Naturalness of Quercus robur stands in Latvia, estimated by structure, species, and processes/ Hariliku tamme Quercus robur puistute looduslikkus Latis, hinnatuna struktuuri, liikide ja arengu jargi.
Due to intensive forest harvest and exploitation for agriculture, all Western European forests of the temperate zone are to a great extent disturbed by man (Peterken, 1996). Very few fragments of the past virgin forest remain (Jones, 1945), causing a deficiency of reference areas to study structure and disturbance dynamics. In northern European forests in the temperate--boreal transition zone, human activities were responsible for a decline of broad-leaved forests and expansion of spruce (Bradshaw & Hannon, 1992; Lindbladh et al., 2000; Niklasson et al., 2002), particularly in the past 300 years (Lindbladh & Foster, 2010). The largest old-growth broad-leaved (Quercus robur, Fraxinus excelsior, Ulmus spp., Tilia cordata, Acer spp.) woodland in lowland Europe is the Bialowieza Primeval Forest on the Polish and Belarusian border (Falinski, 1986). However, even there the territory suffered from felling and overgrazing due to use as a hunting reserve in the late 1800s-1900s (Falinski, 1988), which caused an expansion of Picea abies (Mitchell & Cole, 1998).
The coverage of broad-leaved species in Latvia reached a maximum in the Atlantic period about 6000 years ago (Zunde, 1999). Since that time, climate cooling and human impact caused a decline of broad-leaved forest in Latvia, particularly in the 17th and 18th centuries to supply growing export of Q. robur for ship-building and slash and burn clearage of forest by peasants (Dumpe, 1999; Liepina, 1999). In northern Latvia, the total forest area dropped from 59.2% in 1710 to 19.4% in 1914 (Vasilevskis, 2007). In 1924, broad-leaved forest occupied only 2.3 thousand ha (0.2% of forests) in Latvia (Matiss, 1987). However, by 2008, the forest area had increased to 53%, of which broad-leaved forests contributed 1.1% (data from the Latvian State Forest Register 2009). The Q. robur stands in Latvia are presently mostly small (< 2 ha) and highly fragmented (Zunde, 1999).
It has been argued that the Q. robur primeval forest in lowland Europe formed an open landscape, due to grazing by large herbivores (Vera, 2000). According to this view, wooded meadows and pastures in the rural landscape today might resemble the forests of pre-industrial forestry. In contradiction to the grazing hypothesis, some palaeoecological evidence suggests that the natural Q. robur woodland that existed prior to human settlement formed a closed canopy, with or without grazers (Mitchell, 2005). Thus, knowing that regeneration of Q. robur is limited by light availability caused by competition with vegetation (Humphrey & Swaine, 1997; Kussner, 2003; Harmer & Morgan, 2007), natural disturbances such as fire, wind, and water-logging probably created gaps favouring regeneration (Bradshaw et al., 2003; Bradshaw & Hannon, 2004; Whitehouse & Smith, 2004). Regeneration of Q. robur and other broad-leaved species can be successful in floodplains (Kussner, 2003; Dobrowolska, 2008) and under conifer canopies, provided sufficiently lit conditions free from competitive vegetation (Gotmark et al., 2005; Dobrowolska, 2006; Goris et al., 2007). In Poland (Falinski, 1986) and Russia (Nfesterovs, 1954), Q. robur mostly occurs in mixed woods with a closed canopy, and the tree species in mixed stands with Q. robur differ depending on the growth conditions. The controversy regarding whether Q. robur woods naturally occurred as open of closed woodland may never be fully resolved, but in Eastern Europe there is no evidence that an open landscape with oak existed before human settlement.
In Latvia, open parkland landscapes were created by slash and burn agriculture, whereby scattered Q. robur trees survived within tilled land, hayfields, and pastures (Dumpe, 1999). Regeneration of Q. robur on agricultural land can be successful in patches of unpalatable or spiny vegetation unfavoured by large herbivores, which can provide an explanation for long-term persistence of Q. robur in an open landscape (van Uytvanck et al., 2008). As pasture woodlands are of high conservation value in Europe, partial harvesting has been suggested to create more open stands (Gotmark, 2007; 0kland et al., 2008; Paltto et al., 2008). Also in Latvia, Q. robur trees in protected areas are often cleared from surrounding trees and shrubs to maintain their value for biological diversity (Anon., 2008), creating a setting that might have been typical under a traditional shifting agricultural regime.
Knowledge of the naturalness of Q. robur forests in Latvia could be used to guide conservation management, particularly regarding possible restoration measures. The naturalness of forests can be estimated using criteria within three dimensions: structures, species, and processes (Brumelis et al., 2011b). The aim of the present study was to determine the level of naturalness of Q. robur stands in Latvia, based on the age and size structure, amounts of dead wood, cut stumps, archival inventory data and maps, as well as richness of Woodland Key Habitat (WKH) indicator species (Ek et al., 2002). Considering the history of logging and conversion of forests to farmland in Latvia, we might expect that the naturalness of Q. robur forests is low. Of the 6554 ha of forest area dominated (over 50% relative wood volume) by Q. robur, 14.5% (953 ha) has age over 150 years and only 2.3% (147 ha) is over 200 years old (State Forest Service, 2008). A large part (803 ha) of these forests are WKHs, which in Latvia are identified based on the presence of species that cannot persist under industrial forestry and on the amounts of structural elements, such as coarse wood debris. Considering the legacy of forest harvesting in Latvia, we hypothesize that the structural features of naturalness of Q. robur WKHs, which might be considered to be among the most natural Q. robur forests, have developed over a relatively short period of time. Further, part of the oak forests today are probably secondary forests on abandoned agricultural land.
MATERIAL AND METHODS
Latvia is located in the boreo-nemoral zone, where Q. robur can occur in mixed forest together with coniferous boreal species (Sjors, 1963). The climate is moderate continental with a mean temperature of - 5.3[degrees]C in January and 14.8[degrees]C in July, and precipitation of 700-800 mm, of which about 500 mm falls in the warm period (Central Statistical Bureau of Latvia). Moraine relief dominates with sandy clays and clay sands (Nikodemus et al., 2008). The climate is more continental towards the east.
Stands of Q. robur were chosen from the State Forest Register based on dominance of Q. robur (at least 50% of total wood volume), Q. robur age (> 120 years), stand size (> 2 ha), and WKH criteria (Ek et al., 2002). It was considered that the above criteria would select a subset of stands that might be expected to be among the most natural. From the register, seven stands were selected without prior visit to the site (Fig. 1). These forest stands were Audile (area 4.2 ha), Pededze (7.9 ha), Barkava (11.7 ha), Salenieki (5.2 ha), Kinguru (6.5 ha), Mezotne (18.1 ha), and Rauda (2.9 ha). Four Q. robur stands (Audile, Pededze, Barkava, and Salenieki) were chosen in the Lubana Lowland, central eastern Latvia, as it was generally considered that this area might support old relict stands. In the early 1700s, the largest part of the remaining uncut Q. robur stands were found in the Lubana area, but even so, most of the stands along rivers had been harvested and wood floated by rafts to Riga (Zunde, 1999). The Lubana Lowland had been subject to major floods and the water levels in Lake Lubana (then about 90 km2) could in spring rise by 2-4 m, flooding an area of over 600 [km.sup.2] (Skinkis, 1998). Drainage in the area began in the 1850s, and continued to the 1930s, but with little effect on flooding. All the stands in the Lubana area are located in the previous flood zone of the lake. After a major drainage project, begun in the 1950s, the lake covered about 80 [km.sup.2], and the surrounding forests were much less subject to floods.
[FIGURE 1 OMITTED]
The other three stands, Kinguru, Mezotne, and Rauda, were located in western, central, and south-eastern Latvia, respectively. Except one stand, Kinguru, the sampled forest stands were protected as nature reserves or microreserves. The protected areas were established for conservation of biologically valuable forest habitats. The conservation values of the Mezotne forest, protected as a micro-reserve, are a natural broad-leaved forest and populations of Dendrocopos medius and Osmoderma eremita.
In the studied forest stands in the Lubana area (Audile, Pededze, Barkava, and Salenieki), Q. robur and F. excelsior were dominant species in the tree layer mixed with Alnus glutinosa. Tilia cordata was abundant only in Salenieki. A different pattern of regeneration was observed in the other stands. Picea abies was abundant in the sapling layer at Kinguru and Rauda, and scattered spruce had also entered the upper canopy of the former stand. The Mezotne stand generally lacked a sapling layer. Vegetation in the stands was dominated, in variable composition and cover, by typical nemoral species, such as Paris quadrifolia, Galeobdolon luteum, Pulmonaria officinalis, Anemone nemorosa, and Glechoma hederacea. The Rauda and Kinguru stands supported also boreal herbs, such as Oxalis acetosella. A dense Rhamnus cathartica shrub layer occurred in the Salenieki plot.
Sampling methods and data analyses
One sample plot was established in each of the seven selected Q. robur stands in 2007-2008. The plots were located in the centre of the stands to avoid edge disturbance. Plot size was 20 m x 50 m (0.1 ha). This plot size was considered to be representative of the stand, as visual inspection of the stands suggested similar forest structure at this scale, except at edges.
Stem height was measured for all trees with height over 10 m. Smaller trees were counted in height classes (<1.0 m, 1.1-2.0 m, 2.1-5.0 m, and > 5.0 m). For trees over 10 m in height, cores were removed 0.5 m above tree base with an increment borer (at 1-m height for oak due to the physical difficulty of coring this species). Trees were cored below breast height to obtain a better estimate of true tree age.
A total of 301 trees were measured and from these cores were obtained. The number of tree cores used in age structure analysis was 52 in Audile, 54 in Barkava, 37 in Pededze, 37 in Salenieki, 60 in Kinguru, 48 in Rauda, and only 13 in Mezotne. In Mezotne, trees were sparse with wide canopies. Betula pendula, Populus tremula, and Alnus incana were excluded from the age structure analysis due to their very small number in plots (< 2 in each plot except in Barkava).
A search was made for WKH indicator epiphyte species on cored trees in the plots. Because the epiphytic species composition varied considerably between trees in the selected stands, in addition a search for WKH indicator epiphyte species was made in the entire stands and recorded as present when found. The search was conducted in each stand for approximately 1 h.
Tree cores were glued in mounting boards and sanded to a fine polish. The cores were then scanned and tree rings were measured with Lignovision 1.37 (RINNTECH). If the pith was missed, the age was corrected by adding the number of missing years by estimation based on ring curvature and mean ring width of the innermost 10 increments. In some cases of extensive wood rot around the pith, the distance from the innermost ring to the geometric tree centre was estimated by subtraction of the core length from the tree radius. The number of missing rings was extrapolated based on the mean ring width of the oldest visible 10 increments. This estimation method assumes symmetry of tree rings, which will rarely occur, but this inaccurate method was considered more suitable than omitting the trees from analysis.
The volume of living trees was estimated using volume tables for the respective species (Sacenieks & Matuzans, 1964). For fallen logs over 10 cm in diameter at mid length, the volume was calculated as cylinders using log length and diameter in the middle. Dead standing tree volume was estimated from volume tables for the respective species (Sacenieks & Matuzans, 1964). Cut stumps were counted.
Archival forest inventory data for the stands was obtained from forest management plans stored at the Latvian State Forest Research Institute 'Silava'. For each tree species, the mean tree age in the canopy layer was given as well as the relative timber volume in 10% classes. If a tree species formed more than one canopy layer, the volume and age were given for each cohort.
Archival inventory records
The inventory maps and records for the stands had been produced in different years from 1924 to 1937 (Table 1). The maps indicated that the Kinguru and Rauda stands were open with scattered trees, while the others were shown as closed forest. Ages of Q. robur ranged from 70 to 150 years, with the oldest in Mezotne. Quercus robur dominated in all stands (90-100% of the relative timber volume), except in the Salenieki stand where in 1927 it was reported as absent; the dominant trees there were Betula pendula and Populus tremula. Along with the dominant Q. robur, in 1937 the Barkava stand also contained B. pendula, P. tremula, and A. glutinosa, which were recorded as contributing each less than 10% of the total volume. At Kinguru in 1927, an understorey of B. pendula and P. tremula mixed with P. abies, with age about 15 years, was recorded.
Size and age structure of living trees
The maximum tree height in the plots was 30-36 m. In all plots the upper canopy over 25 m in height (> 20 m in the Barkava plot) was dominated by Q. robur (Fig. 2), mixed with F. excelsior and Ulmus glabra in the Audile site (Fig. 2). While Q. robur was abundant in all plots in the below 1 m height class and in the upper canopy, it was in low numbers or missing in other height classes (Table 2, Fig. 2). The understorey tree layer at 1-10 m height (Fig. 1) was dominated by T. cordata in the Salenieki plot, F. excelsior in the other sites, with also U. glabra at Audile and B. pendula, P. tremula (not shown), and A. glutinosa at Barkava. The successional understorey species was P. abies at Kinguru and Rauda. At Mezotne, F. excelsior formed a subcanopy layer and was abundant together with Q. robur in the below 1 m layer, but otherwise the stand lacked a tall sapling stage.
[FIGURE 2 OMITTED]
The maximum Q. robur ages observed in the Salenieki, Kinguru, and Rauda plots were low (140, 170, and 171 years, respectively) compared to those in the other plots. The maximum age of Q. robur was over 200 years in three plots (Fig. 3), with a maximum age of 242 years in Mezotne. Most (80%) of the tree cores passed close to the pith and thus the age could be estimated fairly accurately. Only four cored trees had rot near the core, resulting in very approximate ages, but these were not the oldest trees, as suggested by size, and thus most likely did not affect the above-mentioned maximum ages.
[FIGURE 3 OMITTED]
In the Audile plot, some F. excelsior and A. glutinosa trees were over 100 years old. Most of the A. glutinosa, U. glabra, and F. excelsior in the plots had ages of 41 to 80 years; a small proportion of the trees were in the 21-40 year age class (Fig. 3). In the Salenieki plot, T. cordata was abundant in the 21-80 year age classes. A peak of P. abies regeneration occurred in about 1930-1950 in the Kinguru and Rauda sites, corresponding to the 61-80 year age class (Fig. 3). Betula was common in the 41-80 year classes at Barkava (not shown).
The highest total dead wood volume (Table 3) was at Audile (198 m3) and the lowest at Barkava (109 [m.sup.3]). In all plots the dead wood volume was comprised mostly of Q. robur (over 100 [m.sup.3] [ha.sup.-1]). The dead Q. robur volume mostly consisted of large, over 30 cm diameter stems (not shown). Significant proportions of dead wood volume were also contributed by A. glutinosa (43 [m.sup.3] [ha.sup.-1]) at Audile and P. abies (52 [m.sup.3] [ha.sup.-1]) at Kinguru. The ratio of dead to living wood volume ranged from 0.10 at Mezotne to 0.38 at Audile. Cut stumps, all of which were Q. robur, were observed in Barkava, and in greater numbers at Kinguru and Mezotne. These three plots, and also Audile, lacked standing dead wood (Table 3).
All the recorded WKH indicator species were found on Q. robur, except for Lecanactis abietina, which was found on P. abies. The most frequent indicator species found in the plots were Arthonia spadicea, A. byssacea, and Homalia trichomanoides (Table 4). No indicator bryophyte spcecies were found at Salenieki. The highest numbers of species occurred in the Pededze stand, followed by Kinguru, Audile, and Rauda. Of the recorded indicator species, six are protected in Latvia: Arthonia spadicea, A. byssacea, A. vinosa, A. leucopellea, Lobaria pulmonaria, and Dicranum viride.
Quercus robur is a typical early successional species that can invade disturbed habitats such as gaps, forest edges, and grassland (Lawesson & Oksanen, 2002). Moreover, Q. robur can persist for multiple generations in a closed canopy at a stand level (Mitchell, 2005), probably due to natural disturbances that create gaps (Bradshaw et al., 2003; Bradshaw & Hannon, 2004; Whitehouse & Smith, 2004).
It might be expected that a natural Q. robur forest would contain structures such as old large trees close to their maximal age, abundant dead wood, and a sapling layer of shade tolerant species, but would lack cut stumps. In the relatively undisturbed mesic deciduous forest in the Bialowieza reserve in Poland, Q. robur reaches 400 years of age, occurs mostly in mixed woods with other broad-leaved tree species, and coarse dead wood amounts are mostly over 100 m3 ha-1 (Falinski, 1988; Bobiec, 2002).
The presence of cut stumps in Barkava, Kinguru, and Mezotne indicates past removal of dead or damaged trees or selective cutting. However, the lack of standing dead wood in these plots suggests that the former was more likely. Despite the evidence of past wood removal, the densities of canopy trees are in the range of those reported for old-growth forests in the nemoral and borenemoral zones (Nilsson et al., 2003). Also, the volumes of dead wood at the sites (75-198 [m.sup.3] [ha.sup.-1]) correspond to, or are close to, volumes (>100 [m.sup.3] [ha.sup.-1]) in old-growth mesic deciduous and riparian forests in Poland (Bobiec, 2002). It has been estimated (Nilsson et al., 2003) that a volume of 130-150 [m.sup.3] [ha.sup.-1] dead wood was common in productive European forest before human exploitation. The mean volume of dead wood in eutrophic bore-nemoral stands in Estonia was estimated to be 198 [m.sup.3] [ha.sup.-1] (Lohmus & Kraut, 2010). Thus, while the dead wood amounts accumulated in the studied Q. robur forests are approaching or have reached those in natural nemoral forest, the lack of standing dead wood in four plots indicates that the level of naturalness in terms of structures is far from that in old-growth forest.
Considering that Q. robur can attain an age of 400-500 years (Falinski, 1986) or even 800 years (Jones, 1959), the sampled stands in Latvia are certainly not old-growth forests. The oldest trees in the plots in the Lubana area (Audile, Pededze, and Barkava) had ages just over 200 years, with origin in about 1800 to 1810. Among the seven studied stands, in one the age of the oldest tree was less than 140 years, and in three stands less than 250 years. Considering the history of forest harvest and conversion to agricultural use in Latvia (Terauds et al., 2011), these stands most likely regenerated naturally after logging or on agricultural land, although natural catastrophic disturbance such as fire cannot be ruled out. Water-logging, which was common in the Lubana area in spring at that time, might have facilitated recruitment of Q. robur (Bradshaw & Hannon, 2004). Forest inventory conducted in the early 1700s indicated that, while Q. robur woodland had already suffered major depletion over much of Latvia, it was abundant along the Pededze River, with stems reaching 1.4 m DBH (Zunde, 1999). However, these stands were close to rivers and thus were most likely prime harvest targets at a time when Quercus forests were already extremely depleted.
The inventory of the Barkava stand in 1937 reported a dominant Q. robur overstorey with a minor component of P. tremula, B. pendula, and A. glutinosa with ages of 70-90 years. The same species are seen in the stand today, but there are no P. tremula, B. pendula, and A. glutinosa over 80 years of age. Considering the cut stumps at Barkava, the continued presence of early successional P. tremula and B. pendula might be due to regeneration in gaps created by natural mortality or perhaps selective cutting. In contrast to the other six stands, stand inventory data from 1927 for the youngest (Salenieki) stand indicated a mixed P. tremula and B. pendula forest without Q. robur. In this stand, the age structure of Q. robur today (Fig. 3) suggests that most of the trees in 1927 were 1-20 years of age, and hence their wood volume would not have been recorded in the inventory. The P. tremula and B. pendula were most likely cut, releasing the Q. robur understorey. Planting or sowing of Q. robur was unlikely in the pre-1930 period, as natural regeneration of clearcuts was predominantly used, followed by sowing of pine (Terauds et al., 2011). The Kinguru and Rauda stands were described in 1926/1929 as having scattered Q. robur about 70-80 years old, with an understorey of P. tremula, B. pendula, and P. abies in the former. In these sites, the open canopy might have been due to previous use as pasture, as was common in Latvia at that time (Dumpe, 1999).
Despite the large numbers of stems of less than 1 m height, recruitment of Q. robur to taller height classes after the post-initiation stage of the canopy has been lacking. In addition, the age structure of plots (Fig. 3) shows that in the Barkava plot, and partly also at Pededze, Kinguru, and Rauda, there was a gap in, or minimal, regeneration of all tree species between about 1870 and 1930, corresponding to the 81-140 year age classes (Fig. 3). A shorter 20-year period of no regeneration occurred at Audile in 1890-1910, and a longer period in Mezotne from 1830 to 1950. These periods might be related to human disturbance to the understorey, for example, by livestock grazing. The method used for identifying overgrowing meadow and pasture woodland of European importance in Latvia is partly based on an observation of a gap in the regeneration of tree species (Larmanis, 2010). It is argued that if a period of no regeneration is observed, then the canopy was more open, i.e., grazed. On the other hand, time gaps in cohort structure can result from poor light availability caused by competition (Harmer et al., 2001; Kussner, 2003). Shading might also explain the succession to relatively more shade tolerant species, such as P. abies (Kinguru and Rauda), F. excelsior (Audile, Pededze, and Barkava), and T. cordata (Salenieki), as reported previously for the Moricsala Reserve in Latvia (Brumelis et al., 2011a). Succession of mixed Q. robur stands to species-rich spruce--deciduous forest has occurred also on Abruka Island in West Estonia (Meikar et al., 2004). Thus, the age structure in the plots, except for Mezotne, probably shows a pattern of natural succession since stand intitiation, with shade tolerant species being more successful than Q. robur. In the stands in the Lubana area, the Q. robur dominated canopy might be expected to change to a mixed broad-leaved forest in the future, as is common in alluvial flood plains, or perhaps to a P. abies successional stage (Falinski, 1986). In the Republic of Bashkortorstan and South-Western Russia, natural succession of Q. robur forests to a mixed canopy is common, but in the long-term a return of Q. robur can occur in large canopy gaps (Nesterovs, 1954).
The Mezotne stand lacked evidence of past regeneration of trees in the understorey over a 120-year period, it contained the least amounts of dead wood (75 [m.sup.3] [ha.sup.-1]) and there were numerous stumps (60 [ha.sup.-1]). This long period of no regeneration cannot be explained solely by a period of canopy closure, and suggests long-term management as an open park forest with or without pasture use. The stand is located in an old rural area and is surrounded by agricultural land, which is consistent with this idea. The present lack of a sapling/shrub layer might explain the copious numbers of Q. robur (22 250 ind. [ha.sup.-1]) and F. excelsior (16 500 ind. [ha.sup.-1]) in the lower than 1 m layer.
Of the recorded eight lichen species, six (except Bacidia rubella and Lecanactis abietina) are protected. However, only one (Dicranum viride) of the five WKH indicator bryophyte species is protected. In a study of epiphytes in nemoral forests of Latvia, in which more stringent criteria of naturalness and a larger number of stands were used (Mezaka et al., 2008), of the species listed as WKH indicator species (Ek et al., 2002), 12 lichens and 7 bryophytes were recorded, of which respectively 8 and 5 were found on Q. robur. Thus, in terms of indicator species numbers, the stands sampled in our study are likely representative of the set of most natural oak forest in Latvia, and we are confident that the search for species in the plots and stands was sufficient. However, we recorded WKH indicator species only, and full species lists would provide additional information on epiphyte diversity in the stands. Even though all the studied stands were relatively young, each contained at least one protected species. The largest number (5) of protected species occurred in Pededze. The species richness of these epiphytes did not appear to be related to stand age, as even the youngest stands (Kinguru and Rauda) supported six and five WKH indicator species, respectively, while the highest richness, observed at Pededze, was eight species. Of the lichen species found in the studied stands, Arthonia leucopellea, A. spadicea, A. vinosa, and Bacidia rubella have been reported as frequent on Q. robur in wooded meadows in Estonia (Leppik & Juriado, 2008) and may be limited by availability of large diameter substrate rather than human disturbance. Arthonia byssacea is rare in Estonia (Thor et al., 2010) and has been found to be associated with old large diameter Q. robur (Juriado et al., 2009). However, we found A. byssacea in five stands, including in the youngest (Salenieki). Perhaps this species requires more shaded conditions in closed forests, as found in the Latvian Q. robur stands compared to the Estonian wooded meadows.
Lobaria pulmonaria and Neckera pennata, recorded each in one stand in the present study, are dependent on habitat connectivity (Snall et al., 2004, 2005; Paltto et al., 2006). Both of these species readily utilize other deciduous species, particularly P. tremula, as a substrate and are probably not limited by presence of Q. robur in the landscape. Homalia trichomanoides, Isothecium alopecuroides, and also Anomodon longifolius (we did not identify Anomodon to species) have been observed to be limited by habitat quality features, such as moisture, substrate pH, tree size, and forest age, and also by present and/or past habitat connectivity (Lobel et al, 2006; Lobel & Rydin, 2009).
Some of the stands may have had past agricultural use (grazing, meadows). Traditionally managed wooded meadows and pasture usually contain old legacy trees. The age of the oldest Q. robur tree in a traditionally managed woodland on Saaremaa Island, Estonia, was estimated to be about 500 years (Laanelaid et al., 2008). This raises the question to what extent the present occurrence of species on oak is due to past non-forest land-use. As pasture woodlands have high conservation value in Europe, partial harvesting has been recommended for generating more open stands (Gotmark, 2007; Okland et al., 2008; Paltto et al., 2008). Also in Latvia, Q. robur trees in protected areas are often cleared from surrounding trees and shrubs to maintain their value for biological diversity (Anon., 2008), creating a setting that might have been typical of that under a traditional shifting agricultural regime. However, none of the studied sites contained legacy trees of a meadow setting, i.e., they might be secondary forests on past agricultural land but they do not represent traditional long-term wooded meadow/pasture management. An exception might be the Mezotne stand. This stand also supports an Osmoderma eremita population. This rare beetle species has a small dispersal distance and requires large diameter and hollow Q. robur that are well lit (Ranius & Hedin, 2001). Therefore, management by shrub cutting might be warranted when required.
The studied Q. robur WKH stands, which can be considered to be among the most natural stands of this forest type in Latvia, were found to be mature (in terms of cutting age, which is 100-120 years for oak in Latvia) to overmature Q. robur woodland developed under minimal to moderate human disturbance. Similarly, in southern Sweden, WKHs have not escaped human disturbance, but nevertheless are among the most natural stands in the region (Ericsson et al., 2005; Jonsson et al., 2009). While the amounts of dead wood were in the range of those reported for old-growth forest, or approaching these, there were no very old Q. robur close to its maximal age, and standing dead wood was lacking in four stands. Nevertheless, protected species were found in the stands. The small size of the stands (2.9-18.1 ha) suggests that the full range of variability of pre-industrial broad-leaved forest cannot be expected, either now or in the future. In the studied stands, considering the long age span of Q. robur, we predict that the Q. robur canopy will persist for at least another 200 years, and certainly the conservation value of the stands will increase with time. The studied Q. robur stands are among the most natural in Latvia, but assessment based on the age and size structure, amounts of dead wood, and past evidence of cutting and possibly agricultural use suggests a rather low level of naturalness. Nevertheless, the richness of WKH indicator species was in some cases high.
The study was supported by the European Social Fund project 'Support for the Daugavpils University doctoral study programme'. We acknowledge the help of Janis Sarmulis, Andris Sarmulis, Didzis Tjarve, and Anna Mezaka during fieldwork. The help of Juris Zarins in accessing inventory records at the Latvian State Forest Research Institute 'Silava' is appreciated. The anonymous comments received during the review process much helped to improve the manuscript.
Anonymous. 2008. LIFE project Protection and Management of the Northern Gauja Valley. Technical Final Report (LIFE03 NAT/LV/00082). http://www.zgauja.lv/images/stories/parzg/ faili/final_report_technical_250308.pdf (visited 2011-07-04).
Bobiec, A. 2002. Living stands and dead wood in the Bialowieza forest: suggestions for restoration management. For. Ecol. Manag., 165, 125-140.
Bradshaw, R. & Hannon, G. 1992. Climatic change, human influence and disturbance regime in the control of vegetation dynamics within Fiby Forest, Sweden. J. Ecol., 80, 625-632.
Bradshaw, R. & Hannon, G. 2004. The Holocene structure of north-west European temperate forest induced from paleoecological data. In Forest Biodiversity: Lessons from History for Conservation (Honnay, K., Verheyen, K., Bossuyt, B. & Hermy, M., eds), pp. 11-25. CAB International.
Bradshaw, R., Hannon, G. E. & Lister, A. M. 2003. A long-term perspective on ungulate-vegetation interactions. For. Ecol. Manag., 181, 267-280.
Brumelis, G., Dauskane, I., Ikauniece, S., Javoisa, B., Kalviskis, K., Madzule, L., Matisons, R., Strazdina, L., Tabors, G. & Vimba, E. 2011a. Dynamics of natural hemiboreal woodland in the Moricsala Reserve, Latvia: the studies of K. R. Kupffer revisited. Scand. J. For. Res., 26(S10), 54-64.
Brumelis, G., Jonsson, B. G., Kouki, J., Kuuluvainen, T. & Shorohova, E. 2011b. Forest naturalness in northern Europe: perspectives on processes, structures and species diversity. Silva Fenn., 45(5), 807-821.
Dobrowolska, D. 2006. Oak natural regeneration and conversion processes in mixed Scots pine stands. Forestry, 79, 503-513.
Dobrowolska, D. 2008. Effect of stand density on oak regeneration in flood plain forests in Lower Silesia, Poland. Forestry, 81, 511-523.
Dumpe, L. 1999. Meza izmantosana Latvija. In Latvijas Meza Vesture lidz 1940. gadam (Strods, H., ed.), pp. 305-357. WWF-Pasaules Dabas Fonds, Riga (in Latvian).
Ek, T., Susko, U. & Auzins, R. 2002. Mezaudzu atslegas biotopu inventarizacijas metodika. Valsts meza dienests, Riga (in Latvian).
Ericsson, T. S., Berglund, H. & Ostlund, L. 2005. History and forest biodiversity of woodland key habitats in south boreal Sweden. Biol. Conserv., 122, 289-303.
Falinski, J. B. 1986. Vegetation Dynamics in Temperate Lowland Primeval Forests. Dr. Junk Publishers, Dordrecht.
Falinski, J. B. 1988. Succession, regeneration and fluctuation in the Bialowieza Forest (NE Poland). Vegetatio, 77, 115-128.
Goris, R., Kint, V., Haneca, K., Geudens, G., Beeckman, H. & Verheyen, K. 2007. Long-term dynamics in a planted conifer forest with spontaneous ingrowth of broad-leaved trees. Appl. Veg. Sci., 10, 219-228.
Gotmark, F. 2007. Careful partial harvesting in conservation stands and retention of large oaks favour oak regeneration. Biol. Conserv., 140, 349-358.
Gotmark, F., Fridman, J., Kempe, G. & Norden, B. 2005. Broadleaved tree species in conifer dominated forestry: regeneration and limitation of saplings in southern Sweden. For. Ecol. Manag., 214, 142-157.
Harmer, R. & Morgan, G. 2007. Development of Quercus robur advance regeneration following canopy reduction in an oak woodland. Forestry, 80, 137-149.
Harmer, R., Peterken, G., Kerr, G. & Poulton, P. 2001. Vegetation changes during 100 years of development of two secondary woodlands on abandoned arable land. Biol. Conserv., 101, 291-304.
Humphrey, J. W. & Swaine, M. D. 1997. Factors affecting the natural regeneration of Quercus in Scottish oakwoods. I. Competition from Pteridium aquilinum. J. Appl. Ecol., 34, 577-584.
Jones, E. W. 1945. The structure and reproduction of the virgin forest of the north temperate zone. New Phytol., 44, 130-148.
Jones, E. W. 1959. Biological flora of the British Isles: Quercus L. J. Ecol., 47, 169-222.
Jonsson, M. T., Fraver, S. & Jonsson, B. G. 2009. Forest history and the development of old-growth characteristics in fragmented boreal forests. J. Veg. Sci., 20, 91-106.
Juriado, I., Liira, J., Paal, J. & Suija, A. 2009. Tree and stand level variables influencing diversity of lichens on temperate broad-leaved trees in boreo-nemoral floodplain forests. Biodivers. Conserv., 18, 105-125.
Kussner, R. 2003. Mortality patterns of Quercus, Tilia, and Fraxinus germinants in a floodplain forest on the River Elbe, Germany. For. Ecol. Manag., 173, 37-48.
Laanelaid, A., Sohar, K. & Meikar, T. 2008. Age and present state of oaks in an oak forest in Saaremaa Island, Estonia. In EuroDendro 2008: The Long History of Wood Utilization.
News of Forest History V (39) (Grabner, M. & Eckstein, D., eds), p. 75. IUFRO. Larmanis, V. 2010. Parkveida plavas un ganibas. In Eiropas Savienibas Aizsargajamie Biotopi
Latvija. Noteiksanas Rokas Gramata (Aunins, A., ed.), pp. 176-181. Latvijas Dabas Fonds, Riga (in Latvian).
Lawesson, J. E. & Oksanen, J. 2002. Niche characteristics of Danish woody species as derived from coenoclines. J. Veg. Sci., 13, 279-290.
Leppik, E. & Juriado, I. 2008. Factors important for epiphytic lichen communities in wooded meadows of Estonia. Folia Cryptogamica Estonica, 44, 75-87.
Liepina, D. 1999. Koktirdznieciba Latvija no 1200. lidz 1940. gadam. In Latvijas Meza Vesture Lidz 1940. gadam (Strods, H., ed.), pp. 251-302. WWF-Pasaules Dabas Fonds, Riga (in Latvian).
Lindbladh, M., Bradshaw, R. & Holmqvist, B. H. 2000. Pattern and process in south Swedish forests during the last 3000 years, sensed at stand and regional scales. J. Ecol., 88, 113-128.
Lindbladh, M. & Foster, D. R. 2010. Dynamics of long-lived foundation species: the history of Quercus in southern Scandinavia. J. Ecol., 98, 1330-1345.
Lobel, S. & Rydin, H. 2009. Dispersal and life history strategies in epiphytic metacommunities: alternative solutions to survival in patchy, dynamic landscapes. Oecologia, 161, 569-579.
Lobel, S., Snall, T. & Rydin, H. 2006. Metapopulation processes in epiphytes inferred from patterns of regional distribution and local abundance in fragmented forest landscapes. J. Ecol., 94, 856-868.
Lohmus, A. & Kraut, A. 2010. Stand structure of hemiboreal old-growth forests: characteristic features, variation among site types, and a comparison with FSC-certified mature stands in Estonia. For. Ecol. Manag., 260, 155-165.
Matiss, J. 1987. Latvijas mezainums. In Latvijas mezi (Buss, M. & Vanags, J., eds), pp. 84-96. Avots, Riga (in Latvian).
Meikar, T., Viilma, K. & Lepp, A. 2004. The effect of man-caused disturbances on the development dynamics of forest communities on Abruka Island (West Estonia). Proc. Estonian Acad. Sci. Biol. Ecol., 53, 208-225.
Mezaka, A., Brumelis, G. & Piterans, A. 2008. The distribution of epiphytic bryophyte and lichen species in relation to phorophyte characters in Latvian natural old-growth broad leaved forests. Folia Cryptogamica Estonica, 44, 88-99.
Mitchell, F. J. G. 2005. How open were European primeval forests? Hypothesis testing using paleoecological data. J. Ecol., 93, 168-177.
Mitchell, F. J. G. & Cole, E. 1998. Reconstruction of long-term successional dynamics of temperate woodland in Bialowieza Forest, Poland. J. Ecol., 86, 1042-1059.
Nesterovs, V. 1954. Vispariga mezkopiba. Latvijas valsts izdevnieciba, Riga (in Latvian).
Niklasson, M., Lindbadh, M. & Bjorkman, L. 2002. A long-term record of Quercus decline, logging and fires in a southern Swedish Fagus-Picea forest. J. Veg. Sci., 13, 765-774.
Nikodemus, O., Karklins, A., Klavins, M. & Melecis, V. 2008. Augsnes ilgstpejiga izmantosana un aizsardziba. LU Akademiskais apgads, Riga (in Latvian).
Nilsson, S. G., Niklasson, M., Hedin, J., Aronsson, G., Gutowski, J. M., Linder, P., Ljungberg, H., Mikusinski, G. & Ranius, T. 2003. Erratum to "Densities of large living and dead trees in old growth temperate and boreal forests". For. Ecol. Manag., 178, 355-370.
Okland, B., Gotmark, F. & Norden, B. 2008. Oak woodland restoration: testing the effects on biodiversity of mycetophilids in Southern Sweden. Biodiv. Conserv., 17, 2599-2616.
Paltto, H., Norden, B., Gotmark, F. & Franc, N. 2006. At which spatial and temporal scale does landscape context affect local density of Red Data Book and indicator species? Biol. Conserv., 133, 442-454.
Paltto, H., Norden, B. & Gotmark, F. 2008. Partial cutting as a conservation alternative for oak (Quercus spp.) forest--response of bryophytes and lichens on dead wood. For. Ecol. Manag., 256, 536-547.
Peterken, G. F. 1996. Natural Woodland: Ecology and Conservation in Northern Temperate Regions. Cambridge University Press, Cambridge.
Ranius, T. & Hedin, J. 2001. The dispersal rates of a beetle, Osmoderma eremita, living in tree hollows. Oecologia, 126, 363-370.
Sacenieks, R. & Matuzans, J. 1964. Mezsaimniecibas Tabulas. Latvijas Valsts Izdevnieciba, Riga (in Latvian).
Sjors, H. 1963. Amphi-Atlantic zonation, nemoral to Arctic. In North Atlantic Biota and Their History (Love, A. & Love, D., eds), pp. 109-125. The Macmillan Company, New York.
Skinkis, C. 1998. Urgent water management problems in Latvia. Proc. Latvian Acad. Sci., 52, 332-337 (in Latvian).
Snall, T., Hagstrom, A., Rudolphi, J. & Rydin, H. 2004. Distribution pattern of the epyphyte Neckera pennata on three spatial scales - importance of past landscape structure, connectivity and local conditions. Ecography, 27, 757-766.
Snall, T., Pennanen, J., Kivisto, L. & Hanski, I. 2005. Modelling epiphyte metapopulation dynamics in a dynamic forest landscape. Oikos, 109, 209-222.
State Forest Service. 2008. Forest Statistics 2007 (MS Excel spreadsheets). CD ROM.
Terauds, A., Brumelis, G. & Nikodemus, O. 2011. Seventy-year changes in tree species composition and tree ages in state-owned forests in Latvia. Scand. J. For. Res., 26, 446-156.
Thor, G., Johansson, P. & Jonsson, M. T. 2010. Lichen diversity and red-listed lichen species relationships with tree species and diameter in wooded meadows. Biodiv. Conserv., 19, 2307-2328.
Van Uytvanck, J., Maes, D., Vandenhaute, D. & Hoffmann, M. 2008. Restoration of woodpasture on former agricultural land: the importance of safe sites and time gaps before grazing for tree seedlings. Biol. Conserv., 141, 78-88.
Vasilevskis, A. 2007. Meziericibas darbu izversana. In Meziericiba Latvija (Zviedre, A., ed.), pp. 28-31. V. elements, Riga (in Latvian).
Vera, F. W. 2000. Grazing Ecology and Forest History. CABI Publishing, Wallingford, Oxon, UK.
Whitehouse, N. J. & Smith, D. N. 2004. 'Islands' in Holocene forests: implications for forest openness, landscape clearance and 'culture-steppe' species. Environ. Archaeol., 9, 203-212.
Zunde, M. 1999. Mezainuma un koku sugu sastava parmainu dinakika un to galvenie ietekmejosie faktori Latvijas teritorija. In Latvijas Meza Vesture Lidz 1940. gadam (Strods, H., ed.), pp. 111-203. WWF-Pasaules Dabas Fonds, Riga (in Latvian).
Sandra Ikauniece (a), Guntis Brumelis (b) *, and Toms Kondratovics (b)
(a) Faculty of Natural Sciences and Mathematics, Daugavpils University, Vienibas iela 13, Daugavpils LV 5400, Latvia
(b) Faculty of Biology, University of Latvia, Kronvalda bulv. 4, Riga LV 1010, Latvia
([mail]) Corresponding author, firstname.lastname@example.org
Received 6 July 2011, revised 28 November 2011, accepted 29 November 2011
Table 1. Stand inventory data for the years 1924-1937. Age and wood volume of dominant tree species are given. * Relative volume was not estimated for tree species with volume < 10, but their presence was recorded Map % Site (description) Species Age volume Audile 1924 (forest) Quercus robur 130 100 Pededze 1934 (forest) Quercus robur 90 90 Fraxinus excelsior 90 10 Salenieki 1927 (forest) Populus tremula 50 50 Betula pendula 50 50 Barkava 1937 (forest) Quercus robur 80 90 Populus tremula 70-90 10 Betula pendula 70-90 < 10* Alnus glutinosa 70-90 < 10* Kinguru 1927 (scattered Quercus robur 80 100 trees) Populus tremula 15 < 10* Betula pendula 15 < 10* Picea abies 15 < 10* Mezotne 1926 (forest) Quercus robur 150 100 Rauda 1929 (scattered Quercus robur 70-80 100 trees) Table 2. Density of stems with height <1 m (number [ha.sup.-1]). Qr-- Quercus robur, Fe--Fraxinus excelsior, Ag--Alnus glutinosa, Tc--Tilia cordata, Pa--Picea abies, Ai--Alnus incana, Ug--Ulmus glabra, Bp-- Betula pendula, and Pt--Populus tremula Site Qr Fe Ag Tc Pa Audile 430 24 500 60 Pededze 640 4 940 10 30 Barkava 370 660 110 Salenieki 390 180 Kinguru 210 20 Mezotne 22 250 16 500 Rauda 70 100 70 Site Ai Ug Bp Pt Total Audile 80 20 25 090 Pededze 10 3 040 8 670 Barkava 50 20 1 710 2 920 Salenieki 570 Kinguru 230 Mezotne 38 750 Rauda 10 250 Table 3. Volume ([m.sup.3] [ha.sup.-1]) of dead and living stems and number of cut stumps (number [ha.sup.-1]). The proportion (%) of dead standing wood volume is also given. For abbreviations of tree species see Table 2 Site Qr Fe Ag Tc Pa Ai Ug Volume dead, [ha.sup.-1] Audile 150 5 43 5 1 Pededze 108 10 13 Barkava 103 <1 2 4 Salenieki 129 Kinguru 34 52 Mezotne 67 8 Rauda 132 4 5 Volume living, [ha.sup.-1] Audile 286 122 49 42 12 Pededze 742 16 36 7 19 Barkava 406 6 21 4 1 Salenieki 352 2 34 Kinguru 505 5 280 Mezotne 736 30 Rauda 463 105 Site Bp Pt Total Proportion Cut standing stumps dead wood, % Volume dead, [ha.sup.-1] Audile 198 50 Pededze 133 0 Barkava 109 0 10 Salenieki 129 69 Kinguru 86 0 60 Mezotne 75 0 60 Rauda 2 143 80 Volume living, [ha.sup.-1] Audile 1 1 513 Pededze 820 Barkava 10 12 460 Salenieki 5 393 Kinguru 790 Mezotne 766 Rauda 568 Table 4. Woodland key habitat indicator epiphytes recorded in the studied forest stands in Latvia Audile Pededze Barkava Saleniece Lichens Acrocordia gemmata + Arthonia byssacea + + + + Arthonia leucopellea + Arthonia spadicea + + + + Arthonia vinosa + Bacidia rubella + Lecanactis abietina Lobaria pulmonaria + Bryophytes Anomodon spp. + + Dicranum viride * Homalia + + trichomanoides Isothecium alopecuroides Neckera pennata Kinguru Mezotne Rauda Lichens Acrocordia gemmata Arthonia byssacea + Arthonia leucopellea Arthonia spadicea + + + Arthonia vinosa + Bacidia rubella Lecanactis abietina + Lobaria pulmonaria Bryophytes Anomodon spp. + Dicranum viride * + Homalia + + + trichomanoides Isothecium + alopecuroides Neckera pennata + * Recorded in the forest stand during WKH inventory but not found in the plots.
|Printer friendly Cite/link Email Feedback|
|Author:||Ikauniece, Sandra; Brumelis, Guntis; Kondratovics, Toms|
|Publication:||Estonian Journal of Ecology|
|Date:||Mar 1, 2012|
|Previous Article:||Trees for marking boundaries of landed property in premodern Estonia/ Puud maavalduste piiritahisena kesk- ja uusajal.|
|Next Article:||Botanical nature writing: an ecocritical analysis/Botaaniline looduskirjandus: kirjandusokoloogiline analuus.|