Skip to main content

Does natural root grafting make trees better competitors?

Cite this dataset

Quer, Elodie; Helluy, Manon; Baldy, Virginie; DesRochers, Annie (2022). Does natural root grafting make trees better competitors? [Dataset]. Dryad.


Natural root grafts (anastomoses) are morphological unions formed between roots of different trees. Common root systems allow translocation of water, nutrients and photosynthesis products between grafted trees, affecting their growth and their physiology. As carbohydrates are redistributed among grafted trees, the formation of a common root system could reduce the negative effect of intraspecific competition for light or soil resources within stands. The aim of this study was to investigate the role of root grafting on intraspecific competition and growth of balsam fir (Abies balsamea (L.) Mill). We studied inter-tree relationships in three natural balsam fir stands of the boreal forest of Quebec (Canada) that contained an average 36% of grafted trees. At each stand, ring width and basal area of trees were measured using dendrochronology techniques. We used mixed linear models to test the effect of root grafting and intraspecific competition on annual basal area increment of trees. Trees before grafting had higher growth rates than trees once grafted. Thus, root grafting did not improve tree growth. Growth of grafted trees was more negatively affected by intraspecific competition than growth of non-grafted trees. Thus, grafted trees cannot be considered as better competitors than non-grafted trees. Under high intraspecific competition, growth of larger grafted trees was less affected than that of smaller trees suggesting that they were able to divert resources at their advantage within a union. Our study demonstrated that grafted trees acted on each other’s growth and provides support for the idea that grafted trees respond to competition for resources more as a community rather than as individual trees.


Field work was performed in summer 2017. At each site, distance between trees were measured then all trees were felled with a chainsaw and cross-sectional disks were collected at ground level (0 m). Cross-sectional disks were air dried and sanded (80-400 grit) to reveal growth rings. Tree ring widths series were measured on cross-sectional disks from the stem base using a Velmex “TA Unislide” measurement system with ACU-Rite linear encoder and QC10V digital readout device (Velmex, Inc., Bloomfield, New York) interfaced with Measure J2X (Version 5.0x) ring-reading software (Project J2X, Voortech Consulting, Holderness, NH). Ring-width series were measured with 0.001 mm precision along two rays per wood disk when possible (i.e. when the rings were clearly visible) and visually cross-dated. To obtain a single series per tree, chronologies were averaged for each tree when two of them were available. A total of 110 tree ring width series were obtained.

We converted tree ring width series into basal area increment (BAI) to calculate annual tree growth (Biondi and Qeadan 2008) following equation (1) :

BAIt = πRt2- πR2t-1 (1)


Where Rt is the stem radius at year t and Rt-1 is the stem radius at year t-1.

Intraspecific competition between trees was estimated using distance-independent (BAL: Basal area of the Larger tree) and distance-dependent (HEG: Hegyi competition index) competition indices. The first index (BAL) was developed by Wykoff et al. (1982) and corresponds to the sum of the of the basal area of neighbor trees that are larger than the subject tree: The second index (HEG) was developed by Hegyi (1974) and corresponds to the sum of the ratios of diameters of a subject tree and its competitors, weighted by the distance from the subject tree. We considered that each tree found in a radius of 6 m or less around the subject tree was a potential competitor.

Usage notes

Excel software is required to open the data files (.csv).


Natural Sciences and Engineering Research Council, Award: RGPIN-2017-05532