Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Spore deposition of wood‐decaying fungi: importance of landscape composition

Spore deposition of wood‐decaying fungi: importance of landscape composition Fragmentation and loss of natural forests due to forestry and other human activities pose major threats to biodiversity all over the world (see, for instance, Harris 1984 ). Thus, from a conservation perspective it is important to document the ecological effects of forestry and to improve our understanding of the subsequent effects on various organisms over a wide range of spatial and temporal scales. Efforts to achieve these goals should include comparative studies of ecological processes over a range of scales in various forest ecosystems with differing intensity and history of management. Although most studies on forestry effects have been performed at a local spatial scale, many ecological processes operate on larger, landscape scales ( Hansson et al. 1995 ). Thus, there is a particular need for studies that consider larger spatial scales. During the last 150 yr, forestry has caused a dramatic decline in the amount of natural and semi‐natural forests in Sweden, and <5% of the remaining forest is in an old‐growth condition (>140 yr) ( Anon. 2001 ). The regions with the longest history of forestry are located in the southern parts and along the Bothnian coast in the east. Consequently, the proportion of old forests is lowest in these areas ( Linder and Östlund 1992 , Östlund 1993 , Axelsson 2001 ). In addition, forestry practices have changed considerably during the 20th century, causing a dramatic decrease in coarse woody debris (CWD). The average volume of CWD in Swedish old‐growth spruce forests ranges between 39 and 133 m 3 ha −1 ( Linder and Östlund 1998 , Jonsson 2000 ), but the current average volume in managed forests with closed canopies is just 3.0 m 3 ha −1 ( Kruys et al. 1999 ). The lower amount of CWD in managed forests, together with a more fragmented occurrence of old‐growth forests, has led to reductions in the diversity of many saproxylic organisms, among them wood‐decaying fungi (e.g. Siitonen 2001 ). Today, 28% of a total of 829 species in the order Aphyllophorales, which are mainly wood‐decomposers, are on the Swedish Red Data List of threatened fungi ( Gärdenfors 2000 ). Nevertheless, some species of wood‐decaying fungi are abundant in managed forests whereas others, although they may have large local populations, are almost exclusively found in old‐growth stands ( Larsson 1997 ). Wood‐decaying fungi are primarily dispersed by airborne basidiospores, although some species are able to spread by mycelial expansion through the soil (see, for instance, Rayner and Boddy 1988 ) or by airborne asexual propagules ( Vasiliauskas et al. 1998 ). There are also examples of species that are dispersed by insect vectors ( Nuss 1982 , Wikars 1997 , Thomsen and Koch 1999 , Vasiliauskas and Stenlid 1999 ). However, for the majority of species, the long‐term survival in fragmented forest landscapes is probably dependent on long‐distance dispersal (>1 km) of airborne basidiospores. The spore dispersal process of wood‐decaying fungi may be affected by habitat loss and forest fragmentation in two major ways. First, fragmentation reduces the size of local populations, causing total spore production from each population to decline. Second, the distance between suitable habitat patches increases. Thus, efficient dispersal of the fungi is a prerequisite in order to maintain gene flow between isolated populations and to prevent local extinctions. Most studies on dispersal of wood‐decaying fungi have so far been performed on a local scale, at distances of 1 km or less (e.g. Penttilä et al. 1999 , Nordén and Larsson 2000 ). Thus, for a better understanding of the dispersal that occurs at a landscape scale, there is a need to study dispersal over longer distances. The aims of this study were to investigate the effect of forest fragmentation and habitat loss on the dispersal of wood‐decaying fungi in two ways: 1) the background spore deposition was compared in forest landscapes subjected to different intensities of forest management, i.e. forest landscapes with different proportions of old spruce forest; 2) by identifying the effective scale of the dispersal processes by correlating the background spore deposition to the proportion of suitable habitat in the surrounding landscape, at different spatial scales. Methods Study species Five species of wood‐decaying fungi were studied, namely Fomitopsis pinicola (Sw.: Fr.) P. Karst., Fomitopsis rosea (Alb. & Schwein.: Fr.) P. Karst., Trichaptum laricinum (P. Karst.) Ryvarden, Gloeoporus taxicola (Pers.: Fr.) Gilb. & Ryvarden and Phlebia centrifuga P. Karst. All of these species have a circumboreal distribution. Fomitopsis pinicola is a common generalist fungus, widely distributed across the forest landscape in Sweden, whereas F. rosea , P. centrifuga and T. laricinum are usually confined to coarse logs in old‐growth boreal forests. All three have decreased dramatically due to modern intensive forestry, and are now on the Swedish Red List of threatened species ( Gärdenfors 2000 ). Gloeoporus taxicola is fairly common in the northern boreal zone, but becomes rarer further south. In Scandinavia, F. rosea and P. centrifuga are confined to spruce, while the other species also may inhabit pine wood. Measuring spore deposition Two slightly different methods were used to measure the spore deposition, both using growing haploid mycelia as bait. Agar spore traps In the first method, monokaryotic mycelia were grown on petri dishes (∅ 90 mm) with Hagem agar ( Stenlid 1985 ). Only one monokaryotic mycelium was cultivated on each petri dish. Deposited spores, if compatible, can germinate and fuse with the monokaryotic mycelium, forming a dikaryotic mycelium (see, for example, Adams et al. 1984 ). For all studied species, dikaryons can then be recognised by the presence of clamp connections, and when these were found in the petri dishes, a spore hit was registered. Since it is not usually possible to distinguish between separate, different spore hits on one and the same petri dish, this method of monitoring spore deposition is considered to yield qualitative data, i.e. merely the presence or absence of spores. Wood‐discs In the second method, haploid mycelia grown on wood‐discs were used. The method is based on the same principle as for the agar method. However, the wood‐discs are more robust and allow longer periods of exposure ( Edman and Gustafsson 2003 ). Compared to the agar method, wood‐discs are more resistant to contamination, rain, drying and freezing temperatures. In addition it is possible to separate different spore hits by identifying the incompatibility zones that appear between the different dikaryons that are formed ( Edman and Gustafsson 2003 ). However, the method has limitations, since not all species produce mycelia that are strong enough to function as reliable spore traps, and consequently this method was only used for F. rosea and P. centrifuga . For both methods, the fungi were inoculated three months before exposure. Only one isolate per species was selected, to prevent cross contamination between field replicates. The isolates had previously been tested and selected for their abilities to grow well and to form clamps readily when paired in the laboratory. Study areas and field methods Effects of forestry intensity – agar spore traps The study was performed in two geographically separate regions in the boreal zone of Sweden ( Ahti et al. 1968 ): one in northern Sweden in the north‐boreal zone (close to 64°N latitude) and the other in southern Sweden in the hemiboreal zone (close to 60°N latitude) ( Fig. 1 ). The forests in both regions are dominated by Norway spruce ( Picea abies (L.) Karst.) and Scots pine ( Pinus sylvestris L.). In all, 13 pairs of circular study plots (2 km radius, 1256 ha) were selected for the study, seven in the northern region and six in the southern region. The paired plots were located relatively close to each other (see below) and were chosen to differ with regard to the proportion of old spruce forest (>80 yr), i.e. one plot in each pair had a significantly higher proportion of old spruce forest than the other plot (see Appendix 1 for details). The proportion of old spruce forest within 2 km radius was calculated from forest stand registers and associated maps. 1 Map of Sweden showing the sampling localities of the spore‐trapping experiments. Each locality includes two sites, one with a relatively high proportion of old spruce forest and one with a lower proportion of old spruce forest. Black circles show the northern localities and black squares the southern localities. All localities were used in the study of forestry intensity, while only the northern localities were used in the scale study. Spores were sampled close to the ground in the centre point of each plot. Centre points were chosen in forest stands with at least five‐metre high trees in order to slow down the wind and facilitate deposition of the spores. The tree cover also protected the spore‐traps from drying out in the sun. To avoid overrepresentation of fruit bodies in the near vicinity, only sites without old spruce forest within 400 m radius from the centres were chosen. In addition, logs and snags within 400 m of the plot‐centres were carefully examined for fruit bodies of the species under investigation, and only sites without fruit bodies were accepted. The distance between the plot‐centres of the pairs ranged from 6 to 10 km and the distance between adjacent pairs from 21 to 75 km. On two occasions during August–October 1999, 12 petri dishes with haploid mycelia of each studied species were placed at the centre of each plot. To compensate for possible diurnal fluctuations in spore settlement the spore traps were exposed for 24 h. After exposure, the petri dishes were sealed, brought to the laboratory and placed in darkness at room temperature for 1–2 months. Thereafter they were examined for the occurrence of clamp connections at 100× magnification. Effects of scale and forest age – wood‐discs In order to study the importance of forest age on the spore deposition and to determine the effective scale of spore dispersal, wood‐discs with monokaryotic mycelia of F. rosea and P. centrifuga were used. Twelve wood‐discs of each species were placed in the centre of the plots and exposed for two weeks in august 2000. After exposure, the wood‐discs were put into petri dishes, which were sealed and placed in darkness at room temperature for 1–2 months. Thereafter they were examined for spore hits as indicated by mycelial clamps and incompatibility zones at the wood disc surface. This investigation was only performed in the northern region, where all seven pairs of plots were used ( Fig. 1 ). The age and size of all forest stands containing spruce within a 3 km radius (2827 ha) of the plot centre were obtained from forest stand registers and associated maps. To evaluate the effects of the area occupied by other tree species in mixed stands, the proportion of spruce was multiplied by the stand area. In addition, the spruce stands were divided into three age classes: >80 yr, >110 yr and >140 yr and the proportion of each class within 1‐, 2‐ and 3‐km radii of the plot centres were calculated ( Fig. 2 ). Since the plots lacked old spruce forest within 400 m from the centre points, the areas within 1‐km radius contained a lower proportion of old spruce forest compared to the areas within 1‐ and 2‐km radius ( Fig. 2 ). 2 Forest data for plots with low proportion of old spruce forest in the northern region (top, n=7) and high proportion of old spruce forest (bottom, n=7) showing the average proportions of spruce forest for three different age classes (>80, >110 and >140 yr) within 1, 2 and 3 km from the plot centres. Statistics Effects of forestry intensity – agar spore‐traps Spore‐trapping data from F. pinicola and P. centrifuga showed non‐normal distributions, so non‐parametric tests were performed on the results for all species. In order to test for differences between paired plots with high and low proportions of old forest, in the northern and southern regions, Wilcoxon's signed rank tests were used. Spore traps with monokaryons grown on nutrient agar usually give discrete presence/absence data, even if more than one spore is responsible for the presence of dikaryotized mycelium. If the spores deposit on the ground at random we can assume that the deposition pattern follows a Poisson distribution. In contrast, the observations follow a binomial distribution with the Poisson probability of p, but we can determine the mean number of spores (m) hitting a spore trap from maximum likelihood calculations, if we know the total number of traps (n) and how many have been hit (k). The general equation becomes: This value is closer to the true number of spore hits than the figure provided by the original discrete data. Therefore, we transformed the data according to the above functions in order to estimate the true spore deposition rates. In addition, to standardize and facilitate the presentation of the spore amounts, average spore deposition values (m) were recalculated in terms of spores×m −2 ×24 h −1 . However, all statistical testing was performed on the original discrete data. Effects of scale and forest age – wood‐discs The relationship between spore deposition and the proportion of old spruce forest (>80, >110 and >140 yr) at different distances from the plot centre (1, 2 and 3 km) was tested using linear regression with number of spore hits as the dependent variable. Partial correlation was used to examine whether the contribution in forest area yielded a significant contribution to the linear regression model when increasing the spatial scale and controlling for smaller scales. The data fulfilled the assumptions of normality. Results Importance of forestry intensity – agar spore traps A comparison between the paired sample plots located within the same geographical region showed that the spore deposition of F. rosea , in both the north and the south, F. pinicola in the south and G. taxicola in the north was significantly higher in areas surrounded by a relatively high proportion of old spruce forest (>80 yr) ( Table 1 , Fig. 3 ). For T. laricinum in the north and P. centrifuga in the south, no significant differences were found. The spore‐traps for T. laricinum and G. taxicola in the south and for P. centrifuga in the north were highly contaminated and therefore excluded from the analyses. In addition, spore deposition of F. pinicola in the north was too high to conduct a relevant t‐test since in 12 of the 14 localities the maximum possible number of spore hits, 12 out of 12, occurred on the agar traps ( Fig. 3 ). 1 Wilcoxon's signed rank test (two‐tailed) for differences in spore deposition, as measured with agar spore traps, between plots with higher and lower proportions of old spruce forests in the northern and southern regions. Bold Z and p‐values indicate significant differences (p<0.05) within a given region. Species north south Z p Z p F. pinicola – 1) – −2.023 0.043 F. rosea −2.366 0.018 −2.023 0.043 G. taxicola −1.992 0.046 – 2) – T. laricinum −0.943 0.345 – 2) – P. centrifuga – 2) – −1.604 0.109 1) Omitted due to high deposition rates. 2) Omitted due to contamination. 3 Spore hits on petri dishes (∅ 9 cm) with homokaryotic mycelia grown on nutrient agar. Old spruce is the percentage of spruce forest age>80 yr old within a 2‐km‐radius of the sampling point. Filled symbols refer to the plots with the higher proportion of old spruce forest in each pair, while open symbols refer to the plots with the lower proportion of old spruce forest, in the northern (circles) and southern (squares) regions, respectively. Average spore deposition rates, transformed according to maximum likelihood calculations, showed that the rates for F. pinicola in the north were never <300 spores×m −2 ×24 h −1 , regardless of the proportion of old forest in the surrounding landscape (Appendix 1). Furthermore, the spore deposition of F. pinicola was probably heavily underestimated, since almost all of the traps were hit, implying an overload of spores. In contrast, spore deposition in the south never exceeded 126 spores×m −2 ×24 h −1 , and in several sites it was only 7 spores×m −2 ×24 h −1 or less (Appendix 1). Spore deposition of F. rosea in the north ranged between 6 and >390 spores×m −2 ×24 h −1 , while the deposition in the south was never >47 spores×m −2 ×24 h −1 . In fact, no spores were trapped at four of the southern sites. Deposition of T. laricinum in the north was generally low compared to the other studied species in that region. No spores were trapped at five of the sites, while the deposition was as high as 93 spores×m −2 ×24 h −1 at one site (Appendix 1). Deposition of G. taxicola in the north was generally high, and always exceeded 95 spores×m −2 ×24 h −1 , except on one occasion (Appendix 1). Phlebia centrifuga in the south displayed the lowest spore deposition rate of all species. Spores of this species were only trapped in four of the 12 landscapes, at rates ranging from 7 to 31 m −2 ×24 h −1 . Effects of scale and forest age – wood‐discs Both F. rosea and P. centrifuga showed a strong relationship between the amount of old forest in the surrounding landscape and the spore deposition in plots with higher proportions of old spruce forest ( Fig. 4 ). For F. rosea significant linear regression models were found for spruce forest >110 and 140 yr old within a 2‐km radius of the sampling locations and for spruce forest older than 140 yr within a 3‐km radius ( Table 2 ). In addition, the amount of the variation in spore deposition explained (adj. R 2 ), increased with increasing scale and forest age. For spruce forest older than 140 yr within a 3‐km radius the amount of variation explained was as high as 91% (adj. R 2 ) ( Table 2 , Fig. 4 ). When increasing the spatial scale, the enlargement in old forest area (>140 yr) significantly contributed to improve the partial correlation coefficient, both when controlling for 1 km (p=0.029) and for 1–2 km (p=0.003). Phlebia centrifuga showed a significant relationship between spore deposition and amount of old forest at all three scales (1, 2 and 3 km radius) ( Table 2 ). The amount of the variation in spore deposition explained (adj. R 2 ) increased with increasing scale and forest age, and for spruce forests older than 140 yr within a 3 km radius, the amount explained was 94% ( Table 2 , Fig. 4 ). In a partial correlation the contribution in larger old forest area (>140 yr) when increasing the spatial scale significantly contributed to the model when controlling for both 1 km (p=0.013) and 1–2 km (p=0.012). 4 Spore deposition of Fomitopsis rosea and Phlebia centrifuga (presented as spores per m 2 and 24 h) plotted against percentage spruce forest>140 yr old for plots with lower proportion of old spruce forest (open circles, n=7) and higher proportion of old spruce forest (black circles, n=7), respectively. Based on data obtained from wood‐discs in the northern region. 2 Results from a simple linear regression with spore deposition, measured by the wood‐disc method, as dependent variable. P‐values and adjusted R 2 values are given for different proportions of spruce forest (>80, >110 and >140 yr) at different scales (1‐, 2‐ and 3‐km radius of the plot centre) both for plots with higher proportion of old forest (n=7) and lower proportion of old forest (n=7). Species Plots with higher amounts of old forest Plots with lower amounts of old forest 1 km 2 km 3 km 1 km 2 km 3 km age p R 2 p R 2 p R 2 p R 2 p R 2 p R 2 F. rosea 80 0.281 – 0.055 – 0.055 – 0.382 – 0.181 – 0.071 – 110 0.059 – 0.031 57 0.051 – 0.620 – 0.255 – 0.095 – 140 0.102 – 0.004 80 0.000 91 0.424 – 0.271 – 0.105 – P. centrifuga 80 0.186 – 0.029 58 0.027 59 0.971 – 0.324 – 0.046 50 110 0.040 52 0.011 71 0.018 65 0.966 – 0.329 – 0.022 62 140 0.184 – 0.005 79 0.000 94 0.565 – 0.196 – 0.046 50 No significant relationships between spore deposition of F. rosea and the proportion of old forests (>80, >110 and>140 yr) were found at any scale in the plots with lower proportions of old forest. For P. centrifuga , on the other hand, the spore deposition was significantly correlated to all three forest age classes within a 3 km radius (p<0.05) ( Table 2 ). Discussion Effects of forestry For both F. rosea and F. pinicola there was a large difference in spore deposition in the paired forest landscapes (2 km radius) between those with relatively high and those with relatively low proportions of old spruce forest (>80 yr). Since the distance between the plot‐centres within pairs ranged between only 6 and 10 km, the results clearly illustrate the importance of dispersal over the range of a few kilometres, i.e. moderate dispersal distances. This is not surprising, since a number of studies have shown that the majority of spores fall very close to the fruiting bodies (e.g. Malloch and Blackwell 1992 , Möykkynen et al. 1997 ). The findings also demonstrate that forestry, even at a local scale, may have a significant effect on the dispersal of species. This, in turn, affects the distribution of fungi at a regional scale where the species may have a more restricted occurrence. For example, Siitonen et al. (2001) found lower abundance of red‐listed polypores in fragmented old‐growth forests in eastern Finland than in comparable, continuous old‐growth forests in Russian Karelia. They suggested that the difference was due to limitations in the dispersal and colonisation ability of the fungi in the fragmented forest landscape. Of particular interest is that F. pinicola , which is a common generalist fungus and is not considered threatened by forestry, displayed lower spore deposition levels in landscapes with lower proportions of old (>80 yr) spruce forests in the southern region. Thus, spore availability appears to be highly correlated to the amount of old forest nearby, even for common species. The extent to which this lower spore availability may affect the viability of the populations remains unclear. However, our data suggests that fragmentation and habitat loss may affect common species to an extent that is not fully understood and deserves further investigations. Importance of forest age The spore deposition was strongly correlated with the amount and age of spruce forest for most of our studied species, suggesting that they are favoured by structures within old forests. Old spruce forests generally contain higher volumes of CWD than younger forests ( Fridman and Walheim 2000 ), which promotes larger populations of wood‐inhabiting fungi. However, other variables connected to forest age may also be involved, such as the quality and continuity of the CWD. For example, several red‐listed species of wood‐decaying fungi show preferences for coarse logs, which are more common in old forests ( Renvall 1995 , Bader et al. 1995 , Kruys et al. 1999 , Sippola et al. 2001 ). However, no relationship between the amount of old forest and spore deposition on wood‐discs was found for F. rosea in plots with low proportions of old forest. This may reflect differences between the species, but it is more likely to be due to the test being insufficiently sensitive, because of the low amount of old forest (>140 yr) in these plots. The average amount was only 2.2 ha in a 1257 ha plot ( Fig. 2 ). Importance of scale For both F. rosea and P. centrifuga , the correlation between the rate of spore deposition and the proportion of old spruce forest was very high for the area located within a 3‐km radius of the measuring point, but decreased at 2‐km and 1‐km radii. Although 3 km was the largest radius tested, and the contribution of sources from larger distances is unknown, the high correlation between the proportion of old forest within 3 km and spore deposition (adj. R 2 >90%) indicates that the trapped spores mainly originated from populations within this distance. This is in accordance with data presented by Nordén and Larsson (2000) , who sampled spores at different distances from fruit bodies of P. centrifuga. They found that the majority of the spores were deposited in close vicinity to the fruit bodies, but also concluded that the deposition rate was still fairly high at a distance of 1 km, which was the largest distance tested. However, a small fraction of the spore deposition in the present study probably constitutes background spore deposition, originating from distant spore sources. Spore deposition curves usually show a steep and exponential decrease with distance, leaving a long tail of low deposition at large distances. For instance, Kallio (1970) showed that 99.9% of the spores released from the root rot fungus Heterobasidion annosum settled within 100 m of the fruit body, but also concluded that some spores of the species spread as far as 70–500 km. Spore deposition amounts For most of the species we found fairly high rates of spore deposition, suggesting that the availability of spores for colonisation is fairly good at a local scale. For example, the spore deposition of the red‐listed T. laricinum , which had by far the lowest deposition rate of the tested species in the north, was still >10 spores m −2 24 h −1 at half of the sampling sites. Fairly high spore deposition rates were also found for the threatened fungi F. rosea and P. centrifuga in several sites in the southern region, where the impact of forestry has been high for several hundred years. This indicates that factors other than dispersal may be important in the colonisation process of spruce logs, such as various aspects of wood quality, including log size, decay stage ( Bader et al. 1995 , Kruys et al. 1999 ) and wood density. Although spore depositions were fairly high overall, the relative differences in deposition between forest landscapes with differing intensities of forestry were also high for several species. There were also considerable differences in deposition between the northern and southern regions. For example, spore deposition of F. rosea was on average ca 10 times higher in the north compared to the south. There was also a large relative difference in spore deposition between species. Although the spore deposition of F. pinicola was highly underestimated in the north, ca 10 times more spores were trapped compared to F. rosea in the landscapes with lower amounts of old forest. Since factors like germination requirements and competition from other mycelia probably also affect colonisation ability ( Stenlid 1993 ), the spatial and temporal window available for colonisation is most likely very small. Thus, dispersal limitation may still be a significant factor in the population dynamics of these fungi, since insufficient densities of spores may be present for successful establishment of mycelia and dikaryotisation. Conclusions Our data strongly suggest that the landscape composition has strong, indirect effects on the availability of spores of wood‐decaying fungi. Both the amount and age of old spruce forests are key factors in predicting the rates of spore deposition. Increases in the intensity of forestry tend to reduce the proportion of old (>80 yr) spruce forests, which in turn negatively affects the fungi and reduces the abundance of their spores. The strong correlation between the rates of spore deposition of F. rosea and P. centrifuga and the proportion of old (>140‐yr) spruce forest within a 3 km radius indicate that this may be a relevant scale at which to consider the needs of these species in conservation and management planning. Furthermore, it emphasizes the importance of very old forests (>140 yr) to wood‐decaying fungi. The age of these stands exceeds the common harvest rotation period, which is ca 80–120 yr, and consequently the ongoing decline in old‐growth forests clearly constitutes a risk for the long‐term survival of these species in the boreal forest landscape. However, the significance of old forests is probably connected to the abundance of CWD, and increasing the CWD volumes in younger stands may enhance the possibility for wood‐decaying species to survive also in the managed forest landscape. Acknowledgements We thank David Rönnblom, Lars Persson, Christer Ahlenius and Mats Englund at Holmen Skog, Kjell‐Ove Häggström at Sveaskog, Eva Ståhl and Bo Ernstsson at AssiDomän, Urban Ekskär at SCA and The Local Board of Forestry (Skogsvårdsstyrelsen) for providing forest stand maps and valuable information on different forest sites. We also thank Kristoffer and Eva Hylander for help with maximum likelihood calculations of spore deposition data. This study has been financed by The Science Research Council (VR); grants to L. Ericson and The Swedish Council for Forestry and Agricultural Research (SJFR); grants to J. Stenlid. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecography Wiley

Spore deposition of wood‐decaying fungi: importance of landscape composition

Download PDF
9 pages

Loading next page...
1
 
/lp/wiley/spore-deposition-of-wood-decaying-fungi-importance-of-landscape-1Jscl7iKR4

References (33)

Publisher
Wiley
Copyright
Copyright © 2004 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-7590
eISSN
1600-0587
DOI
10.1111/j.0906-7590.2004.03671.x
Publisher site
See Article on Publisher Site

Abstract

Fragmentation and loss of natural forests due to forestry and other human activities pose major threats to biodiversity all over the world (see, for instance, Harris 1984 ). Thus, from a conservation perspective it is important to document the ecological effects of forestry and to improve our understanding of the subsequent effects on various organisms over a wide range of spatial and temporal scales. Efforts to achieve these goals should include comparative studies of ecological processes over a range of scales in various forest ecosystems with differing intensity and history of management. Although most studies on forestry effects have been performed at a local spatial scale, many ecological processes operate on larger, landscape scales ( Hansson et al. 1995 ). Thus, there is a particular need for studies that consider larger spatial scales. During the last 150 yr, forestry has caused a dramatic decline in the amount of natural and semi‐natural forests in Sweden, and <5% of the remaining forest is in an old‐growth condition (>140 yr) ( Anon. 2001 ). The regions with the longest history of forestry are located in the southern parts and along the Bothnian coast in the east. Consequently, the proportion of old forests is lowest in these areas ( Linder and Östlund 1992 , Östlund 1993 , Axelsson 2001 ). In addition, forestry practices have changed considerably during the 20th century, causing a dramatic decrease in coarse woody debris (CWD). The average volume of CWD in Swedish old‐growth spruce forests ranges between 39 and 133 m 3 ha −1 ( Linder and Östlund 1998 , Jonsson 2000 ), but the current average volume in managed forests with closed canopies is just 3.0 m 3 ha −1 ( Kruys et al. 1999 ). The lower amount of CWD in managed forests, together with a more fragmented occurrence of old‐growth forests, has led to reductions in the diversity of many saproxylic organisms, among them wood‐decaying fungi (e.g. Siitonen 2001 ). Today, 28% of a total of 829 species in the order Aphyllophorales, which are mainly wood‐decomposers, are on the Swedish Red Data List of threatened fungi ( Gärdenfors 2000 ). Nevertheless, some species of wood‐decaying fungi are abundant in managed forests whereas others, although they may have large local populations, are almost exclusively found in old‐growth stands ( Larsson 1997 ). Wood‐decaying fungi are primarily dispersed by airborne basidiospores, although some species are able to spread by mycelial expansion through the soil (see, for instance, Rayner and Boddy 1988 ) or by airborne asexual propagules ( Vasiliauskas et al. 1998 ). There are also examples of species that are dispersed by insect vectors ( Nuss 1982 , Wikars 1997 , Thomsen and Koch 1999 , Vasiliauskas and Stenlid 1999 ). However, for the majority of species, the long‐term survival in fragmented forest landscapes is probably dependent on long‐distance dispersal (>1 km) of airborne basidiospores. The spore dispersal process of wood‐decaying fungi may be affected by habitat loss and forest fragmentation in two major ways. First, fragmentation reduces the size of local populations, causing total spore production from each population to decline. Second, the distance between suitable habitat patches increases. Thus, efficient dispersal of the fungi is a prerequisite in order to maintain gene flow between isolated populations and to prevent local extinctions. Most studies on dispersal of wood‐decaying fungi have so far been performed on a local scale, at distances of 1 km or less (e.g. Penttilä et al. 1999 , Nordén and Larsson 2000 ). Thus, for a better understanding of the dispersal that occurs at a landscape scale, there is a need to study dispersal over longer distances. The aims of this study were to investigate the effect of forest fragmentation and habitat loss on the dispersal of wood‐decaying fungi in two ways: 1) the background spore deposition was compared in forest landscapes subjected to different intensities of forest management, i.e. forest landscapes with different proportions of old spruce forest; 2) by identifying the effective scale of the dispersal processes by correlating the background spore deposition to the proportion of suitable habitat in the surrounding landscape, at different spatial scales. Methods Study species Five species of wood‐decaying fungi were studied, namely Fomitopsis pinicola (Sw.: Fr.) P. Karst., Fomitopsis rosea (Alb. & Schwein.: Fr.) P. Karst., Trichaptum laricinum (P. Karst.) Ryvarden, Gloeoporus taxicola (Pers.: Fr.) Gilb. & Ryvarden and Phlebia centrifuga P. Karst. All of these species have a circumboreal distribution. Fomitopsis pinicola is a common generalist fungus, widely distributed across the forest landscape in Sweden, whereas F. rosea , P. centrifuga and T. laricinum are usually confined to coarse logs in old‐growth boreal forests. All three have decreased dramatically due to modern intensive forestry, and are now on the Swedish Red List of threatened species ( Gärdenfors 2000 ). Gloeoporus taxicola is fairly common in the northern boreal zone, but becomes rarer further south. In Scandinavia, F. rosea and P. centrifuga are confined to spruce, while the other species also may inhabit pine wood. Measuring spore deposition Two slightly different methods were used to measure the spore deposition, both using growing haploid mycelia as bait. Agar spore traps In the first method, monokaryotic mycelia were grown on petri dishes (∅ 90 mm) with Hagem agar ( Stenlid 1985 ). Only one monokaryotic mycelium was cultivated on each petri dish. Deposited spores, if compatible, can germinate and fuse with the monokaryotic mycelium, forming a dikaryotic mycelium (see, for example, Adams et al. 1984 ). For all studied species, dikaryons can then be recognised by the presence of clamp connections, and when these were found in the petri dishes, a spore hit was registered. Since it is not usually possible to distinguish between separate, different spore hits on one and the same petri dish, this method of monitoring spore deposition is considered to yield qualitative data, i.e. merely the presence or absence of spores. Wood‐discs In the second method, haploid mycelia grown on wood‐discs were used. The method is based on the same principle as for the agar method. However, the wood‐discs are more robust and allow longer periods of exposure ( Edman and Gustafsson 2003 ). Compared to the agar method, wood‐discs are more resistant to contamination, rain, drying and freezing temperatures. In addition it is possible to separate different spore hits by identifying the incompatibility zones that appear between the different dikaryons that are formed ( Edman and Gustafsson 2003 ). However, the method has limitations, since not all species produce mycelia that are strong enough to function as reliable spore traps, and consequently this method was only used for F. rosea and P. centrifuga . For both methods, the fungi were inoculated three months before exposure. Only one isolate per species was selected, to prevent cross contamination between field replicates. The isolates had previously been tested and selected for their abilities to grow well and to form clamps readily when paired in the laboratory. Study areas and field methods Effects of forestry intensity – agar spore traps The study was performed in two geographically separate regions in the boreal zone of Sweden ( Ahti et al. 1968 ): one in northern Sweden in the north‐boreal zone (close to 64°N latitude) and the other in southern Sweden in the hemiboreal zone (close to 60°N latitude) ( Fig. 1 ). The forests in both regions are dominated by Norway spruce ( Picea abies (L.) Karst.) and Scots pine ( Pinus sylvestris L.). In all, 13 pairs of circular study plots (2 km radius, 1256 ha) were selected for the study, seven in the northern region and six in the southern region. The paired plots were located relatively close to each other (see below) and were chosen to differ with regard to the proportion of old spruce forest (>80 yr), i.e. one plot in each pair had a significantly higher proportion of old spruce forest than the other plot (see Appendix 1 for details). The proportion of old spruce forest within 2 km radius was calculated from forest stand registers and associated maps. 1 Map of Sweden showing the sampling localities of the spore‐trapping experiments. Each locality includes two sites, one with a relatively high proportion of old spruce forest and one with a lower proportion of old spruce forest. Black circles show the northern localities and black squares the southern localities. All localities were used in the study of forestry intensity, while only the northern localities were used in the scale study. Spores were sampled close to the ground in the centre point of each plot. Centre points were chosen in forest stands with at least five‐metre high trees in order to slow down the wind and facilitate deposition of the spores. The tree cover also protected the spore‐traps from drying out in the sun. To avoid overrepresentation of fruit bodies in the near vicinity, only sites without old spruce forest within 400 m radius from the centres were chosen. In addition, logs and snags within 400 m of the plot‐centres were carefully examined for fruit bodies of the species under investigation, and only sites without fruit bodies were accepted. The distance between the plot‐centres of the pairs ranged from 6 to 10 km and the distance between adjacent pairs from 21 to 75 km. On two occasions during August–October 1999, 12 petri dishes with haploid mycelia of each studied species were placed at the centre of each plot. To compensate for possible diurnal fluctuations in spore settlement the spore traps were exposed for 24 h. After exposure, the petri dishes were sealed, brought to the laboratory and placed in darkness at room temperature for 1–2 months. Thereafter they were examined for the occurrence of clamp connections at 100× magnification. Effects of scale and forest age – wood‐discs In order to study the importance of forest age on the spore deposition and to determine the effective scale of spore dispersal, wood‐discs with monokaryotic mycelia of F. rosea and P. centrifuga were used. Twelve wood‐discs of each species were placed in the centre of the plots and exposed for two weeks in august 2000. After exposure, the wood‐discs were put into petri dishes, which were sealed and placed in darkness at room temperature for 1–2 months. Thereafter they were examined for spore hits as indicated by mycelial clamps and incompatibility zones at the wood disc surface. This investigation was only performed in the northern region, where all seven pairs of plots were used ( Fig. 1 ). The age and size of all forest stands containing spruce within a 3 km radius (2827 ha) of the plot centre were obtained from forest stand registers and associated maps. To evaluate the effects of the area occupied by other tree species in mixed stands, the proportion of spruce was multiplied by the stand area. In addition, the spruce stands were divided into three age classes: >80 yr, >110 yr and >140 yr and the proportion of each class within 1‐, 2‐ and 3‐km radii of the plot centres were calculated ( Fig. 2 ). Since the plots lacked old spruce forest within 400 m from the centre points, the areas within 1‐km radius contained a lower proportion of old spruce forest compared to the areas within 1‐ and 2‐km radius ( Fig. 2 ). 2 Forest data for plots with low proportion of old spruce forest in the northern region (top, n=7) and high proportion of old spruce forest (bottom, n=7) showing the average proportions of spruce forest for three different age classes (>80, >110 and >140 yr) within 1, 2 and 3 km from the plot centres. Statistics Effects of forestry intensity – agar spore‐traps Spore‐trapping data from F. pinicola and P. centrifuga showed non‐normal distributions, so non‐parametric tests were performed on the results for all species. In order to test for differences between paired plots with high and low proportions of old forest, in the northern and southern regions, Wilcoxon's signed rank tests were used. Spore traps with monokaryons grown on nutrient agar usually give discrete presence/absence data, even if more than one spore is responsible for the presence of dikaryotized mycelium. If the spores deposit on the ground at random we can assume that the deposition pattern follows a Poisson distribution. In contrast, the observations follow a binomial distribution with the Poisson probability of p, but we can determine the mean number of spores (m) hitting a spore trap from maximum likelihood calculations, if we know the total number of traps (n) and how many have been hit (k). The general equation becomes: This value is closer to the true number of spore hits than the figure provided by the original discrete data. Therefore, we transformed the data according to the above functions in order to estimate the true spore deposition rates. In addition, to standardize and facilitate the presentation of the spore amounts, average spore deposition values (m) were recalculated in terms of spores×m −2 ×24 h −1 . However, all statistical testing was performed on the original discrete data. Effects of scale and forest age – wood‐discs The relationship between spore deposition and the proportion of old spruce forest (>80, >110 and >140 yr) at different distances from the plot centre (1, 2 and 3 km) was tested using linear regression with number of spore hits as the dependent variable. Partial correlation was used to examine whether the contribution in forest area yielded a significant contribution to the linear regression model when increasing the spatial scale and controlling for smaller scales. The data fulfilled the assumptions of normality. Results Importance of forestry intensity – agar spore traps A comparison between the paired sample plots located within the same geographical region showed that the spore deposition of F. rosea , in both the north and the south, F. pinicola in the south and G. taxicola in the north was significantly higher in areas surrounded by a relatively high proportion of old spruce forest (>80 yr) ( Table 1 , Fig. 3 ). For T. laricinum in the north and P. centrifuga in the south, no significant differences were found. The spore‐traps for T. laricinum and G. taxicola in the south and for P. centrifuga in the north were highly contaminated and therefore excluded from the analyses. In addition, spore deposition of F. pinicola in the north was too high to conduct a relevant t‐test since in 12 of the 14 localities the maximum possible number of spore hits, 12 out of 12, occurred on the agar traps ( Fig. 3 ). 1 Wilcoxon's signed rank test (two‐tailed) for differences in spore deposition, as measured with agar spore traps, between plots with higher and lower proportions of old spruce forests in the northern and southern regions. Bold Z and p‐values indicate significant differences (p<0.05) within a given region. Species north south Z p Z p F. pinicola – 1) – −2.023 0.043 F. rosea −2.366 0.018 −2.023 0.043 G. taxicola −1.992 0.046 – 2) – T. laricinum −0.943 0.345 – 2) – P. centrifuga – 2) – −1.604 0.109 1) Omitted due to high deposition rates. 2) Omitted due to contamination. 3 Spore hits on petri dishes (∅ 9 cm) with homokaryotic mycelia grown on nutrient agar. Old spruce is the percentage of spruce forest age>80 yr old within a 2‐km‐radius of the sampling point. Filled symbols refer to the plots with the higher proportion of old spruce forest in each pair, while open symbols refer to the plots with the lower proportion of old spruce forest, in the northern (circles) and southern (squares) regions, respectively. Average spore deposition rates, transformed according to maximum likelihood calculations, showed that the rates for F. pinicola in the north were never <300 spores×m −2 ×24 h −1 , regardless of the proportion of old forest in the surrounding landscape (Appendix 1). Furthermore, the spore deposition of F. pinicola was probably heavily underestimated, since almost all of the traps were hit, implying an overload of spores. In contrast, spore deposition in the south never exceeded 126 spores×m −2 ×24 h −1 , and in several sites it was only 7 spores×m −2 ×24 h −1 or less (Appendix 1). Spore deposition of F. rosea in the north ranged between 6 and >390 spores×m −2 ×24 h −1 , while the deposition in the south was never >47 spores×m −2 ×24 h −1 . In fact, no spores were trapped at four of the southern sites. Deposition of T. laricinum in the north was generally low compared to the other studied species in that region. No spores were trapped at five of the sites, while the deposition was as high as 93 spores×m −2 ×24 h −1 at one site (Appendix 1). Deposition of G. taxicola in the north was generally high, and always exceeded 95 spores×m −2 ×24 h −1 , except on one occasion (Appendix 1). Phlebia centrifuga in the south displayed the lowest spore deposition rate of all species. Spores of this species were only trapped in four of the 12 landscapes, at rates ranging from 7 to 31 m −2 ×24 h −1 . Effects of scale and forest age – wood‐discs Both F. rosea and P. centrifuga showed a strong relationship between the amount of old forest in the surrounding landscape and the spore deposition in plots with higher proportions of old spruce forest ( Fig. 4 ). For F. rosea significant linear regression models were found for spruce forest >110 and 140 yr old within a 2‐km radius of the sampling locations and for spruce forest older than 140 yr within a 3‐km radius ( Table 2 ). In addition, the amount of the variation in spore deposition explained (adj. R 2 ), increased with increasing scale and forest age. For spruce forest older than 140 yr within a 3‐km radius the amount of variation explained was as high as 91% (adj. R 2 ) ( Table 2 , Fig. 4 ). When increasing the spatial scale, the enlargement in old forest area (>140 yr) significantly contributed to improve the partial correlation coefficient, both when controlling for 1 km (p=0.029) and for 1–2 km (p=0.003). Phlebia centrifuga showed a significant relationship between spore deposition and amount of old forest at all three scales (1, 2 and 3 km radius) ( Table 2 ). The amount of the variation in spore deposition explained (adj. R 2 ) increased with increasing scale and forest age, and for spruce forests older than 140 yr within a 3 km radius, the amount explained was 94% ( Table 2 , Fig. 4 ). In a partial correlation the contribution in larger old forest area (>140 yr) when increasing the spatial scale significantly contributed to the model when controlling for both 1 km (p=0.013) and 1–2 km (p=0.012). 4 Spore deposition of Fomitopsis rosea and Phlebia centrifuga (presented as spores per m 2 and 24 h) plotted against percentage spruce forest>140 yr old for plots with lower proportion of old spruce forest (open circles, n=7) and higher proportion of old spruce forest (black circles, n=7), respectively. Based on data obtained from wood‐discs in the northern region. 2 Results from a simple linear regression with spore deposition, measured by the wood‐disc method, as dependent variable. P‐values and adjusted R 2 values are given for different proportions of spruce forest (>80, >110 and >140 yr) at different scales (1‐, 2‐ and 3‐km radius of the plot centre) both for plots with higher proportion of old forest (n=7) and lower proportion of old forest (n=7). Species Plots with higher amounts of old forest Plots with lower amounts of old forest 1 km 2 km 3 km 1 km 2 km 3 km age p R 2 p R 2 p R 2 p R 2 p R 2 p R 2 F. rosea 80 0.281 – 0.055 – 0.055 – 0.382 – 0.181 – 0.071 – 110 0.059 – 0.031 57 0.051 – 0.620 – 0.255 – 0.095 – 140 0.102 – 0.004 80 0.000 91 0.424 – 0.271 – 0.105 – P. centrifuga 80 0.186 – 0.029 58 0.027 59 0.971 – 0.324 – 0.046 50 110 0.040 52 0.011 71 0.018 65 0.966 – 0.329 – 0.022 62 140 0.184 – 0.005 79 0.000 94 0.565 – 0.196 – 0.046 50 No significant relationships between spore deposition of F. rosea and the proportion of old forests (>80, >110 and>140 yr) were found at any scale in the plots with lower proportions of old forest. For P. centrifuga , on the other hand, the spore deposition was significantly correlated to all three forest age classes within a 3 km radius (p<0.05) ( Table 2 ). Discussion Effects of forestry For both F. rosea and F. pinicola there was a large difference in spore deposition in the paired forest landscapes (2 km radius) between those with relatively high and those with relatively low proportions of old spruce forest (>80 yr). Since the distance between the plot‐centres within pairs ranged between only 6 and 10 km, the results clearly illustrate the importance of dispersal over the range of a few kilometres, i.e. moderate dispersal distances. This is not surprising, since a number of studies have shown that the majority of spores fall very close to the fruiting bodies (e.g. Malloch and Blackwell 1992 , Möykkynen et al. 1997 ). The findings also demonstrate that forestry, even at a local scale, may have a significant effect on the dispersal of species. This, in turn, affects the distribution of fungi at a regional scale where the species may have a more restricted occurrence. For example, Siitonen et al. (2001) found lower abundance of red‐listed polypores in fragmented old‐growth forests in eastern Finland than in comparable, continuous old‐growth forests in Russian Karelia. They suggested that the difference was due to limitations in the dispersal and colonisation ability of the fungi in the fragmented forest landscape. Of particular interest is that F. pinicola , which is a common generalist fungus and is not considered threatened by forestry, displayed lower spore deposition levels in landscapes with lower proportions of old (>80 yr) spruce forests in the southern region. Thus, spore availability appears to be highly correlated to the amount of old forest nearby, even for common species. The extent to which this lower spore availability may affect the viability of the populations remains unclear. However, our data suggests that fragmentation and habitat loss may affect common species to an extent that is not fully understood and deserves further investigations. Importance of forest age The spore deposition was strongly correlated with the amount and age of spruce forest for most of our studied species, suggesting that they are favoured by structures within old forests. Old spruce forests generally contain higher volumes of CWD than younger forests ( Fridman and Walheim 2000 ), which promotes larger populations of wood‐inhabiting fungi. However, other variables connected to forest age may also be involved, such as the quality and continuity of the CWD. For example, several red‐listed species of wood‐decaying fungi show preferences for coarse logs, which are more common in old forests ( Renvall 1995 , Bader et al. 1995 , Kruys et al. 1999 , Sippola et al. 2001 ). However, no relationship between the amount of old forest and spore deposition on wood‐discs was found for F. rosea in plots with low proportions of old forest. This may reflect differences between the species, but it is more likely to be due to the test being insufficiently sensitive, because of the low amount of old forest (>140 yr) in these plots. The average amount was only 2.2 ha in a 1257 ha plot ( Fig. 2 ). Importance of scale For both F. rosea and P. centrifuga , the correlation between the rate of spore deposition and the proportion of old spruce forest was very high for the area located within a 3‐km radius of the measuring point, but decreased at 2‐km and 1‐km radii. Although 3 km was the largest radius tested, and the contribution of sources from larger distances is unknown, the high correlation between the proportion of old forest within 3 km and spore deposition (adj. R 2 >90%) indicates that the trapped spores mainly originated from populations within this distance. This is in accordance with data presented by Nordén and Larsson (2000) , who sampled spores at different distances from fruit bodies of P. centrifuga. They found that the majority of the spores were deposited in close vicinity to the fruit bodies, but also concluded that the deposition rate was still fairly high at a distance of 1 km, which was the largest distance tested. However, a small fraction of the spore deposition in the present study probably constitutes background spore deposition, originating from distant spore sources. Spore deposition curves usually show a steep and exponential decrease with distance, leaving a long tail of low deposition at large distances. For instance, Kallio (1970) showed that 99.9% of the spores released from the root rot fungus Heterobasidion annosum settled within 100 m of the fruit body, but also concluded that some spores of the species spread as far as 70–500 km. Spore deposition amounts For most of the species we found fairly high rates of spore deposition, suggesting that the availability of spores for colonisation is fairly good at a local scale. For example, the spore deposition of the red‐listed T. laricinum , which had by far the lowest deposition rate of the tested species in the north, was still >10 spores m −2 24 h −1 at half of the sampling sites. Fairly high spore deposition rates were also found for the threatened fungi F. rosea and P. centrifuga in several sites in the southern region, where the impact of forestry has been high for several hundred years. This indicates that factors other than dispersal may be important in the colonisation process of spruce logs, such as various aspects of wood quality, including log size, decay stage ( Bader et al. 1995 , Kruys et al. 1999 ) and wood density. Although spore depositions were fairly high overall, the relative differences in deposition between forest landscapes with differing intensities of forestry were also high for several species. There were also considerable differences in deposition between the northern and southern regions. For example, spore deposition of F. rosea was on average ca 10 times higher in the north compared to the south. There was also a large relative difference in spore deposition between species. Although the spore deposition of F. pinicola was highly underestimated in the north, ca 10 times more spores were trapped compared to F. rosea in the landscapes with lower amounts of old forest. Since factors like germination requirements and competition from other mycelia probably also affect colonisation ability ( Stenlid 1993 ), the spatial and temporal window available for colonisation is most likely very small. Thus, dispersal limitation may still be a significant factor in the population dynamics of these fungi, since insufficient densities of spores may be present for successful establishment of mycelia and dikaryotisation. Conclusions Our data strongly suggest that the landscape composition has strong, indirect effects on the availability of spores of wood‐decaying fungi. Both the amount and age of old spruce forests are key factors in predicting the rates of spore deposition. Increases in the intensity of forestry tend to reduce the proportion of old (>80 yr) spruce forests, which in turn negatively affects the fungi and reduces the abundance of their spores. The strong correlation between the rates of spore deposition of F. rosea and P. centrifuga and the proportion of old (>140‐yr) spruce forest within a 3 km radius indicate that this may be a relevant scale at which to consider the needs of these species in conservation and management planning. Furthermore, it emphasizes the importance of very old forests (>140 yr) to wood‐decaying fungi. The age of these stands exceeds the common harvest rotation period, which is ca 80–120 yr, and consequently the ongoing decline in old‐growth forests clearly constitutes a risk for the long‐term survival of these species in the boreal forest landscape. However, the significance of old forests is probably connected to the abundance of CWD, and increasing the CWD volumes in younger stands may enhance the possibility for wood‐decaying species to survive also in the managed forest landscape. Acknowledgements We thank David Rönnblom, Lars Persson, Christer Ahlenius and Mats Englund at Holmen Skog, Kjell‐Ove Häggström at Sveaskog, Eva Ståhl and Bo Ernstsson at AssiDomän, Urban Ekskär at SCA and The Local Board of Forestry (Skogsvårdsstyrelsen) for providing forest stand maps and valuable information on different forest sites. We also thank Kristoffer and Eva Hylander for help with maximum likelihood calculations of spore deposition data. This study has been financed by The Science Research Council (VR); grants to L. Ericson and The Swedish Council for Forestry and Agricultural Research (SJFR); grants to J. Stenlid.

Journal

Ecography – Wiley

Published: Feb 1, 2004

There are no references for this article.