Variation in fine root biomass of three European tree species: Beech (Fagus sylvatica L.), Norway spruce (Picea abies L. Karst.), and Scots pine (Pinus sylvestris L.)

Abstract

Fine roots (<2 mm) are very dynamic and play a key role in forest ecosystem carbon and nutrient cycling and accumulation. We reviewed root biomass data of three main European tree species European beech, (Fagus sylvatica L.), Norway spruce (Picea abies L. Karst.) and Scots pine (Pinus sylvestris L.), in order to identify the differences between species, and within and between vegetation zones, and to show the relationships between root biomass and the climatic, site and stand factors. The collected literature consisted of data from 36 beech, 71 spruce and 43 pine stands. The mean fine root biomass of beech was 389 g m⁻², and that of spruce and pine 297 g m⁻² and 277 g m⁻², respectively. Data from pine stands supported the hypothesis that root biomass is higher in the temperate than in the boreal zone. The results indicated that the root biomass of deciduous trees is higher than that of conifers. The correlations between root biomass and site fertility characteristics seemed to be species specific. There was no correlation between soil acidity and root biomass. Beech fine root biomass decreased with stand age whereas pine root biomass increased with stand age. Fine root biomass at tree level correlated better than stand level root biomass with stand characteristics. The results showed that there exists a strong relationship between the fine root biomass and the above-ground biomass.

Key words:

• Below-ground biomass
• boreal forests
• climatic variables
• soil variables
• temperate forests

Introduction

In forest ecosystems the below-ground carbon pool often exceeds the above-ground pool. Root systems account for 10 – 45% of the total tree stand carbon pool, and most of it is allocated to stumps and coarse roots (Santantonio et al., 1977; Fogel, 1983; Helmisaari et al., 2002). Fine root (<2 mm) biomass is relatively small (Keyes & Grier, 1981; Vogt et al., 1996); nevertheless fine roots are very dynamic and play a key role in forest ecosystem carbon and nutrient cycling and accumulation (Berg, 1984; Joslin & Henderson, 1987; Hendrick & Pregitzer, 1993; Helmisaari et al., 2002). There is less information about root system carbon pools and it is more difficult to generalize than that of the above-ground pools, mainly due to methodological difficulties in determining the fine root biomass, labor-intensive nature of such studies, and variety of internal and external factors affecting stand level root biomass (e.g. Persson, 1983; Vogt & Persson, 1991).

In previous reviews fine root biomass has been related to several abiotic and biotic factors such as soil properties, tree species, stand characteristics, and climatic factors (Jackson et al., 1996; Vogt et al., 1986, 1996; Cairns et al., 1997; Leuschner & Hertel, 2003; Chen et al., 2004). In most cases, fine root biomass and the abiotic or biotic factors have not been correlated or only weakly correlated when datasets from different sources have been included. The relationships have been often clearer and stronger in studies carried out with the same species, within a limited geographical area, or in controlled experiments. Based on the earlier findings the following hypotheses can be presented.

1. Tree fine root biomass is higher in the temperate than in the boreal zone (Vogt et al., 1986; Jackson et al., 1996).

2. Deciduous trees have a higher fine root biomass than conifer trees in the temperate zone (Vogt et al., 1986, 1996; Leuschner & Hertel, 2003).

3. The higher the soil fertility the lower the fine root biomass (Keyes & Grier, 1981; Vogt et al., 1983, 1986, 1987; Jackson et al., 1996; Finér & Laine, 1998; Helmisaari et al., 2007), even though there can be species-specific differences (Leuschner & Hertel, 2003).

4. The higher the soil acidity the lower the fine root biomass, but there can be species-specific differences and differences related to the length of exposure to the acid deposition (Jentschke et al., 2001; Godbold et al., 2003; Leuschner & Hertel, 2003; Leuschner et al., 2004).

5. The fine root biomass increases until canopy closure, thereafter it does not increase or decrease (Vogt et al., 1983; Vanninen et al., 1996; Helmisaari et al., 2002; Claus & George, 2005), but there can be species specific variation from this pattern (Leuschner & Hertel, 2003).

6. Fine root biomass at tree level correlates better than the stand level fine root biomass with the stand characteristics (Chen et al., 2004; Helmisaari et al., 2007).

The aims of our study were as follows: (1) to review and summarize the fine root biomass data of major forest-forming species in Europe, which was available in published literature, (2) to determine differences between species within the vegetation zones, (3) to determine differences in fine root biomass between temperate vs. boreal zone, and (4) to reveal possible relationships between fine root biomass and the climatic, site and stand factors.

Material and methods

Collection of root biomass data

The data collected for this review had to fulfill the following criteria:

1. Only published data were used.

2. We limited our data set to common European tree species, for which large number of published information on fine root biomass exists. These species were European beech (Fagus sylvatica L.), Norway spruce (Picea abies L. Karst.) and Scots pine (Pinus sylvestris L.). Data from very young stands (age < 10 years) were not included. The data sets included stands of Norway spruce and Scots pine from temperate to boreal zones and beech stands only from the temperate zone.

3. Fine root biomass was determined with the coring method. The data had been collected at one sampling event or represented means of several seasonal or annual samplings. The number of core samples and the sampling depth varied between studies. It was assumed that the sampling covered the depth of majority of fine root biomass.

4. Only fine root biomass < 2 mm in diameter was included. Living (biomass) and dead roots (necromass) were treated separately.

5. Fine root data from stands or plots where soil manipulations had been done were excluded.

Additional background information collected for the forest stands:

1. Sites were classified into three fertility classes: poor, medium, rich. The classification was in some cases done already in original papers (e.g. Leuschner & Hertel, 2003) or it was done by the authors using the information found in the original papers.

2. The following site characteristics were collected from the original papers: the location of the site, elevation, mean annual precipitation, mean annual air temperature, topsoil pH, and topsoil C/N ratio. The following tree stand characteristics were recorded: tree stand age, height, mean diameter, density, volume and basal area.

The data basis

The collected material consisted of data from 36 beech, 71 Norway spruce and 43 Scots pine stands (Tables I –  III, Figure 1). In addition, 18 of the beech stands, 20 of the spruce, and 8 pine stands were included from the review of Leuschner & Hertel (2003) concerning tree stands in the temperate zone. All beech stands were located in the temperate zone within a narrow longitudinal gradient, but there was substantial variation in the climatic, tree, and soil characteristics between these stands. There were spruce and pine stands from both temperate and boreal zones with large variation in the climatic, tree, and soil characteristics between the stands. There were spruce stands also at high elevations whereas the pine stands were mainly located in lowlands. All site characteristics were not available from all stands.

Figure 1. The location of reviewed stands. The stands located in Finland, Estonia, Norway and Sweden were classified in the boreal zone.

Table I. Sources of data

Vegetation zone	Species	Reference
Temperate	Fagus	Al Sayegh-Petkovšek (2000), Bolte & Villanueva (2006), Claus & George (2005), Epron et al. (1999), Grebenc (1995), Grebenc & Kraigher (2007a,b), Grebenc et al. (2005), Hendriks & Bianchi (1995), Hobbie et al. (Ecosystems, 2006, submitted), Košir (1976), Kutnar et al. (2002), Kraigher et al. (2007), Leuschner & Hertel 2003, Leuschner et al. (2004), Schmid (2002), Vilhar & Kraigher (1999)
Temperate	Picea	Ammer & Wagner (2005), Bolte & Villanueva (2006), Borken et al. (2007), Brunner et al. (2002), Claus & George (2005), Genenger et al. (2003), Godbold et al. (2003), Grebenc (2005), Grebenc et al. (2007), Hobbie et al. (Ecosystems, 2006, submitted), Konôpka & Pavlenda (2004), Košir (1976), Kraigher (1997, 1999), Kraigher et al. (1996), Kutnar et al. (2002), Leuschner & Hertel (2003), Murach (1984), Schmid (2002), Vilhar & Kraigher (1999)
Temperate	Pinus	Janssens et al. (2002), Hobbie et al. (Ecosystems, 2006, submitted), Konŏpka et al. (2005), Leuschner & Hertel (2003), Oleksyn et al. (1999), Vanguelova et al. (2005), Xiao et al. (2002)
Boreal	Picea	Eldhuset et al. (2006), Finér et al. (2003), Helmisaari & Hallbäcken (1999), Helmisaari et al. (2007), Lõhmus & Oja (1983), Majdi (2001), Majdi & Persson (1995), Möttönen et al. (2003), Ohashi et al. (2007), Ostonen et al. (2005), Püttsepp et al. (2006)
Boreal	Pinus	Finér & Laine (1998), Helmisaari et al. (2002, 2007), Mälkönen (1974), Mamayev (1977, 1986), Persson (1979), Vanninen & Mäkelä (1999)

Table II. Descriptive statistics for the beech, spruce and pine stands in the temperate zone

Variable	Mean	Min	Max	SD	N
Fagus
Latitude, °N	50.5	45	53	2.1	36
Longitude,°E	10.4	5.4	18.1	2.8	36
Elevation, m	427	115	900	186	33
Mean annual precipitation, mm	882	462	1624	313	33
Mean annual temperature,°C	7.7	6.0	9.2	0.77	33
Stand density, trees, ha^−1	788	108	4768	1280	17
Basal area, m^2ha^−1	27	17	52	9.6	13
Basal area/tree, m^2	0.07	0.0	0.23	0.07	11
Mean diameter, cm	33	7	54	33	15
Height, m	24	11	34	6.5	14
Stem volume, m^3ha^−1	379	179	800	284	4
Stand age, years	111	30	250	46	36
Topsoil, mineral, pH	4.4	2.9	7.0	1.2	27
Soil organic layer, C/N	30	22	35	7.0	3
Fine root biomass, g m^−2	389	116	960	206	36
Fine root necromass, g m^−2	920	84	3500	948	22
Fine root biomass per tree, kg	13.8	0.7	38.2	10.7	17
Picea
Latitude, °N	48.6	40	52	3.2	47
Longitude,°E	12.0	7.5	19.2	3.0	47
Elevation, m	714	150	1720	398	47
Mean annual precipitation, mm	1056	591	2000	317	43
Mean annual temperature,°C	6.5	3.0	9.6	1.7	41
Stand density, trees, ha^−1	1090	272	5858	1470	22
Basal area, m^2ha^−1	41	18	62	12.4	13
Basal area/tree, m^2	0.08	0.0	0.18	0.06	13
Mean diameter, cm	26	6	53	11	16
Height, m	19	7	31	7.5	11
Stem volume, m^3ha^−1	589	280	720	178	5
Stand age, years	89	24	200	38	46
Topsoil, mineral, pH	3.8	2.8	5.5	0.7	28
Soil organic layer, C/N	31	24	34	3.9	6
Fine root biomass, g m^−2	281	63	640	134	47
Fine root necromass, g m^−2	176	55	368	91	15
Fine root biomass per tree, kg	5.0	0.4	10.7	3.4	22
Pinus
Latitude, °N	51.8	51	52	0.4	14
Longitude,°E	10.1	−0.51	18.1	5.1	14
Elevation, m	81	16	180	48	14
Mean annual precipitation, mm	692	591	767	692	13
Mean annual temperature,°C	8.6	8.2	9.8	0.5	11
Stand density, trees, ha^−1	2998	374	10016	4134	5
Basal area, m^2ha^−1	32	20	41	10.9	3
Basal area/tree, m^2	0.05	0.01	0.11	0.05	3
Mean diameter, cm	17	6	30	11	4
Height, m	16	6	22	6.8	4
Stem volume, m^3ha^−1	–	–	–	–	–
Stand age, years	85	12	131	35	14
Topsoil, mineral, pH	3.9	3.3	6.7	1.0	11
Soil organic layer, C/N	–	–	–	–	–
Fine root biomass, g m^−2	377	138	725	188	14
Fine root necromass, g m^−2	168	41	311	110	9
Fine root biomass per tree, kg	2.9	0.2	7.2	2.9	5

Table III. Descriptive statistics for the spruce and pine stands in the boreal zone

Variable	Mean	Min.	Max.	S.D.	N
Picea
Latitude, °N	62.5	56	68	3.2	24
Longitude, °E	24.2	11.1	30.0	5.7	24
Elevation, m	177	3	310	80	24
Mean annual precipitation, mm	661	460	1145	174	24
Mean annual temperature, °C	4.1	1.2	7.5	2.5	9
Stand density, trees, ha^−1	1208	374	2863	633	21
Basal area, m^2ha^−1	29	11	54	13.3	15
Basal area/tree, m^2	0.03	0.01	0.08	0.02	15
Mean diameter, cm	22	10	32	6	11
Height, m	16	8	28	6.4	17
Stem volume, m^3ha^−1	176	57	671	176	18
Stand age, years	79	24	181	47	24
Topsoil, mineral, pH	4.2	3.9	4.4	0.2	6
Soil organic layer, C/N	34	27	47	8.3	11
Fine root biomass, g m^−2	330	100	748	157	24
Fine root necromass, g m^−2	221	50	383	103	7
Fine root biomass per tree, kg	3.3	1.1	8.8	2.0	21
Pinus
Latitude, °N	61.6	58	69	2.3	29
Longitude, °E	27.1	16.4	38.0	5.7	29
Elevation, m	125	48	185	38	25
Mean annual precipitation, mm	641	419	750	89	22
Mean annual temperature, °C	2.8	1.9	3.7	0.9	7
Stand density, trees, ha^−1	1487	199	7425	1444	29
Basal area, m^2ha^−1	18	2	36	8.5	26
Basal area/tree, m^2	0.03	0.0	0.11	0.03	26
Mean diameter, cm	17	2	37	9	26
Height, m	14	2	24	6.4	26
Stem volume, m^3ha^−1	156	8	311	91	18
Stand age, years	69	15	220	49	29
Topsoil, mineral, pH	4.9	3.5	6.0	0.8	9
Soil organic layer, C/N	43	33	53	5.7	11
Fine root biomass, g m^−2	229	26	467	102	29
Fine root necromass, g m^−2	72	39	126	32	5
Fine root biomass per tree, kg	3.0	0.2	12.0	2.7	29

Statistical tests

One way Anova was used to test the differences between species and vegetation zones and the paired comparisons were done with Tukey's test. Pearson correlation coefficients were calculated between fine root biomass and climatic, tree stand and soil characteristics of the stand, and linear regressions were fit to the data. The analyses were done with the Anova, Correlation, and Regression procedures of the SPSS® Base for Windows 14.0 statistical package.

Results

Stand level fine root biomass

The mean living beech fine root biomass was 389 ± 206 g m⁻², that of spruce 297 ± 143 g m⁻², and that of pine 277 ± 150 g m⁻². The necromasses were 920 ± 948 g m⁻², 190 ± 95 g m⁻², and 133 ± 100 g, respectively. In the temperate zone the fine root biomass of beech was significantly higher than that of spruce (p = 0.015; Table II). The fine root biomass of spruce and pine did not differ from each other either in the temperate zone or in the boreal zone (Tables II and III). For spruce the fine root biomass did not differ between the temperate and boreal zones, but for pine it was significantly (p = 0.002) higher in the temperate zone than in the boreal zone.

There was a significant positive correlation between beech and pine fine root biomass and latitude in the temperate zone (Table IV). Among pine stands the variation in latitude was only 1° (Table II), and most of the beech data were from a narrow latitude range without any latitudinal trend in fine root biomass. There were no significant correlations between the fine root biomass and the elevation of the stand in either of the vegetation zones.

Table IV. The Pearson correlation coefficients between the fine root biomass (g m²) and the climatic, site and stand factors.^a

Variable	Fagus, temperate		Picea, temperate		Picea, boreal		Pinus, temperate		Pinus, Boreal
Latitude, °N	0.382*	(36)	−0.077	(47)	0.005	(24)	0.535*	(14)	0.278	(29)
Elevation, m	−0.089	(33)	0.080	(47)	0.201	(24)	−0.198	(14)	−0.070	(25)
Mean annual precipitation, mm	−0.058	(33)	0.293	(43)	0.406*	(24)	0.080	(13)	−0.323	(22)
Mean annual temperature,°C	−0.238	(33)	0.011	(41)	−0.258	(9)	0.094	(11)	−0.314	(7)
Stand density, trees, ha^−1	0.109	(17)	0.036	(22)	0.422	(21)	−0.102	(5)	0.023	(29)
Basal area, m^2ha^−1	−0.462	(13)	−0.609*	(13)	−0.721*	(15)	–	(–)	0.357	(26)
Stand age, years	−0.364*	(36)	0.216	(46)	−0.244	(24)	0.720*	(14)	0.034	(29)
Topsoil, mineral, pH	−0.377	(27)	−0.076	(28)	−0.743	(6)	−0.265	(11)	−0.017	(9)
Soil organic layer, C/N	–	(–)	0.038	(6)	0.579	(11)	–	(–)	0.563	(11)
^aStatistically significant correlations (p < 0.05) are indicated by asterisks. The number of observations is in parenthesis.

There was a significant positive correlation between the mean annual precipitation and the fine root biomass of spruce in the boreal zone (Table IV), but most of the data were from a narrow precipitation range without any relationship between the amount of precipitation and the fine root biomass.

In beech (p = 0.08, N = 35) and spruce (p = 0.007, N = 70) stands there was a significant difference in the fine root biomasses between different fertility classes, but no such differences existed in pine stands (Figure 2). The fine root biomass of spruce stands was lowest in the poorest sites, whereas in beech stands it was the highest in the poorest sites.

Figure 2. Mean fine root biomass in different site fertility classes for beech, spruce and pine stands. Error bars indicate standard deviations, and in each species group letters show the fertility classes which differ significantly from each other according to the Tukey's test (p < 0.05).

The mineral soil acidity did not correlate with the fine root biomass (Table IV). However, there was a significant relationship between pine fine root biomass and the C/N ratio in the soil organic layer (Figure 3).

Figure 3. Relationship between C/N ratio of organic soil layer and fine root biomass for Scots pine stands.

The fine root biomass of beech decreased with stand age and that of pine increased with stand age in the temperate zone (Figure 4). There was one exceptionally old beech stand in the data. However, the relationship between beech stand age and fine root biomass changed only slightly if that stand was excluded from the analysis (y = −1.705 + 577.416, p = 0.053).

Figure 4. Relationship between stand age and fine root biomass in the temperate zone for beech and Scots pine stands.

The fine root biomass of spruce decreased with the increase in the basal area of the stand in both vegetation zones (Figure 5). At the stand level, there were no other significant correlations between the fine root biomass and the stand characteristics.

Figure 5. Relationship between stand basal area and fine root biomass for spruce in the temperate and boreal zones.

Root biomass at tree level

The fine root biomass per tree was calculated by dividing the fine root biomass per hectare with the stem number per hectare. It was significantly higher for beech than for spruce (p = 0.001) or pine (p = 0.013) (Tables II and III). The fine root biomass per tree did not differ between spruce or pine in either of the zones. The ratio of the stand basal area to the fine root biomass did not differ significantly between studied tree species (p = 0.681).

The root biomass per tree increased with stand age in beech and in spruce stands in the temperate zone and in pine stands in both vegetation zones (Table V, Figure 6). The fine root biomass per tree increased with basal area per tree in beech and in spruce stands in the temperate zone and in pine stands in the boreal zone (Figure 7).

Figure 6. Relationship between stand age and fine root biomass per tree for beech and spruce in the temperate zone and for Scots pine in the temperate and the boreal zone.

Figure 7. Relationship between stand basal area per tree and fine root biomass per tree for beech and spruce in the temperate zone and for Scots pine in the boreal zone.

Table V. The Pearson correlation coefficients between biomass per tree (g) and the stand age and basal area per tree.^a

Variable	Fagus, temperate		Picea, temperate		Picea, boreal		Pinus, temperate		Pinus, Boreal
Stand age, years	0.501*	(17)	0.681*	(22)	−0.015	(21)	0.898*	(5)	0.827*	(29)
Basal area/tree, m^2	0.761*	(11)	0.781*	(13)	0.435	(15)	–	(–)	0.846*	(26)
^aStatistically significant correlations (p < 0.05) are indicated by asterisks. The number of observations is in parenthesis.

Discussion

The reviewed papers showed that the average fine root biomasses for the temperate zone species were within the range reported by Noguchi et al. (2007) for other temperate forests (49 – 749 g m⁻²). The fraction of roots <5 mm in diameter was larger than that of <2 mm roots. However, the biomasses reported by Vogt et al. (1996) for <5 mm diameter roots in boreal forests (60 – 165 g m⁻²) were smaller than those reviewed in this study. The root collection methods could differ between studies.

We could reach our aim to find out the differences in fine root biomass between the temperate vs. boreal zones only partly, because for the deciduous trees distributed both in the boreal and the temperate zone the fine root data were too scarce. Silver birch (Betula pendula Roth.) is economically the most important deciduous tree species in the boreal zone, and it is also widely distributed in the temperate zone of Europe (Evans, 1984; Krüssmann, 1976). Fine root biomass (<2 mm in diameter) varied from 154 to 538 g m⁻² in three young and middle-aged silver birch stands (Hobbie et al., Ecosystems, 2006, submitted) (Mamayev, 1977; Uri et al., 2007).

Fine root biomass had a large variation between sites. Despite of the scatter caused by this variation, there were trends supporting or rejecting our hypotheses on the relationships between fine root biomass, and climatic, tree stand and soil characteristics.

The results supported only partly the first hypothesis that the fine root biomass is higher in the temperate than in the boreal zone. However, this was true for pine but not for spruce. It is unlikely that these differences are related to variation in the sampling depth, which was less than 40 cm mineral soil in most of the studies from the boreal Scots pine forests but up to 60 – 100 cm mineral soil in most of the temperate Scots pine forests. Jackson et al. (1996) reported in their review that the rooting depth of the boreal forests is shallower than that of temperate forests. In boreal forests 80 – 90% of roots are in the top 30 cm, or even top 20 cm, since Helmisaari et al. (2007) reported that only 7.5% of Scots pine and Norway spruce fine roots were in the mineral soil layer 20 – 30 cm deep, while in temperate forests 78% of roots are found in the top 50 cm of soil (Jackson et al., 1996). Thus, the underestimate in boreal Scots pine fine root biomass caused by not sampling deeper than 30 or 40 cm in the mineral soil is relatively small, and the minor error related to different sampling depths is not likely to have caused the observed difference between the zones. Spruce sampling depths did not systematically differ between the zones.

The differing age of the Scots pine stands between zones may have been another source of error, as Scots pine fine root biomass correlated positively with stand age in the temperate zone. In the data on temperate Scots pine forests only 14% of the stands were younger than 40 years old whereas the share of young stands was over two times higher (32%) in the data from the boreal region. The greatest difference between Scots pine stands from temperate and boreal zones was, however, stand density. Temperate Scots pine stands were much denser than boreal stands. When the fine root biomass was calculated on the basis of stocking, there were no differences between zones in the mean fine root biomass per pine tree.

Contrary to our results, Jackson et al. (1996) concluded that fine root biomass was generally higher in temperate deciduous and coniferous forests than in the boreal forests. Their data were, however, based only on a few data points from these regions, and European forests were poorly represented. Based on data presented here from European forests we cannot generally conclude that fine root biomass is higher in the temperate zone. This is in agreement with the findings of Vogt et al. (1996).

Some of the observed significant relationships were based on clustered data. For instance, the significant correlation between beech and pine fine root biomass and latitude in the temperate zone was based on clustered data. For beech the cluster consisting of a few southernmost stands (latitude < 46°N) had low beech fine root biomasses which may be partly related to the low sampling depth (down to 20 cm mineral soil). Thus, especially for deeper-rooted beech (Bolte et al., 2004; Claus & George, 2005), the differences in the sampling depth may have been an additional source of error. Most of the beech data was from a narrow latitude range without any latitudinal trend in the beech fine root biomass. Among the temperate pine stands the variation in latitude was only 1° and the data could not be used for analyzing the relationship between the fine root biomass and latitude.

The results supported the second hypothesis that the fine root biomass of deciduous trees is higher than that of conifers. Vogt et al. (1996), Leuschner & Hertel (2003), and Noguchi et al. (2007) reported similar results, whereas those of Jackson et al. (1996) are contradictory to ours. In our collected data beech stands were less dense than spruce or pine stands but beech trees were taller, and growing on fertile site types. Especially the fine root biomass per tree was several times larger for beech than for conifers in the temperate zone. The high fine root necromass of beech indicates that beech fine roots may also have a high turnover rate. This conclusion is supported by fine root life span estimates using minirhizotron images of individual roots in a 30-year-old common-garden experiment with Fagus sylvatica, Pinus sylvestris, and Picea abies in central Poland (Withington et al., 2006). In that study median life span of two smallest root orders (with max. diameter 0.3 – 0.4 mm) of beech was 0.57 years, and was by 15% shorter than that of pine (0.67 years), and by 19% shorter than that of spruce (0.70 years).

The correlations between site fertility characteristics and fine root biomass seemed to be species specific. There was a higher fine root biomass of beech on the poorest sites, and of spruce on medium fertility sites. Also, the positive correlation between the organic layer C/N ratio and pine fine root biomass indicated that the low release of mineral nitrogen increased fine root biomass, thus partly supporting the third hypothesis. The data which reported C/N ratios were solely from Finland. Eight of the pine sites were from the study by Helmisaari et al. (2007), who reported a stronger relationship between the fine root biomass of both Norway spruce and Scots pine and the C/N ratio in Finland, where nitrogen is the main growth-limiting factor on upland soils. Only a few papers reported soil C/N ratios from the temperate region. It may also be that this ratio is ecologically less important in the European temperate region where nitrogen deposition is high, and in many areas nitrogen is no more a growth-limiting factor.

The data did not support the fourth hypothesis – there was no correlation between fine root biomass and soil acidity. Godbold et al. (2003) studied the effect of soil acidity on Norway spruce fine root density in the organic layer, and reported that on the most acidic sites fine root density was higher than on the least acidic sites. As in the beech stands studied by Leuschner et al. (2004) no change in fine root biomass was found with increasing soil acidity.

Precipitation correlated positively only with spruce fine root biomass in the boreal zone. The data was highly scattered and mostly from areas with relatively low annual rainfall (460 – 709 mm year⁻¹), where there was no relationship between fine root biomass and the amount of annual precipitation. The data points (2) with high rainfall (>1100 mm year⁻¹) in boreal forests were from maritime, western Sweden from young stands (24 and 31 years old), which may have partly contributed to data scatter. We did not find significant correlations between fine root biomass and precipitation in the temperate zone, whereas this relationship was reported in several earlier papers. Leuschner & Hertel (2003) found significant correlations between fine root biomass and precipitation for spruce and beech also in the temperate forests. According to Leuschner et al. (2004) beech can grow in regions with both high and low rainfall (500 to 2000 mm year⁻¹), even if it is a drought-sensitive species (Backes & Leuschner, 2000). However, not only the total annual rainfall but its seasonal distribution is important for soil water availability and its effects on fine roots. Along a precipitation gradient in Germany, Leuschner et al. (2004) concluded that beech fine root biomass was highly sensitive to summer droughts resulting in a reduction at sites with low precipitation amounts.

The significant correlation between the fine root biomass of beech and pine with stand age in the temperate zone only partly supported the fifth hypothesis that fine root biomass increases by canopy closure, and thereafter it does not increase or decrease. Many studies have reported that fine root biomass peaks at canopy closure (Helmisaari et al., 2002; Claus & George, 2005), but the stand level data gathered for this study largely varied in this respect. The tree level correlations were better in agreement: fine root biomass increased with stand age, except for spruce in the boreal zone. For beech there was an increasing trend on tree level, but a decreasing trend on stand level. There were less data on the tree level as stand densities were reported in only half of the reviewed beech stands. Many of the highest stand level beech fine root biomasses (>600 g m⁻²) were in stands younger than 40 years old. Our stand level results confirm the results by Claus & George (2005) on beech stands having the highest fine root biomass at younger age. On the contrary, temperate Scots pine forests had clearly lowest fine root biomasses at young age. Scots pine is a light-demanding early-successional species, and reaches canopy closure later than late-successional beech.

The tree level correlations between fine root biomass and tree age and basal area were better than respective correlations on stand level. Only the correlation between spruce fine root biomass and stand basal area in the boreal zone was better on stand level than on tree level. These higher correlations on tree level compared to stand level supported the sixth hypothesis. When tree level characteristics are observed, stand density variation can be omitted. This approach has earlier been used by, for example, Chen et al. (2004) for boreal and temperate forests, and Helmisaari et al. (2007) for boreal Scots pine and Norway spruce forests. We agree with Chen et al. (2004) who concluded that above-ground variables such as tree size distribution should be included to improve usefulness of fine root biomass data for large-scale studies. Also, fine root biomass studies reporting needle biomass would be ecologically valuable as needle: fine root biomass ratio is reported to vary according to, for example, site fertility and stand characteristics and population origin (Oleksyn et al., 1999; Helmisaari et al., 2007).

Conclusions

The average fine root biomass of common European tree species falls within the same range as that in the corresponding vegetation zone in the other parts of the world. Many sources of errors and methodological issues are involved in the estimation of fine root biomass at stand level, and the variation observed in this study between sites may partly originate from these factors. However, here, as in some earlier reviews significant relationships between fine root biomass, and climatic, site and stand characteristics could be found. Our results show clearly that at tree level there exists a strong relationship between the fine root biomass and the above-ground biomass. There is a need to standardize the fine root biomass estimation methods and to collect more information on the relationships between fine root biomass and easily measurable or available climatic, site and stand factors to be better able to include the fine roots in ecosystem carbon pool and flux analyses in Europe and elsewhere.

Acknowledgements

Financial support for this work by European COST Action E38 ‘Woody Root processes’ is gratefully acknowledged. The authors thank Dr. Ain Kull for help in designing the map of the location of stands and Dr. Ivan Janssens for providing data.