birds

Biological Conservation 95 (2000) 343±351 www.elsevier.com/locate/biocon Assessing conservation values of forest stands based on specialised lichens...

184KB Sizes 0 Downloads 11 Views

Biological Conservation 95 (2000) 343±351

www.elsevier.com/locate/biocon

Assessing conservation values of forest stands based on specialised lichens and birds Helen Uliczka *, Per Angelstam Swedish University of Agricultural Sciences, Department of Conservation Biology, GrimsoÈ Wildlife Research Station, S-730 91 Riddarhyttan, Sweden Received 11 June 1998; accepted 24 December 1999

Abstract We compared the presence of sets of epiphytic macrolichens and resident birds as indicators of old and deciduous trees in each of three forest types in south central Sweden: mixed/deciduous (MIX), old coniferous (OLD) and managed (MAN). Lichen diversity and abundance was greatest in MIX but the mean number of species per stand was highest in OLD. The number of species per stand increased with the proportion of old trees in the stand. The diversity and mean number of resident bird species was highest in MIX while OLD and MAN did not di€er. The number of bird species and the proportion of deciduous trees per stand were positively correlated. Thus the conservation of stands of old trees that also contain deciduous (non-commercial) trees will favour both lichens and birds and probably a wide range of other plants and animals. The study shows that the present Swedish use of only the ¯oristic component for assessment of conservation values is not sucient. Adding area-demanding specialists such as certain resident birds is a step forward towards a reliable indicator system. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Indicator species; Boreal forest; Forest conservation; Resident birds; Epiphytic lichens

1. Introduction Loss of natural forest habitats is a common cause of the recent decline in biological diversity in many countries (Heywood, 1995; Stanners and Bourdeau, 1995; Dudley et al., 1996). For a long time forest management focused on the production of wood, with few or no consideration given to other values in forest ecosystems. Recently, however, the conservation and enhancement of forest biodiversity have started to be included in sustainable forest ecosystem management (Noss, 1993; Christiansen et al., 1996; DeGraaf and Miller, 1996; Angelstam and Pettersson, 1997). To be e€ective, this requires that the status of biological diversity can be assessed. The results of this assessment should be available and understandable by all participants in the practical management and planning of the forest landscapes (Noss, 1990; Angelstam, 1997, 1998; Hunter, 1999). One way to monitor and identify the current status of a forest, is to select specialised species as indicators of * Corresponding author. Fax: +46-581-697310. E-mail address: [email protected] (H. Uliczka), [email protected] (P. Angelstam)

particular forest conditions or properties. This approach was originally introduced to classify the productivity of a forest site by using the composition of vascular plants (Cajander, 1926, 1949; Hill et al., 1975; Pregitzer and Barnes, 1982; HaÈgglund and Lundmark, 1984; La Roi et al., 1988). Since then it has gradually been broadened to become a tool for environmental monitoring (Spellerberg, 1991). Some species of lichens need trees of particular ages (Uliczka and Angelstam, 1999) and are sensitive to changes in the local climate (Nash and Olafsen, 1995; Gauslaa and Solhaug, 1996) and in local forest continuity (Tibell, 1992; Rose, 1992, 1996; Kuusinen, 1996a). The presence of these species could thus indicate speci®c habitat conditions and the likely occurrence of other, less sensitive, species. Today, the Swedish National Board of Forestry (Anon., 1994) uses lichens, vascular plants, bryophytes and fungi in Sweden as indicators of forests with high conservation values. By contrast, there has been little use of animals as indicators of forest qualities in practical nature conservation although there is good evidence for using birds as indicators. For example Jansson (1998) found that when the lesser spotted woodpecker (Dendrocopos

0006-3207/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0006-3207(00)00022-7

344

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

minor) was present in a patch of deciduous forest, the probability of the presence of three other selected bird species, long-tailed tit (Aegithalos caudatus), marsh tit (Parus palustris) and blue tit (P. caeruleus) was very high. The presence of the long-tailed tit was the second best predictor of the presence of the other species. Similarly, Martikainen et al. (1998) showed that the presence of white-backed woodpeckers (Dendrocopos leucotos) was closely linked to the presence of rare beetles. For a whole forest ecosystem with di€erent disturbance regimes and successional stages, it is likely that several indicator species with di€erent ecologies are necessary since many properties covering multiple scales must be assessed (Noss, 1990; Angelstam et al., 1993; Angelstam, 1997; Nilsson, 1997). In Sweden, a very long history of forest management has proved to be detrimental to many of the native forest-dwelling plant and animal species (Hansson, 1992, 1997). According to recent calculations about 1040 forest species are threatened or have already disappeared (Berg et al., 1994), i.e. 70% of all species threatened or lost in Sweden. Analyses of the habitat requirement of those species show that old forest stands, deciduous trees and dead wood are forest properties commonly required. This paper is an evaluation of selected sets of epiphytic lichens as indicators of deciduous trees and old trees in a typical managed landscape. We compare the lichen results with the occurrence of resident bird species in the same habitats to illustrate that di€erent taxa may complement each other in an indicator species system, by having di€erent habitat requirements. The objective of the study was to see whether there are some simple criteria for selecting forest stands to conserve birds as well as lichens.

than 120 years make up a further 3.7%. Lakes and rivers cover 8% of the area, and mires 18%. Agricultural land takes up about 3%.

2. Study area

We decided to record 33 epiphytic macrolichens (Table 2) that have been noted as indicators of forest biotopes of conservation value in this region (e.g. Nitare and NoreÂn, 1992; Hermansson et al., 1993; Anon., 1994; Uliczka and Angelstam, 1999). Lichen species known to

The study area covered 10,600 ha within the GrimsoÈ Wildlife Research Area, which is situated in south central Sweden (59 400 N, 15 250 E). It is a part of ``Bergslagen'' which used to be a mining district and has the longest history of exploitation in the Swedish boreal forest. The area was intensively logged for several centuries for the iron industry and later for commercial forestry (Wieslander, 1936; Attman, 1986). The forests are dominated by Norway spruce (Picea abies) and/or Scots pine (Pinus sylvestris) (BjoÈrkhem and Lundmark, 1975). The deciduous forest component is mostly birch (Betula sp.) and aspen (Populus tremula). The forest (72% of the total area) consists of 84% conifer-dominated, 11% mixed deciduous/coniferous and 5% deciduous-dominated forests. Deciduous and mixed deciduous/coniferous forests older than 80 years amount to 1.6% of the forest area and pine forests older

3. Methods 3.1. Selection of forest types and stands We surveyed resident birds and epiphytic lichens in 30 stands in each of three di€erent forest types: Mixed (MIX): Unmanaged, mixed deciduous-coniferous (mainly birch, aspen and spruce) forest with a patchy distribution in the area (Enoksson et al., 1995). The selected patches were mainly on former farmland and wooded meadows, where most of this forest type is found. Old forest (OLD): The only remaining unmanaged old growth forest consists of small patches of pine forest on locally remote hilltops inaccessible for forestry and on islands in large peat bogs. Managed (MAN): Managed, formerly clear-cut, pinedominated forest on glacial till and sediments. All pine, spruce, birch and aspen trees were counted in these 90 sample stands (radius 10 m) which were separated by a minimum distance of 200 m (see Table 1). Tree age was determined for 5±10 trees per stand by taking an increment core at 30 cm above ground level. Since managed stands are even-aged, these stands were selected so as to cover the existing range of tree age classes. Breast height diameter was measured on all trees that were sampled for lichens. Based on these measurements the rest of the sampled trees in the stands were classi®ed into 20-year interval age-classes. 3.2. Lichens

Table 1 Mean number of trees per tree species and survey stand (diameter 20 m)standard error, in the forest types MIX, OLD and MANa MIX OLD MAN n Tree species trees/point (%) trees/point (%) trees/point (%) Pine Spruce Birch Aspen

3.91.28 (7)a 24.14.0 (45)a 17.74.8 (33)a 8.11.8 (15)a

28.15.2 (64)b 12.22.5 (28)b 3.62.0 (8)b 00b

42.58.5 (76)b 1481 13.04.4 (23)b 2233 0.20.14 (0)b 646 00b 243

a Numbers within parentheses are the percentages of the tree species in the forest type. Di€erent letters (a and b) denotes a signi®cant di€erence (P<0.05) in analysis of variance.

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

be very abundant (Moberg and HolmaÊsen, 1982; Kuusinen, 1996b), such as Hypogymnia physodes and Platismatia glauca, were excluded from the set. The species in bold in Table 2 were named ``signal species'' by the National Board of Forestry in a nation-wide inventory of forest stands of importance for the maintenance of biodiversity (Anon., 1994). Nomenclature follows Santesson (1993). We sampled epiphytic lichens on the stem and on the branches of a maximum of 20 trees per sample stand, from the ground level up to 2-m height. Individual thalli of fruticose lichens were counted and foliose lichens were estimated as per cent cover of the examined tree area. Following McCune (1990), who found a positive relation between the thalli size and % coverage, we grouped the abundance of fruticose and foliose lichens as follows: 1=1±5 thalli or cover of 0.1±5%; 2=6±10 thalli or cover of 6±10%; 3=11±15 thalli or cover of 11± 15% etc. Table 2 List of lichen species included in the surveya all of which occur regionally)

345

Trees that were older than the lowest permitted age for harvesting according to productivity class of the site (Anon., 1985; Anon., 1993) were classi®ed as old in our correlations with lichens. The range of hardwood tree ages was 40±80 years, of spruce 80±120 years and of pine 100±180+ years (Uliczka and Angelstam, 1999). 3.3. Birds Similarly, we listed 22 resident bird species that we might expect to see in our stands (Table 3). These ranged from small to medium size, and from generalists to specialists in their habitat preferences (see JaÈrvinen et al., 1977; JaÈrvinen and VaÈisaÈnen, 1977; VaÈisaÈnen et al., 1986; Angelstam and Mikusinski, 1994; Enoksson et al., 1995). The inventory was made by point-counts for 6 min in 30 stands in each forest type, in April 1993. All pointcounts were made in two ways: by recording just those species that could be seen or heard within a circle of 25m radius, and by recording all species that could be heard from the survey-point.

Forest type b

Lichen species

MIX OLD MAN All (n=30) (n=30) (n=30) (n=90)

Usnea ®lipendula Bryoria capillaris Usnea hirta Bryoria fuscescens Imshaugia aluerites Bryoria furcellata Hypogymnia farinacea Usnea sub¯oridana Evernia prunastri Xantoria parietina Peltigera praetextata Ramalina farinacea Leptogium saturninum Alectoria sarmentosa Anaptychia ciliaris Bryoria fremontii Collema subnigrescens Total No. of species: No. of signal species: Mean No. of species/stand Mean No. of signal species/stand Diversity index (Simpson's)

24 20 12 18 2 4 5 11 11 8 7 7 4 1 1 15 4 4.4b 0.5b 9.54

23 24 27 17 28 21 21 8 1

27 18 19 14 7 9 8 2

3 1 11 4 5.8a 1.5a 7.84

8 2 3.6b 0.5b 5.98

74 62 58 49 37 34 34 21 12 8 7 7 4 3 1 1 1 17 6

a The following 16 species that were looked for were not found in any of the stands: Bryoria nadvornikiana, Collema furfuraceum, Evernia divaricata, Evernia mesomorpha, Hypogymnia vittata, Letharia vulpina, Lobaria pulmonaria, Lobaria scrobiculata, Nephroma bellum, Nephroma laevigatum, Nephroma resupinatum, Parmeliella triptophylla, Plasmatia norvegica, Peltigera canina, Ramalina thrausta, Vulpicida juniperinus. b Species are ranked according to the number of stands (n=90) in which they occurred in the forest types MIX, OLD and MAN. Signal species are indicated in bold. Di€erent letters (a and b) denotes a signi®cant di€erence (P<0.001) in analysis of variance.

Table 3 List of sedentary bird species included in the surveya (all of which occur regionally) and the number of stands in which they were recorded in the forest types MIX, OLD and MAN. Species are ranked by frequency on the 90 survey pointsb Forest type Bird species

MIX (n=30)

OLD (n=30)

MAN (n=30)

All (n=90)

Parus montanus Parus ater Parus major Regulus regulus Dendrocopos major Dryocopus martius Loxia spp. Carduelis spinus Parus cristatus Garrulus glandarius Sitta europaea Picus viridis Parus caeruleus Carduelis choris Pyrrhula pyrrhula Aegithalos caudatus Certhia familiaris Parus palustris Total No. of species/stand Mean No. of species/stand Diversity index (Simpson's)

7 16 16 9 5 3

16 1 4 2 4 5 4 2 9 5

4 7 1 8 8 11 14 6 2 3

1 1 3

1

27 24 21 19 19 19 18 14 14 12 10 10 9 5 4 3 2 2 18

6 3 4 10 10 8 3 1 3 2 16 3.4a 10.97

2 14 1.9b 7.58

11 2.2b 7.53

a The following 4 species that were looked for were not found in any of the stands: Dendrocopos minor, Dendrocopos leucotos, Picoides tridactylus, Picus canus. b Di€erent letters (a and b) denote a signi®cant di€erence (P<0.001) in analysis of variance.

346

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

4. Results 4.1. Stand composition The tree species composition showed that aspen was absent in OLD and MAN. All four tree species had a signi®cantly di€erent occurrence in MIX compared to OLD and MAN, which were not signi®cantly di€erent from each other (Table 1). The mean 20-year age-class of the sampled trees was 40±60 years in MIX, 80±100 years in OLD and 60±80 years in MAN and these di€erences were signi®cant between all pairs of forest types (ANOVA: F2, 1530=170.5; SMIX-OLD=157.77, SMIX-MAN=15.76, SOLD-MAN=87.9; P=0.0001). 4.2. Lichens The forest types MIX, OLD and MAN contained respectively 440, 505 and 588 trees that were examined for lichens. From the set of 33 species that were searched for, we found a total of 17 (Table 2). Of these, 15 species occurred in MIX, 11 species in OLD and 8 species in MAN. Similarly, Simpson's diversity index was greatest in MIX, followed by OLD and least in MAN. In contrast the mean number of lichen species, and of signal lichen species, per stand was highest in OLD. OLD di€ered signi®cantly from the other forest types but they did not di€er from each other (Table 2). The eight lichen species found in MAN were also present in the other forest types. MIX had six species that were not found in OLD or MAN. These species were exclusively con®ned to aspen in this study. Two signal species, Alectoria sarmentosa and Bryoria fremontii, were only found on trees older than 120 years and were speci®c to OLD (Table 2) The total mean of the index of abundance of lichen thalli per tree in all age classes (see Fig. 1a±c) was highest in MIX (3.793.23), decreased to OLD (2.061.62) and was lowest in MAN (0.860.54). The sum of the means for all species of the index for each age-class increased the most in MIX, from 0 to a maximum of 7.74 in age-class 81±100 years (Fig. 1). The increase was lowest in MAN, from 0.078 to 1.614 in age-class 61±80 years. This di€erence was statistically signi®cant in the pairwise comparison between MIX and MAN (z=1.922, P<0.05, Wilcoxon signed-rank test) and between OLD and MAN (z=1.992, P<0.05), but not between OLD and MIX (z=1.214, P=0.22). Hence, not only the number of species but also the amount of lichens was highest in middle-aged mixed forests. Six signal species were present; four in MIX and OLD, respectively, and two in MAN (Table 2). These two species (Bryoria furcellata and Hypogymnia farinacea) were present in more than twice as many stands

Fig. 1. Summed mean of the index of lichen abundance per species and tree divided into signal species and other species in the age-classes of the forest types (a) MIX, (b) OLD and (c) MAN. Figures above stacks indicate the number of trees in each age-class/number of lichen species within the stack.

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

in OLD as in MAN and MIX (Table 2). In Table 2, where the lichens are ranked according to their frequency of occurrence, they are ranked 6 and 7, while the other four species are all found amongst the last ®ve on the list. There was a signi®cant positive relationship between the number of signal species and the number of other species (r2=0.138, d.f.=88, P<0.001) (Fig. 2). The number of lichen species, and signal species, per stand increased in relation to the percentage of old trees (of those examined) in the stand (r2= 0.22 and 0.12, d.f.=88, P<0.001). However, there was no relation between either the number of all lichen species (r2=0.007) or the number of signal species (r2=0.037) and the percentage of deciduous trees in the stand. 4.3. Birds We recorded a total of 18 resident bird species out of the 22 in the chosen set; 16 species in MIX, 14 in OLD and 11 in MAN (Table 3). The mean number of species in MIX was signi®cantly greater than that in OLD or MAN. Simpson's index showed a higher diversity in MIX than in OLD and MAN (Table 3). The more conservative data, including only those species seen or heard within the 25 m radius circle, gave the same results with a mean species number of 1.41.3 for MIX, 0.61.1 for OLD and 0.40.7 species for MAN.

347

Four species, specialising on deciduous trees, were exclusive to MIX, namely nuthatch (Sitta europaea), green woodpecker (Picus viridis), long-tailed tit and marsh tit; only tree creeper, (Certhia familiaris) was exclusive to OLD. All species found in MAN were present in OLD and, except for crossbills (Loxia sp.) which are conifer seed specialists, in MIX (Table 3). There was a positive correlation between the number of species and the percentage of deciduous trees (r2=0.179, d.f.=88, P<0.001). The four species in our list that were not encountered require extensive areas of old-growth forest, which are absent in our study area (VaÈisaÈnen et al., 1986; Angelstam and Mikusinski, 1994). 5. Discussion Our study shows that the bird and lichen species studied depended on di€erent forest features (Table 4). While the mean number of bird species was associated with the di€erences in the proportion of deciduous trees, the mean number of lichen species followed the pattern of mean tree age. However, the abundance and total number of lichen species were also related to tree species diversity. Hence these taxonomic groups could complement each other in an indicator system for assessment of forest conservation values. 5.1. Lichens The species richness of lichens was negatively a€ected by the lack of old trees in both mixed and coniferous forest. Because several of the lichen species we found are classi®ed as sensitive to air-borne pollution, we reject an alternative explanation that the low species richness was due to air pollution: for example Bryoria sp., Usnea sp. and Peltigera praetextata, all with a pollution sensitivity index of 6±7 on a 0 to 9 scale (Hultengren et al., 1991), were common. Only in stands that have been inaccessible for harvesting did we ®nd those lichen species suggested as indicators of forest of high conservation value e.g. Alectoria sarmentosa and Bryoria fremontii, or red-listed Table 4 Relative di€erences between the occurrences of the species groups and the structural di€erences between the habitats

Fig. 2. Relationship between the number of signal lichen species per survey stand and the number of other lichen species in the same stand (the target lichen set has been divided into signal species and other species) in all forest types. A central circle and rays indicate the number of stands (n=90). a=indicates presence of one or more thalli of the signal species Alectoria sarmentosa, Bryoria fremontii, Collema subnigrescens or Leptogium saturninum.

Stand data

MIX

OLD

MAN

Mean No. of lichen species Mean No. of signal species Mean No. of bird species Prop. of deciduous trees Mean tree age Habitat distribution

Low Low High High Low Fragmented

High High Low Low High Fragmented

Low Low Low Low Medium Continuous

348

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

species, e.g. Collema subnigrescens (Anon., 1994; HallingbaÈck, 1994). This suggests that lichens can persist for a long time if there is a continuous presence of old trees. However, the very low abundance of these species could be a sign of an ongoing decline within these stands. This, in turn, means that they may not function as ``source'' areas for the dispersal of the lichens into the surrounding managed forest. An old mixed forest usually contains sensitive lichen species, such as those with cyanobacterial photobionts in the Lobarion-community (KarstroÈm, 1992, 1993; Gauslaa, 1994; Kuusinen, 1995, 1996a). However, in our study we found mostly green-algal species and not the expected increase in species with an increasing proportion of deciduous trees in the stand. We argue that the main reason was the very low abundance of old deciduous trees in our study area. The ``signal'' indicator species con®ned to deciduous trees were very few compared to those growing on conifers. If the cyanobacterial lichens still exist in the region, they have evidently not been able to disperse into the younger mixed deciduous/coniferous stands. The coniferous trees in MIX were <120 years of age, which could explain why the slowly colonising signal species on conifers were not present there. However, the lichen abundance in younger age-classes was higher in mixed and old forest than in the managed forest. This may be due to a higher amount of dispersing propagules from lichens growing on older trees within these stands; i.e. more lichens generate more lichens. The total abundance of all species declined in the older age-classes in all forest types. This was especially conspicuous in the oldest age-class of MIX. In this ageclass, and also in the older age-classes of MAN and OLD, the predominant tree species was pine. Pine has a lower abundance of several lichen species than the other tree species (Uliczka and Angelstam, 1999), which could be an explanation for this decline. 5.2. Birds Our results suggest that the proportion of deciduous trees in a forest stand is more determining for the presence of the selected set of bird species than the age of the trees. There was no signi®cant di€erence between the two coniferous forest types in the mean number of species per survey point, and the diversity was almost the same. Nevertheless, more species were found in OLD than in MAN. Similarly, BostroÈm (1988) found that of nine resident bird species found in natural north boreal pine forest, only two species were found in managed forest. Also in southern Sweden, Nilsson (1979) found more species in natural than in managed coniferous forest. In young forest plantations, the number of species was less than one third of that in the natural forest. According to Nilsson (1979), tree creeper, which

was exclusive for OLD, had higher densities in mixed natural forest than in managed coniferous forest. Virkkala et al. (1994) found tree creeper together with the crested tit (Parus cristatus) to be more abundant in the old-growth nature reserves than in other forest types. It also increased in number with the proportion of oldgrowth forest in the reserve. We argue, that the main reason for the absence of a signi®cant di€erence in species diversity between OLD and MAN in our study was that the patches of old coniferous forest are very small, <20 ha, and also make up <4% of the landscape. The absence of birds characteristic of old-growth forest in OLD, is most likely a consequence of the combination of the small size of the stands and their isolation from other old forest stands (AndreÂn, 1994). Species associated with old-growth habitats (Virkkala et al., 1994) such as three-toed woodpecker (Picoides tridactylus) and grey-headed woodpecker (Picus canus) have disappeared as regular breeders from this region, and lesser spotted woodpecker is very rare (Anon., 1986; Angelstam and Mikusinski, 1994). Hence, the absence of a di€erence between OLD and MAN was not unexpected. In Finland the woodpecker guild densities were higher in old-growth nature reserves than in managed forests, while the tit guild showed no preference for the reserves (Virkkala et al., 1994). Only the small-sized bird species demanding deciduous trees were present in our study area. They have smaller area requirements and the proportion of mixed forest in this area is still above a threshold value for their existence (Jansson and Angelstam, 1999). 5.3. Indicator species or habitat criteria? The use of vertebrate indicator species has been criticised by Landres et al. (1988). It has in practical use often been characterised by the ``absence of precise de®nitions and procedures...''. The value of indicators of forest continuity has also been under dispute as being anecdotal or based on intuitional knowledge (Ohlson et al., 1997). Peterken (1993) mentions that the popular interest in ``Ancient Woodland Indicators'' in Britain has resulted in an example of an indicator list where the presence of species was not backed up by evidence of woodland history. According to the nested subset hypothesis (Patterson, 1987) ``a species comprising proper depauperate insular biota are a subset of those in richer biota'', Cutler (1990) argued that this hypothesis is also applicable to habitat islands of a formerly continuous habitat. When compared to old forest and mixed forest, the managed forest in this study (with the exception of the irregularly occurring crossbill) contained a subset of both sets of species. Hence our macrolichen study suggests that some of these can be valid as indicators of the diversity of other macrolichens; where the signal lichen species

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

were present, other species of the set were present too. Similarly, Nilsson et al. (1995) showed that where the lichen Lobaria pulmonaria occurred, several red-listed beetles dependent on hollow trees were also found. Kuusinen (1995) suggested the lichens Lobaria pulmonaria, Nephroma spp, Pannaria pezizoides and Parmeliella triptophylla as indicators of long forest continuity in Finland. Accordingly, such indicator species could be used for identifying ``key-biotopes'' as core areas for a number of habitat specialists (Nitare and NoreÂn, 1992). The observed relationships between species and these simple habitat measurements suggest that the requirements of specialised species can be translated to quantitative habitat guidelines that can be used in practical forest management. Our results suggest that high stand age (older than the upper age for clear-cutting, i.e. >120 years) and a high proportion (>40%) of deciduous trees are two good and easily assessed habitat criteria of forests that serve as suitable habitat for many specialist species. Esseen et al. (1996) and Pettersson et al. (1995) have also shown that high stand age (>150 years) was a requirement of epiphytic macrolichens, e.g. Alectoria sarmentosa and their associated fauna. The abundance of lichens is important since insects use the lichen thalli as shelter and the insect abundance is higher when lichen abundance is high. Our observation that the lichen abundance was highest in MIX coincides with the fact that MIX also had the highest number of birds. If there is a causal link between lichen and bird abundance then lichen abundance is a potential forest biodiversity indicator. However, neither of the above-mentioned factors is given in present-day forest descriptions. Firstly, there is no resolution within age classes over 120 or 140 years. Secondly, the deciduous component often forms linear elements or very small patches which are not accounted for in forest management plans, but require special inventories, for example by remote sensing using infrared spectral frequencies, now commonly available. Finally, there are no attempts to separate aspen and birch in the stand descriptions, which is essential since these trees host di€erent species, e.g. several rare lichen species are associated with aspen (Kuusinen, 1996b; Uliczka and Angelstam, 1999). The problem of translating habitat requirements into forest criteria is similar for birds and lichens. Old growth is usually characterised by large amounts of dead wood and structural complexity both vertically and horizontally. As for the old and the deciduous trees, neither of these components is assessed in forest management plans. To maintain biodiversity in managed forests, identi®cation of these important habitat characteristics is crucial, but this has to be complemented by knowledge about the amount of habitat required for maintaining viable populations. With both these aspects of habitats at hand, the viability status of species'

349

populations could be assessed in practice by measuring the amount and connectivity of essential habitat components. For example, long-tailed tit was found to need at least 15% of deciduous trees in suitable age classes in the landscape for an 80% probability of occurrence (Jansson and Angelstam, 1999). Ideally, indicator species with large area requirements, so called `umbrella species' (sensu Noss, 1990), ensuring the presence also of other species in the same habitat (e.g. Martikainen et al., 1998), should be selected to represent all the di€erent forest disturbance regimes and stand types (Noss, 1990; Ohlson et al., 1997; Angelstam, 1998). To conclude, our study shows that the present Swedish use of only the ¯oristic component for assessment of conservation values of forest key-habitats (Anon., 1994) is insucient. Adding area-demanding specialists, such as certain resident birds, is a step forward towards a reliable biodiversity indicator system. However, it is also important to secure a stable patch dynamics of di€erent forest habitats for the maintenance of viable populations of all naturally occurring boreal species. In fact, when specialist species' quantitative requirements are well known and de®ned, these can be translated into simple forest criteria. These criteria should include the necessary amount of vital qualities for forest species conservation, like non-commercial tree species and old forest stands.

Acknowledgements We thank Per-Anders Esseen, Gunnar Jansson, KarlErik Renhorn, Mikko Kuusinen, Brian N.K. Davis and two anonymous referees for valuable comments on earlier drafts of this manuscript.

References AndreÂn, H., 1994. E€ects of habitat fragmentation on birds and mammals in landscapes with di€erent proportion of habitat: a review. Oikos 71, 355±366. Angelstam, P., 1997. Landscape analysis as tool for the scienti®c management of biodiversity. Ecological Bulletins 46, 140±170. Angelstam, P., 1998. Towards a logic for assessing biodiversity in boreal forest. In: Bachmann, P., KoÈhl, M., PaÈivinen, R. (Eds.), Assessment of Biodiversity for Improved Forest Planning. Kluwer Academic Publishers, Dordrecht, pp. 301±313. Angelstam, P., Mikusinski, G., 1994. Woodpecker assemblages in natural and managed boreal and hemiboreal forest Ð a review. Annales Zoologici Fennici 31, 157±172. Angelstam, P., Pettersson, B., 1997. Principles of present Swedish forest biodiversity management. Ecological Bulletins 46, 191±203. Angelstam, P., Rosenberg, P., Pettersson, B., 1993. Hur maÈter vi naturvaÊrdsmaÊlet? Skog och Forskning 93 (1), 28±33. Anon., 1985. Praktisk Skogshandbok. Sveriges skogsvaÊrdsfoÈrbund, Barometerns civiltryckeri AB, Kalmar. Anon, 1986. FaÊglar i Sverige. SOF, Stockholm.

350

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351

Anon, 1993. FoÈreskrifter och AllmaÈnna raÊd till SkogsvaÊrdslagen. Skogsstyrelsen, JoÈnkoÈping. Anon, 1994. Signalarter i Projekt Nyckelbiotoper. Skogsstyrelsen, JoÈnkoÈping. Attman, A., 1986. Svenskt jaÈrn och staÊl 1808±1914. Jernkontorets bergshistoriska skriftserie 21, 180. Berg, AÊ., EhnstroÈm, B., Gustafsson, L., HallingbaÈck, T., Jonsell, M., Weslien, J., 1994. Threatened plant, animal and fungus species in Swedish forests Ð distribution and habitat associations. Conservation Biology 8, 718±731. BjoÈrkhem, U., Lundmark, J.-E., 1975. Markkartering av GrimsoÈs ViltforskningsomraÊde. Rapport 23, SkogshoÈgskolan, Stockholm. BostroÈm, U., 1988. FaÊgelfaunan i olika aÊldersstadier av naturskog och kulturskog i norra Sverige. VaÊr FaÊgelvaÈrld 47, 68±76, (in Swedish with English summary). Cajander, A.K., 1926. The Theory of Forest Types. Acta Forestalia Fennica No. 29, 1±108. Cajander, A.K., 1949. Forest Types and Their Signi®cance. Acta Forestalia Fennica No. 56, 1±71. Christiansen, N.L., Bartuska, A.M., Brown, J.H., Carpenter, S., D'Antonio, C., Francis, R. et al., 1996. The report of the ecological society of America committee on the scienti®c basis for ecosystem management. Ecological Applications 6, 665±691. Cutler, A., 1990. Nested faunas and extinction in fragmented habitats. Conservation Biology 5, 496±505. DeGraaf, R., Miller, R.I., 1996. Conservation of Faunal Diversity in Forested Landscapes. Chapman and Hall, London. Dudley, N., Gilmour, D., Jeanrenaud, J.-P., 1996. Forests for Life. WWF International and IUCN, Gland. Enoksson, B., Angelstam, P., Larsson, K., 1995. Deciduous trees and resident birds- the problem of fragmentation within a coniferous forest landscape. Landscape Ecology 10, 267±275. Esseen, P.-A., Renhorn, K.-E., Pettersson, R.B., 1996. Epiphytic lichen biomass in managed and old-growth boreal forests: e€ect of branch quality. Ecological Applications 6, 228±238. Gauslaa, Y., 1994. Lungenever, Lobaria pulmonaria, som indikator paÊ artsrike kontinuitetsskoger. Blyttia 3, 119±129, (in Norwegian with English summary). Gauslaa, Y., Solhaug, K.A., 1996. Di€erences in the susceptibility to light stress between epiphytic lichens of ancient and young forest stands. Functional Ecology 10, 344±354. HallingbaÈck, T., 1994. Ekologisk Katalog oÈver Lavar. ArtDatabanken, SLU, Uppsala. Hansson, L., 1992. Nature Conservation by Ecological Principles Ð a Boreal Perspective. Elsevier, London. Hansson, L., 1997. Boreal ecosystems and landscapes: structures, processes and conservation of biodiversity. Ecological Bulletins 46. Hermansson, J., Lundquist, R., HallingbaÈck, T., 1993. Signalarter foÈr Dalarna. SkogsvaÊrdsstyrelsen, JoÈnkoÈping. Heywood, V.H., 1995. Global Biodiversity Assessment. Cambridge University Press, Cambridge. Hill, M.O., Bunce, G.H., Shaw, M.W., 1975. Indicator species analysis, a divisive polythetic method of classi®cation, and its application to a survey of native pinewoods in Scotland. Journal of Ecology 63, 597±613. Hultengren, S., Martinsson, P.-O. StenstroÈm, J. 1991. Lavar och luftfoÈroreningar. KaÈnslighetsklassning och indexberaÈkning av epifytiska lavar. Rapport 3967, NaturvaÊrdsverket, Solna. Hunter, M.L., 1999. Maintaining biodiversity in Forest Ecosystem. Cambridge University Press, Cambridge. HaÈgglund, B., Lundmark, J.E., 1984. Handledning i bonitering med skogshoÈgskolans boniteringssystem. Del 3- Markvegetationstyperskogsmarks¯ora. Skogsstyrelsen, JoÈnkoÈping. Jansson, G., 1998. Guild indicator species on a landscape scale Ð an example with four avian habitat specialists. Ornis Fennica 75, 119± 127.

Jansson, G., Angelstam, P., 1999. Threshold levels of habitat composition for the presence of the long-tailed tit (Aegithalos caudatus) in a boreal landscape. Landscape Ecology 14, 283±290. JaÈrvinen, O., Kuusela, K., VaÈisaÈnen, R., 1977. E€ects of modern forestry on the number of breeding birds in Finland 1945±1975. Silva Fennica 11, 284±294. JaÈrvinen, O., VaÈisaÈnen, R.A., 1977. Long-term changes of the North European land bird fauna. Oikos 29, 225±228. KarstroÈm, M., 1992. Steget foÈre- en presentation. (The project one step ahead Ð a presentation). Svensk Botanisk Tidskrift 86, 103±114. KarstroÈm, M., 1993. Indikatorarter som biologisk inventeringsmetod. In: Olsson, G. (Ed.), Indikatorarter foÈr Identi®ering av Naturskogar i Norrbotten. NaturvaÊrdsverket, Solna, pp. 19±96. Kuusinen, M., 1995. Epiphytic lichen diversity on Salix caprea and Populus tremula in old-growth forests of Finland. Mitt. EidgenoÈss. Forsch. anst. Wald Schnee Landsch 70, 125±132. Kuusinen, M., 1996a. Cyanobacterial macrolichens on Populus tremula as indicators of forest continuity in Finland. Biological Conservation 75, 43±49. Kuusinen, M., 1996b. Epiphyte ¯ora and diversity on basal trunks of six old-growth forest tree species in southern and middle boreal Finland. Lichenologist 28, 443±463. Landres, P.B., Verner, J., Thomas, J.W., 1988. Ecological uses of vertbrate indicator species. Conservation Biology 2, 316±328. La Roi, G., Strong, W.L., Pluth, D.J., 1988. Understory plant community classi®cations as predictors of forest site quality for lodgepole pine and white spruce in west-central Alberta. Canadian Journal of Forest Research 18, 875±887. Martikainen, P., Kaila, L., Haila, Y., 1998. Threatened beetles in White-backed Woodpecker habitats. Conservation biology 12, 293± 301. McCune, B., 1990. Rapid estimation of abundance of epiphytes on branches. The Bryologist 93, 39±43. Moberg, R., HolmaÊsen, I., 1982. Lavar, en FaÈlthandbok. Interpublishing AB, Stockholm. Nash, T.H., Olafsen, A.G., 1995. Climate change and the ecophysiological response of Arctic lichens. Lichenologist 27, 559±565. Nilsson, S.G., 1979. E€ects of forest management on the breeding bird community in southern Sweden. Biological Conservation 16, 135± 143. Nilsson, S.G., 1997. Forests in the temperate-boreal transition: natural and man-made features. Ecological Bulletins 46, 61±71. Nilsson, S.G., Arup, U., Baranowski, R., Ekman, S., 1995. Treedependent lichens and beetles as indicators in conservation forests. Conservation Biology 9, 1208±1215. Nitare, J., NoreÂn, M., 1992. Nyckelbiotoper kartlaÈggs i nytt projekt vid Skogsstyrelsen. (Woodland key-habitats of rare and endangered species will be mapped in a new project of the Swedish National Board of Forestry). Svensk Botanisk Tidskrift 86, 219±226. Noss, R.F., 1990. Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology 4, 355±363. Noss, R.F., 1993. Sustainable forestry or sustainable forests?. In: Aplet, G.H., Johnson, N., Olson, J.T., Sample, V.A. (Eds.), De®ning sustainable forestry. Island press, Washington, D.C, pp. 17±41. Ohlson, M., SoÈderstroÈm, L., HoÈrnberg, G., Zackrisson, O., Hermansson, J., 1997. Habitat qualities versus long-term continuity as determinants of biodiversity in boreal old-growth swamp forests. Biological Conservation 81, 221±231. Patterson, B.D., 1987. The principle of nested subsets and its implications for biological conservation. Conservation Biology 1, 323± 334. Peterken, G.F., 1993. Woodland Conservation and Management. Chapman and Hall, London. Pettersson, R.B., Ball, J.B., Renhorn, K.-E., Esseen, P.-A., SjoÈberg, K., 1995. Invertebrate Communities in Boreal Forest Canopies as In¯uenced by forestry and lichens with implications for Passerine Birds. Biological Conservation 74, 57±63.

H. Uliczka, P. Angelstam / Biological Conservation 95 (2000) 343±351 Pregitzer, K.S., Barnes, B.V., 1982. The use of ground ¯ora to indicate edaphic factors in upland ecosystems of the McCormick experimental forest, Upper Michigan. Canadian Journal of Forest Research 12, 661±672. Rose, F., 1992. Temperate forest management: its e€ect on bryophyte and lichen ¯oras and habitats. In: Bates, J., Farmer, A.M. (Eds.), Bryophytes and Lichens in a Changing Environment. Oxford Science Publications, Oxford, pp. 211±233. Rose, F., 1996. Lichenological indicators of age and environmental continuity in woodlands. In: Brown, D.H., Hawksworth, D.L., Bailey, R.H. (Eds.), Lichenology: progress and problems. Academic Press, London, pp. 279±307. Santesson, R., 1993. The Lichens and Lichenicolos Fungi of Sweden and Norway. SBT-foÈrlaget, Lund. Spellerberg, I.E., 1991. Monitoring ecological change. Cambridge University Press, Cambridge. Stanners, D., Bourdeau, P., 1995. Europe's environment. The Dobris assessment. European Environment Agency, Copenhagen.

351

Tibell, L., 1992. Crustose lichens as indicators of forest continuity in boreal coniferous forests. Nordic Journal of Botany 12, 427±450. Uliczka, H., Angelstam, P., 1999. Occurrence of epiphytic macrolichens in relation to tree species and age in managed boreal forest. Ecography 22, 396±405. Virkkala, R., RajasaÈrkkaÈ, A., VaÈisaÈnen, R.A., Vickholm, M., Virolainen, E., 1994. Conservation value of nature reserves: do holenesting birds prefer protected forest in southern Finland? Annales Zoologici Fennici 31, 173±186. VaÈisaÈnen, R.A., JaÈrvinen, O., Rauhala, P., 1986. How are extensive, human-caused habitat alteration expressed on the scale of local bird populations in boreal forests? Ornis Scandinavica 17, 282± 292. Wieslander, G., 1936. The shortage of forest in Sweden during the 17th and 18th centuries. Sveriges skogsvaÊrdsfoÈrbunds tidskrift 34, 593±633.