Polyculture affects biomass production of component species but not total standing biomass and soil carbon stocks in a temperate forest plantation

Over-yielding of stand biomass did not occur in a tree polyculture comprised of Betula pendula, Alnus glutinosa and Fagus sylvatica selected for contrasting traits. This was due to antagonistic interactions between the component species. Fine root dynamics and soil C stocks were unaffected by species mixture. Increasing CO2 fixation in tree biomass through afforestation and forest management actions has potential for cost-effective climate mitigation. The influences of tree mixture on biomass production and subsequent soil C accumulation in polyculture still remain uncertain. We studied the polyculture of Alnus glutinosa (L.) Gaertn, Betula pendula Roth and Fagus sylvatica L. in a plantation forest to examine the effectiveness of species mixtures as a tool for increased biomass production and soil C accumulation. Tree biomass was estimated by developing species-specific allometric models and 3 years tree measurement. Fine root biomass and production were estimated by root coring and root-mesh methods. The ‘relative yield of mixture’ approach was used to examine the mixture effect. In mixture, an additive effect was observed in A. glutinosa (13% increase in basal diameter relative to the monoculture); however, there was no overall effect of mixture on total standing biomass due to the suppression of F. sylvatica (2.75 g m−2 reduction in woody biomass). Fine root biomass production showed no mixture effect. The quantity and quality of soil C (top 0.5 m) was not affected by tree mixture. We conclude that the contrasting growth responses of the A. glutinosa, B. pendula and F. sylvatica in polyculture resulted in no over-yielding of standing biomass despite the complementary traits of the component species.

3 Tree species identity can enhance soil C stock through aboveground litter production and fine root turnover 60 under the influence of soil and site characteristics. The magnitude of C inputs to soil by above ground litter-61 flux may depend on the quality and quantity of litter, biodegradation and soil properties (Jandl et al. 2007).

62
Dead roots and rhizodeposition of tree root systems are likely more potential to stabilize soil C than 63 aboveground litter due to slow decay rate and mineral interactions (Vesterdal et al. 2013). In addition, many 64 previous studies reported that the stock and stability of soil C under these processes fluctuate over stand age 65 (Chen et al. 2013). In general, most of the factors affecting C inputs are species specific, therefore the 66 impacts of tree plantation on soil C storage are highly variable depending on the species selection. because of the variability in competition and growth rates. In the present study, we used two fast-growing 77 species with a late successional tree in the polyculture stand to study whether the interactions of these 78 contrasting traits affect the biomass production and soil C stocks compared to monoculture conditions. 79 Assessing forest biomass currently lacks methodological robustness. Since our study was confined to a 80 single location, however, the traditional approach of tree harvesting was followed to estimate aboveground 81 biomass, which is a more reliable method than others (Weiskittel et al. 2015 (Clarke, 1940). This is classified as Dystric Cambisols according to the FAO system and recognized as the

131
As most of the trees were not perfectly round, the geometric mean of the highest and lowest diameter was 132 calculated to estimate DBH.

133
After harvesting the bole, branches, dry leaves (most of the trees were leafless, hence the biomass was 134 termed as woody biomass) and catkins were separated. The fresh weight of all separated parts was measured 135 using an electrical balance (OHAUS, 5000 Series, Xtreme W, T51XW), the dry mass of tree components 136 was determined from oven dried subsamples. For each species, four tree parameters (DBH, basal diameter, 137 branch dry-mass and tree height) were considered to predict the above-ground woody biomass by 138 developing allometric models. Based on the goodness of fit indices, basal diameter models were selected for 139 B. pendula and F. sylvatica and the DBH model for A. glutinosa. The selected two parameters (basal 140 diameter and DBH) were checked with three non-linear models, viz. power, exponential and logarithm, of 141 which power models were found as the best fit for predicting woody biomass. The following three equations 142 were developed for biomass estimation (Appendix Figure S1 and   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 standing biomasses of each plot, the basal diameter and DBH data of individual trees were used for the 150 above-mentioned allometric equations. washed through a set of mesh sieves (2.0-0.5 mm) with tap water and the fine roots were sorted following 166 the handpicking approach based on the physical characteristics of the root matrix. Non-tree roots such as 167 herbaceous roots were soft (non-lignified) and lighter in colour than tree roots, grass roots were white, soft 168 and more elastic, while the moss roots were black and rigid. Some non-tree roots had fine root hairs which 169 were absent in three tree roots in our study. To distinguish the roots of the different tree species, fine roots 170 were compared to 'specimen roots' of three species collected during the field studies. The roots were 171 distinguished based one colour, texture and branching pattern, often using a magnifying glass. The sorted 172 fine roots within each soil core were dried at 70°C till constant weight and dry mass was recorded.

175
To estimate annual fine root (< 2 mm) production, a root-mesh technique (Godbold et al. 2003; Lukac and 176 Godbold, 2010) was used. In this method, a nylon mesh strip (7 cm × 25 cm, 1 mm mesh size) was pushed 177 into the ground vertically with a steel blade and hammer. Four strips were inserted at 50 cm distance from 178 each target trees (three trees in each monoculture and three component trees in each polyculture plots were 179   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 randomly selected for this experiment) and retrieved after two intervals (6 months each). The fine roots that 180 crossed through the net were collected and dry mass was determined. The roots of other species (non-tree 181 species or other tree species in mixed plot) were carefully separated following the methods described in 182 previous section. The root biomass turnover rate was calculated as annual root production divided by the 183 mean standing root biomass.   226 T-test: An independent sample t-test was performed to compare the actual and predicted biomass in the 227 mixed species plots, assuming the actual and predicted biomass as two different groups of cases. Similarly 228 species level biomass, tree height and DBH between mono and polyculture plots were examined by t-test.

229
ANOVA and t-test were conducted with SPSS 16.0 (SPSS Inc., Chicago, IL) and the level of significance P 230 <0.05 was accepted in all cases.

275
The C stock in F. sylvatica soil was significantly lower than that in B. pendula at 0-10 cm (p= 0.019) and 276 lower than that in B. pendula and A. glutinosa at 10-20 cm soil layers (p=0.004 and 0.013, respectively), but 277 no effect of species polyculture on soil C was observed (Table 4) 3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 survive well in the understory, the growth of young F. sylvatica is much higher when directly exposed to 299 sunlight (in monoculture) than growing with diffused light in the understory of a polyculture stand. 300 Although increased A. glutinosa biomass in polyculture was not statistically significant, the quantity was 301 substantial when we compared with decreased biomass of F. sylvatica. The increasing trend in biomass 302 production, A. glutinosa showed a positive growth trend in polyculture. Therefore our predictions of growth 303 and biomass increase at species level in polyculture were not fully supported by the tree species we used, 304 rather A. glutinosa (increased diameter growth) and F. sylvatica (decreased growth and biomass production)   Table A3). High soil organic C in mixed stand 341 are generally attributed to high litter inputs (Forrester et al. 2013), hence no obvious variation in soil C was 342 observed between mono and polyculture stands in the present study.

Plantation plots (n=4)
Fine root biomass at different soil depths (g m -2 ) Fine root production (g m -2 ) Fine root turnover (rate y -1 ) 0-10 cm 10-20 cm 20-30 cm Total Single plot Mixed plots Appendix Table A1 Exponential and logarithm models to examine the relationship between woody biomass and DBH and basal diameter (diameter at 22.5 cm). General model y = a e bx (exponential) and y = y0+ ln x (Logarithm), where y = woody biomass of plant (kg), x = tree variables (here D and d denotes DBH and basal dia. (diameter) in mm respectively), a and b are regression coefficients.