EA: Stem CO₂ efflux to the atmosphere; EA-LEAF: Foliage CO₂ efflux to the atmosphere; ES: Soil CO₂ efflux; ES-ROOT: Root-respired CO₂ that diffused to the atmosphere through soil; FT: Internal CO₂ flux through xylem; RR: Root respiration; RS: Stem respiration.

Quercus pyrenaica Willd is a vigorous root-resprouting species that has been intensively coppiced for firewood, charcoal and woody pastures for centuries. Nowadays, coppicing has mostly ceased, and symptoms of stand degradation-slow stem growth, branch dieback, and scarce acorn production-are widely observed [1]. Disproportionate respiratory costs of massive root systems grown after centennial coppicing have been suggested as a potential driver of Q. pyrenaica decay [2]; nevertheless, assessments on carbon expenditures in Q. pyrenaica have not been essayed to date, and causes for coppice stagnation remain unknown.

The quantification of the relative weight of respiratory sinks is crucial for a better understanding of tree carbon budgets [3-5]. However, our comprehension of respiratory processes, particularly of woody organs, is limited relative to our knowledge of photosynthesis [5,6]. In resprouting deciduous species, nonstructural carbohydrates are stored in large amounts in woody organs [7] that can contain a large proportion of living parenchyma [8]. The penalty in terms of respiratory carbon loss associated to storage tissues [9,10] could be of particular relevance in carbon budgets of the root-resprouting Q. pyrenaica , since long lasting coppicing might lead to disproportionate accumulation of resources in roots [11].

Trees of Q. pyrenaica long subjected to coppicing have developed massive systems of living roots (Figure 1) [11] that store but also may consume a great portion of carbohydrates assimilated aboveground. For a better understanding of the role of respiratory carbon loss in Q. pyrenaica decay, we gathered and extended previous published work on Q. pyrenaica biomass and respiration of woody tissues [11-14] to scale up stem and root respiration (RS and RR, respectively) to the tree and stand levels. We aimed to compare respiratory expenditures of above- and below-ground woody organs across one growing season, and in relation to other forests to provide an insight of the relative magnitude of carbon invested for respiration in Q. pyrenaica coppices. We expected RS and particularly RR to be important carbon sinks, and therefore high woody respiration and RR/RS ratios relative to other forest stands.

Figure 1: Photograph of the root system of a hydraulically excavated clone of Quercus pyrenaica located in the Monte Matas de Valsaín (Segovia, Spain).

To estimate annual respiratory loss of woody organs in a coppice system of Q. pyrenaica , we reviewed and reanalyzed our previous work on Q. pyrenaica root development and biomass [11], and internal and external stem CO₂ fluxes [12-14], together with unpublished data of soil CO₂ efflux. Data was obtained from eight stems belonging to a Q. pyrenaica clone located in the Monte Matas de Valsaín (Segovia, Spain). The clonal assignment of the stems was based on molecular markers [15,16]. Four 24 h long measurement campaigns were conducted across the growing season of 2013, at the end of which stems and roots were harvested for biomass quantification and scaling up of RS and RR to the tree and stand levels. The root system was hydraulically excavated with a high-pressure water pump down to 1 m depth over an 81 m² area (Figure 1). Details on stand characteristics, aboveground features of the monitored stems, excavation methodology, and above and below-ground biomass measurements can be seen in our previosly published work [11].

Stem CO₂ efflux to the atmosphere (EA) was measured in every stem at ground level (0.1 m) with a portable infrared gas analyzer (LI-6400, Li-Cor Inc., Lincoln, NE, USA) and a soil chamber (LI-6400-09) [12,13]. Accumulated EA per tree was calculated by scaling up stem EA to functional woody biomass (i.e., sapwood and bark) of the eight stems and their branches. Stem respiration (RS) was estimated as the sum of EA and the internal CO₂ flux through xylem (FT) [RS=EA+FT]. FT was calculated as the product of the sap flux and the CO₂ dissolved in sap solution (sap [CO₂ *]). Sap [CO₂ *] was in turn estimated by Henry’s law from gas phase [CO₂], sap temperature, and sap pH [13,17].

Soil CO₂ efflux (ES) was measured in four collars located below the canopy of the clone with a portable infrared gas analyzer and a soil chamber [14]. Daily ES was averaged among collars on a soil surface area basis. Since roots extended beyond the excavated area (81 m²), a buffer of 0.63 m was added to estimate clonal extension (102 m²). This buffer distance was chosen to meet actual stand density (8 stems/102 m^-2=784 stems ha-1≈781 stems ha^-1=stand stem density) and scale up results to the stand level. RR was calculated as the sum of the rootrespired CO₂ that diffused to the atmosphere through soil (ES-ROOT) and FT at the base of the stem [RR=ES-ROOT+FT] [18]. ES-ROOT was estimated from ES and the relative contribution of autotrophic respiration to ES. Seasonality of root autotrophic contribution to ES was obtained from two studies in a Mediterranean Quercus cerris coppice cut one [19] and 17 [20] years before respiration measurements; spring contributions reported in these studies were attributed to the first campaign (DOY 143-144), summer contributions to the second (DOY 183-184) and third (DOY 218-219) campaigns, and autumn contributions to the fourth campaign (266-267).

Figure 2 shows seasonal and diel variation in EA, FT and ES on a soil surface area basis. The contribution of FT to total aboveground respiration was less than 10% [12], whereas FT belowground was 2% of the ES-ROOT [14]. This is in part due to low xylem [CO₂] in Q. pyrenaica , lower than 0.5% on average, which is one order of magnitude lower than the [CO₂] reported in other species using this methodology [21]. Averaged over the growing season, ES was the greatest respiratory flux to the ecosystem (39 mol CO₂ clone^-1 day^-1), three and four times greater than RS (12 mol CO₂ clone^-1 day^-1) and RR (8-9 mol CO₂ clone^-1 day^-1), respectively (Figure 2) (Table 1). Due to the high magnitude of ES, the relative contribution of autotrophic respiration to ES substantially determined the root-to-shoot ratio of respiration (RR/RS). To illustrate this, the root contribution to ES reported for a Mediterranean oak coppice-ranging between 14 and 27% [19,20]-yielded RR/RS ratios ranging between 0.42 to 1.18 across the growing season (Table 1). Instead, if heterotrophic and autotrophic contributions to ES were considered equal, as generally assumed by some authors for different forest biomes [18,22], RR/RS would increase up to values ranging between 1.36 to 2.18.

Figure 2: Diel variations in stem CO₂ efflux to the atmosphere (EA), internal CO₂ transport through xylem (FT) and soil CO₂ efflux (ES) on four dates over the growing season in an abandoned coppice of Quercus pyrenaica . Fluxes registered from one stem segment and one soil collar intensively monitored (18 times day^-1) are shown. Other stem segments and collars used to average EA, FT and ES are not displayed in this figure as they were monitored less intensively (4 times day^-1). EA and FT data was obtained from previous published work [12] and expressed on a soil surface area basis. Shaded areas indicate night-time.

Table 1: Above- and below-ground fluxes of respired CO₂ in an eightstemmed clone of Quercus pyrenaica . Stem CO₂ efflux to the atmosphere (EA), soil CO₂ efflux (ES), and stem and root respiration (RS and RR, respectively) were measured during four 24-hours campaigns throughout 2013 growing season. RR was calculated as the product of ES multiplied by the autotrophic contribution to ES. Autotrophic contribution to ES was obtained from two reports of a Mediterranean coppice of Quercus cerris cut one [19] and 17 years [20] before measurements. Estimations of RR and RR/RS ratios from contributions reported in both studies (recently coppiced vs mature stand) are shown.

Above- and below-ground functional woody biomass was similar in the surveyed clone: 1026 Kg above-ground and 972 Kg below-ground [11]. Consequently, seasonal deviations of RR/RS from unity reflected differences in the metabolic activity between below- and above-ground organs over time. The relatively high ratio of RR/RS above one during spring (Table 1) may be explained by earlier growth of roots than stems [23], particularly by intense fine root growth and belowground cambial activity for root elongation in spring [24]. The progressive decrease in RR/RS observed onward (ratios below one) may result from a gradual reduction in the root activity, relative to enhanced aboveground metabolism. In other oak coppices the highest root contribution to ES was registered on the dormant season [19,20], which would support a major role of belowground respiration and relatively high RR/RS during winter. Nonetheless, the fact that RR/RS ratios below one were predominantly observed along the growing season evidenced an unexpected high weight of aboveground woody respiration as a carbon sink. The accumulation of woody biomass in stems and branches in this over-aged coppice (cut for the last time around 1967), together with the remarkably high proportion of living parenchyma observed in Q. pyrenaica stems [8], contribute to high woody respiratory costs above ground.

Annual respiratory fluxes were compared with those reported in 15 other studies for different forest stands (Table S1). Leaf area index (LAI), a key eco-physiological determinant of carbon gas exchange [25], was also considered for comparison. Average ES, EA and LAI across the stands were 776 g Cm^-2year^-1, 162 g Cm^-2year^-1 and 4.2 m²m^-2, respectively. In our site, ES and EA scaled up to the stand and the whole year were 1164.2 and 297.0 gCm^-2year^-1, respectively, values 1.5 and 1.8 times higher than those averaged across 15 different stands, suggesting greater carbon losses from soil and stem respiration in the studied Q. pyrenaica coppice. Meanwhile, LAI in our site was 3.8, a value that falls within the reported range for this species [26] and for the genus Quercus [25], and slightly lower than that averaged across different forest types (Table S1).

In conclusion, the relative importance of roots and stems as carbon sinks shifted along the growing season in accordance to their phenology and metabolic activity. The overall respiratory carbon losses from both woody organs were high in comparison with values exhibited by other species. Relatively high values of EA, together with slightly low values of LAI point to an unexpected high importance of aboveground woody respiration in carbon budgets of Q. pyrenaica coppice stands. It also supports the hypothesis of a physiological imbalance between carbon sources and sinks as a causal factor of poor aboveground tree performance in most Q. pyrenaica stands previously coppiced [11,14,16,27]. Nonetheless, temporal patterns of leaf carbon exchange and heterotrophic and autotrophic components of ES should be analyzed in future studies to elucidate the role of root-to-shoot imbalance in Q. pyrenaica coppice decline. A greater number of monitored individuals across several Q. pyrenaica coppices will also provide stronger empirical support to this hypothesis, given that our scaling up from eight stems to the stand level should be taken with caution, and that only one surveyed site hinders statistical comparisons.

We are grateful to Javier Donés for economic and logistical support. We also thank Elena Zafra, Guillermo González, César Otero, Manuel Iglesias, Paula Guzman, Aida Rodríguez, Jose Carlos Miranda, Rosa Ana López, Eva María Miranda and Matías Millerón for their enthusiastic help with field work. This work was funded by the Comunidad de Madrid through CAM P2009/AMB-1668 and P2013/ MAE-2760 projects and by the Organismo Autónomo de Parques Nacionales through PPNN 1148/201 project. Salomón RL was supported by a Ph.D. scholarship from the Universidad Politécnica de Madrid. Jesús Rodríguez-Calcerrada was supported by a Juan de la Cierva contract from the Ministerio de Economía y Competitividad.