Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2013 Nov 7;112(9):1911–1920. doi: 10.1093/aob/mct243

A meta-analysis of cambium phenology and growth: linear and non-linear patterns in conifers of the northern hemisphere

Sergio Rossi 1,*, Tommaso Anfodillo 2, Katarina Čufar 3, Henri E Cuny 4, Annie Deslauriers 1, Patrick Fonti 5, David Frank 5,6, Jožica Gričar 7, Andreas Gruber 8, Gregory M King 5,6, Cornelia Krause 1, Hubert Morin 1, Walter Oberhuber 8, Peter Prislan 3, Cyrille B K Rathgeber 4
PMCID: PMC3838565  PMID: 24201138

Abstract

Background and Aims

Ongoing global warming has been implicated in shifting phenological patterns such as the timing and duration of the growing season across a wide variety of ecosystems. Linear models are routinely used to extrapolate these observed shifts in phenology into the future and to estimate changes in associated ecosystem properties such as net primary productivity. Yet, in nature, linear relationships may be special cases. Biological processes frequently follow more complex, non-linear patterns according to limiting factors that generate shifts and discontinuities, or contain thresholds beyond which responses change abruptly. This study investigates to what extent cambium phenology is associated with xylem growth and differentiation across conifer species of the northern hemisphere.

Methods

Xylem cell production is compared with the periods of cambial activity and cell differentiation assessed on a weekly time scale on histological sections of cambium and wood tissue collected from the stems of nine species in Canada and Europe over 1–9 years per site from 1998 to 2011.

Key Results

The dynamics of xylogenesis were surprisingly homogeneous among conifer species, although dispersions from the average were obviously observed. Within the range analysed, the relationships between the phenological timings were linear, with several slopes showing values close to or not statistically different from 1. The relationships between the phenological timings and cell production were distinctly non-linear, and involved an exponential pattern

Conclusions

The trees adjust their phenological timings according to linear patterns. Thus, shifts of one phenological phase are associated with synchronous and comparable shifts of the successive phases. However, small increases in the duration of xylogenesis could correspond to a substantial increase in cell production. The findings suggest that the length of the growing season and the resulting amount of growth could respond differently to changes in environmental conditions.

Keywords: Cambium, cell differentiation, cell production, climate change, conifers, growth, meristem, phenology, productivity, secondary wall formation, xylogenesis

INTRODUCTION

The timing of seasonal activities, or phenology, is a fundamental aspect of both plant and animal functioning at organization levels ranging from the individual to the ecosystem (Forrest et al., 2010). Phenology is an integrated response to the environmental conditions accomplishing growth and reproduction of species. For this reason, archives of phenological observations have assumed particular importance in the last decades because shifts in phenological events have pointed out changes in the environment (Menzel et al., 2006; Anderson et al., 2013) or a mismatch among species along food chains (Johnson et al., 2010). In plants, the phenology of growth represents a trade-off between environmental constraints and resource availability, and identifies the period when resources are acquired and used (Nord and Lynch, 2009). The growing season in temperature-limited environments, for example, is an optimization between frost damage avoidance and maximization of annual carbon assimilation (Chuine, 2010).

The cambium is the secondary meristem of plants that produces layers of phloem and xylem cells that envelope the wood cylinders of stem, branches and roots, and results in seasonal radial growth. In temperate, boreal and some tropical ecosystems, the cambium undergoes winter dormancy, producing annual tree rings. During development, the xylem undergoes several different biochemical processes throughout its sequential stages of maturation (Prislan et al., 2009). Thus, secondary growth represents an intriguing model of complex phenological processes of single cells that gradually and successively proliferate during a growing season. In contrast to the primary meristems (leaf and flower buds), cambial activity (e.g. cell division and differentiation) occurs within the plants and cannot be directly observed during the growing season.

Species native to colder climates are associated with a short growing season and low growth and productivity. At higher latitudes and altitudes, growth has to be completed within a limited time period and in less favourable conditions than in temperate climates (Chuine, 2010). Species adjust their phenology with shifted or compressed growth and reproduction phases, according to specific regional environmental drivers, local adaptations and individual plasticity to climate (Nord and Lynch, 2009; Diez et al., 2012). However, it is unclear if and how this pattern applies to the cambium and its phenology across taxonomic groups and locations. In addition, changes of environmental conditions could have several effects by modifying the phenology of native species or, in the long run, driving migration towards higher latitudes and altitudes. Accordingly, the timings of phenological events are expected, and have already been in part observed, to change with the ongoing global warming (Menzel et al., 2006). The challenge is to determine to what extent these modifications in phenology are associated with variations in forest productivity, in particular carbon uptake by trees (Forrest et al., 2010).

Biological systems are composed of numerous interacting components that can be characterized by multivariate relationships such as covariation, causality and feedback, that may be directly proportional (i.e. linear) or not. Linear models are the most widely known and applied methods to quantify proportional changes, and have been demonstrated to be a valid representation of several processes, or at least of parts of them (Sparks et al., 2009). However, a linear pattern can be considered as a special case: generally, nature follows more complex relationships, revealing non-linear or even non-monotonic behaviour (Burkett et al., 2005; Porter and Semenov, 2005). Biological relationships may be, or appear to be, approximately linear within a central, bounded interval of variation, but with marked deviations from linearity at the extreme margins of the range (Sparks et al., 2000). Living organisms experience limiting factors that generate shifts and discontinuities, or contain thresholds beyond which responses change abruptly (Stenseth and Mysterud, 2002; Körner, 2012). Coexistence or interaction of factors can have cumulative effects and produce disproportional responses of biological systems (Burkett et al., 2005; Körner and Basler, 2010). The detection and understanding of non-linear relationships of ecological processes benefit from analyses that span a wide enough response range – be it spatial, temporal or system response states – that extend beyond linear regimes.

The questions that this paper aims to answer are: do the phenological phases of xylem production and differentiation occur independently of each other? If any relationships exist among the cambial phenological timings, what is their form (e.g. linear or not) and what does this imply about cambial and growth processes? Answers to these questions could contribute to a more complete understanding of the growth dynamics of forest ecosystems and their possible large-scale responses to climate change.

MATERIALS AND METHODS

Study species and tree selection

Data were collected from nine conifer species at numerous sites distributed in six countries (Table 1). The sites consisted of permanent plots located at different altitudes and latitudes in Europe and Canada and represented temperate-cold to cold environments. In each site, adult trees with upright and injury-free stems were selected for the study. Trees with polycormic stems, partially dead crowns, reaction wood or evident damage due to parasites were preferentially avoided. Between one and 20 trees were sampled per site and year.

Table 1.

Studied species, location of the sites, years of monitoring and number of sampled trees

Species Site Country Latitude Longitude Altitude (m a.s.l.) Study years Number of trees
Abies alba Amance France 48°44′N 06°19′E 270 2006–2007 15–20
Abies alba Grandfontaine France 48°29′N 07°09′E 643 2007–2009 5
Abies balsamea Arvida Canada 48°26′N 71°09′W 80 1999–2000 18
Abies balsamea Liberal 23 Canada 49°58′N 72°30′W 380 1998–2000 10
Abies balsamea Liberal 24 Canada 49°58′N 72°30′W 430 1998–2001 5–10
Abies balsamea Simoncouche Canada 48°13′N 71°15′W 338 2005–2008 10
Larix decidua Cinque Torri 1 Italy 46°27′N 12°08′E 2085 2001–2005 5
Larix decidua Cinque Torri 2 Italy 46°27′N 12°08′E 2156 2002–2005 5
Larix decidua Cinque Torri 3 Italy 46°27′N 12°08′E 2085 2004–2005 5
Larix decidua Lötschental N08 Switzerland 46°23′N 07°45′E 800 2008 4
Larix decidua Lötschental N13 Switzerland 46°23′N 07°45′E 1350 2007–2008 4
Larix decidua Lötschental N16 Switzerland 46°23′N 07°45′E 1600 2007–2008 4
Larix decidua Lötschental N19 Switzerland 46°23′N 07°45′E 1900 2007–2008 4
Larix decidua Lötschental N22 Switzerland 46°23′N 07°45′E 2150 2007–2008 4
Larix decidua Lötschental S16 Switzerland 46°23′N 07°45′E 1600 2007–2008 4
Larix decidua Lötschental S19 Switzerland 46°23′N 07°45′E 1900 2007–2008 4
Larix decidua Lötschental S22 Switzerland 46°23′N 07°45′E 2150 2007–2008 4
Larix decidua Val di Susa Italy 45°03′N 06°40′E 2030 2003–2004 5
Picea abies Cinque Torri 1 Italy 46°27′N 12°08′E 2085 2001–2005 5
Picea abies Cinque Torri 2 Italy 46°27′N 12°08′E 2156 2002–2005 1
Picea abies Cinque Torri 3 Italy 46°27′N 12°08′E 2085 2004–2005 5
Picea abies Grandfontaine France 48°29′N 07°09′E 643 2007–2009 5
Picea abies Lötschental N08 Switzerland 46°23′N 07°45′E 800 2008 4
Picea abies Lötschental N13 Switzerland 46°23′N 07°45′E 1350 2007–2008 4
Picea abies Lötschental N16 Switzerland 46°23′N 07°45′E 1600 2007–2008 4
Picea abies Lötschental N19 Switzerland 46°23′N 07°45′E 1900 2007–2008 4
Picea abies Lötschental S16 Switzerland 46°23′N 07°45′E 1600 2007–2008 4
Picea abies Lötschental S19 Switzerland 46°23′N 07°45′E 1900 2007–2008 4
Picea abies Menina Planina Slovenia 46°16′N 14°48′E 1200 2009–2011 6–8
Picea abies Panska reka Slovenia 46°00′N 14°40′E 400 2009–2011 6
Picea abies San Vito di Cadore Italy 46°26′N 12°13′E 1000 2003 1
Picea mariana Bernatchez Canada 48°51′N 70°20′W 611 2002–2010 5–10
Picea mariana Camp Daniel Canada 50°41′N 72°11′W 487 2002–2010 5–10
Picea mariana Mistassibi Canada 49°43′N 71°56′W 342 2002–2010 5–10
Picea mariana Simoncouche Canada 48°13′N 71°15′W 338 2002–2010 5–10
Pinus cembra Cinque Torri 1 Italy 46°27′N 12°08′E 2085 2001–2005 5
Pinus cembra Cinque Torri 2 Italy 46°27′N 12°08′E 2156 2002–2005 5
Pinus cembra Cinque Torri 3 Italy 46°27′N 12°08′E 2085 2004–2005 5
Pinus cembra Patscherkofel krummholz Austria 47°12′N 11°27′E 2180 2007 5
Pinus cembra Patscherkofel timberline Austria 47°12′N 11°27′E 1950 2007 6
Pinus cembra Patscherkofel treeline Austria 47°12′N 11°27′E 2110 2007 6
Pinus cembra Val di Susa Italy 45°03′N 06°40′E 2030 2003–2004 5
Pinus leucodermis Pollino Italy 39°54′N 16°12′E 2053 2003–2004 10
Pinus sylvestris Grandfontaine France 48°29′N 07°09′E 643 2007–2009 5
Pinus sylvestris Tschirgant dry-mesic Austria 47°14′N 10°50′E 750 2007–2010 5–8
Pinus sylvestris Tschirgant xeric Austria 47°14′N 10°50′E 750 2007–2010 6–8
Pinus uncinata Val di Susa Italy 45°03′N 06°40′E 2030 2003–2004 5
Open in a new tab

Xylem sampling and preparation

Xylem formation was studied at all sites for one or more years during the 1998–2011 period. The different research teams followed the same technique of sample and data collection according to a common protocol. Wood microcores were collected weekly, or occasionally fortnightly, on the stem from 30 cm below to 30 cm above breast height (1·3 m) using surgical bone sampling needles or a Trephor (Rossi et al., 2006a). Samples usually contained several previous xylem rings including the developing annual increment with the cambial zone and the adjacent phloem. The microcores were stored in solutions of propionic or acetic acid mixed with formaldehyde and ethanol at 5 °C to avoid tissue deterioration.

Microcores were dehydrated in ethanol and d-limonene, and embedded in paraffin (Rossi et al., 2006a) or glycol methacrylate (Gruber et al., 2010). The microcores collected in Switzerland were not embedded (Moser et al., 2010). Transverse sections of 10–30 µm thickness were cut from the samples with rotary or sledge microtomes, stained with cresyl violet acetate (Rossi et al., 2006b) or safranin and astrablue (Moser et al., 2010), and examined under bright-field and polarized light to differentiate the developing and mature xylem cells.

Microscopic observations

In each sample, the radial number of cells in the cambial zone, as well as cells in the radial enlargement, secondary cell wall thickening and mature phases were counted along three radial rows. In cross-section, cambial cells were characterized by thin cell walls and small radial diameters (Rossi et al., 2006b). Enlarging cells were composed of a protoplast still enclosed in the thin primary wall but with radial diameter at least twice that of a cambial cell. Observations under polarized light discriminated between enlarging and wall-thickening tracheids. The developing secondary walls shine when observed under polarized light, whereas no glistening is observed in enlarging cells (Abe et al., 1997). Lignification was identified by a colour change from violet to blue with cresyl violet acetate or from blue to red when using safranin and astrablue. Completed colour change over the whole cell wall and lumina free of cell contents marked the end of lignification and tracheid maturity (Gričar et al., 2005).

The number of cells in each developmental phase for every section was averaged over the three rows and used to assess onset and ending of each phase of cell differentiation. In spring, when at least one tangential row of differentiating or mature cells was observed, the differentiation or maturation phase was considered to have begun, respectively. In late summer, when no differentiating cell was observed, xylem formation was considered complete. Cambium phenology was represented by the dates corresponding to (1) the first enlarging cell; (2) the first wall-thickening cell; (3) the first mature cell; (4) the last enlarging cell; and (5) the last wall-thickening cell. The timings of cambium phenology, computed in day of the year (DOY), as well as the number of radial cells produced annually by the cambium were calculated for each tree, site and year. The number of xylem cells varies around the stem and among samples, even if collected on the same day. Thus, cell production was quantified for each tree by fitting a sigmoid curve to the total number of cells produced during the year and estimating the upper asymptote of the function (Deslauriers et al., 2008). The trees showing low goodness of fit or no representative asymptote were excluded from the analysis (29 of 927, representing 3·1 % of all tree–year combinations).

Statistical analyses

Relationships between phenological timings, durations of cell differentiation and cell production were performed using correlations and regressions. Before statistical analyses, variable distributions were inspected for normality. The distribution of the total cell production diverged significantly from normality and data were transformed using the common logarithm. The dates of the first enlarging cell showed a partially asymmetric distribution, but no transformation was applied due to the relatively low skewness.

Bootstrapped Spearman correlations (rS) were calculated by resampling the original data set 10 000 times (Chernick, 2008). Bootstrap calculated robust statistics irrespective of the huge size of the sample, allowing only the relationships really correlated to emerge as significant. The resulting distribution of the coefficient rS was skewed, clearly not normal, and with an upper limit at 1. Thus, a Fisher's z-transformation using the inverse hyperbolic tangent of rS was applied to obtain statistics following a normal distribution. Correlations were considered significant when both 95 % confidence intervals of the distributions were either higher or lower than zero (Chernick, 2008).

Reduced major axis (RMA) regressions were performed when correlations were significant. RMA minimizes the sum of the product of both vertical and horizontal deviations, which is equivalent to the area of triangles formed by the deviation of a point from the line in the vertical and horizontal directions (Smith, 2009). RMA was more appropriate than the conventional ordinary least squares (OLS) method because both dependent and independent variables were measured with error. The bootstrap procedure with 10 000 replications was adopted for computing 95 % confidence intervals of the intercept and slope. For the identification of the dependent and independent variables of the regressions, the temporal rule in which effects, if any, might propagate only from preceding to subsequent phenological phases was assumed. The last cell of the tree ring had to be produced before completing enlargement; accordingly, the radial number of cells within the tree ring was considered as an independent variable in respect to the ending of cell enlargement and wall thickening. However, the assignment of the variables was only a formal procedure, because RMA regression symmetrically defines the bivariate relationship, regardless of which variable is dependent and which is independent (Smith, 2009).

RESULTS

Phenological timings and cell production

The study concerned an analysis of cambium phenology based on >800 tree–year combinations in Europe and Canada. Overall, the first enlarging cells, corresponding to the onset of xylem differentiation, were observed between mid-April (DOY 100) and the end of June (DOY 180), although very occasionally enlargement was observed earlier in Abies alba, Pinus sylvestris and Picea abies (Fig. 1). On average, the first wall-thickening cells were detected 15 d after the onset of the enlargement phase, between the beginning of May and mid-July, with 75 % of observations occurring in June (DOY 150–180). The first mature cells appeared between mid-May and the end of July, on average 17 d after the onset of cell wall thickening (Fig. 1). The earliest onsets of each phenological phase were observed at the lower altitudes, while cell differentiation began later in the sites at the higher altitudes.

Fig. 1.

Fig. 1.

Open in a new tab

Frequency distributions of the dates of cambium phenology (first enlarging, wall-thickening and mature cell, ending of cell enlargement and cell wall lignification) and the radial number of cells produced annually by the cambium of nine conifer species of Europe and Canada. Horizontal boxes represent the upper and lower quartiles, whiskers achieve the 10th and 90th percentiles, and the mean and median are drawn as plus signs and vertical solid lines, respectively. DOY, day of year.

On average, 35 xylem cells were produced by the cambium along a radial row of a tree ring, within a range of 5–144, with most observations (84 %) being between ten and 50 (Fig. 1). The number of cells showed the highest dispersion among the analysed variables, with a coefficient of variation of 62 %, at least five times larger than the phenological timings. The distribution of these data was right-skewed to exponential, with only 5 % of individuals producing >80 cells. In general, the highest cell production was observed in P. abies and A. alba, while fewer cells were produced by Picea mariana and P. sylvestris.

The ending of cell enlargement generally occurred between the beginning of July and mid-September (on average in mid-August), although in 1·6 % of cases cessation of the enlargement phase was observed in June and October (Fig. 1). The last cells in the phase of lignification, which corresponded to the ending of both cell wall thickening and lignification and xylem differentiation, were observed in September, on average 40 d after the conclusion of the enlargement phase. Most trees completed cell maturation in the period between September (DOY 250) and the beginning of October (DOY 280).

Duration of xylem differentiation

Durations of the phases of cell differentiation were extremely variable, as evidenced by the wide range of the distributions (Fig. 2). Cell enlargement lasted 80 d on average, with 81 % of observations between 50 and 110 d. Only 2·4 and 2·5 % of observations showed enlargement phases shorter than 30 d and longer than 150 d, respectively. Picea mariana tended to have the shortest period of cell enlargement. Secondary wall differentiation and the overall duration of xylogenesis lasted on average 105 and 120 d, respectively (Fig. 2). Longer periods of wall thickening and lignification and xylogenesis were observed in A. alba and P. sylvestris in France.

Fig. 2.

Fig. 2.

Open in a new tab

Frequency distributions of the durations of cell enlargement, cell wall thickening and xylogenesis observed in tree rings of nine conifer species of Europe and Canada. Horizontal boxes represent upper and lower quartiles, whiskers achieve the 10th and 90th percentiles, and the mean and median are drawn as plus signs and vertical solid lines, respectively.

Relationship between phenological timings and xylem cell production

The dates of emergence of the first enlarging, wall-thickening and mature cells were strongly correlated with positive and significant rS (Fig. 3). Thus, late onset of cell enlargement was associated with delayed onset of secondary cell wall formation and late emergence of mature cells. Similarly, the ending of cell enlargement was significantly and positively correlated with the ending of cell wall lignification. The models estimated by RMA regressions between these variables showed slopes varying between 0·99 and 1·08 (Table 2). In contrast, no correlation was observed between the phases of onset and ending of differentiation (Fig. 3). Except for the date of the first enlarging cells, all phenological timings were significantly correlated with the logarithm of the number of xylem cells in a radial row, even if with different slopes. The logarithm of cell production was positively correlated with the emergence of the first wall-thickening and mature cells, and negatively correlated with the ending of cell enlargement and lignification. Accordingly, later dates of first observed wall-thickening and mature cells corresponded to lower annual xylem cell production, and greater numbers of cells were observed when the phases of cell enlargement and cell wall lignification were completed later in autumn (Fig. 3).

Fig. 3.

Fig. 3.

Open in a new tab

Correlation matrix between the dates of cambium phenology (first enlarging, wall-thickening and mature cell, ending of cell enlargement and cell wall lignification) and the radial number of cells produced annually by the cambium of nine conifer species of Europe and Canada. The enclosed vertical bars represent the mean and 95 % confidence interval based on the Fisher's z-transformation of the bootstrapped Spearman correlation coefficient calculated by 10 000 replications. The correlation is considered significant when both confidence intervals are either higher or lower than zero. Note the log10 scale on the axis for the number of cells. DOY, day of year.

Table 2.

Intercepts and slopes resulting from the RMA regressions comparing the dates of cambium phenology and the radial number of cells produced annually by the cambium of nine conifer species of Europe and Canada

Dependent variable Independent variable Intercept Slope CI of intercept CI of slope R2
First wall-thickening cell First enlarging cell 15·48 0·99 –10·33 to 39·54 0·83 to 1·17 0·77
First mature cell First enlarging cell 28·32 1·02 –4·88 to 58·87 0·81 to 1·25 0·60
First mature cell First wall-thickening cell 12·44 1·02 –13·81 to 37·39 0·86 to 1·19 0·78
Log10 radial number of cells First wall-thickening cell 3·98 –0·01 2·54 to 5·33 –0·02 to –0·006 0·14
Log10 radial number of cells First mature cell 4·17 -0·01 2·54 to 5·68 –0·02 to –0·006 0·13
Ending of cell enlargement Log10 radial number of cells 119·60 70·68 92·07 to 146·08 52·90 to 89·13 0·30
Ending of cell wall lignification Log10 radial number of cells 151·82 76·34 106·56 to 199·42 44·36 to 106·76 0·19
Ending of cell wall lignification Ending of cell enlargement 22·63 1·08 –35·13 to 80·47 0·81 to 1·34 0·40
Open in a new tab

CI indicates the bootstrapped 95 % confidence interval of the estimated parameters.

The number of cells is transformed using the common logarithm.

Relationship between duration of cell differentiation and cell production

All variables reported in Fig. 4 showed significant correlations with positive rS. Consequently, higher annual xylem cell production (in the form of common logarithm) was associated with longer durations of cell enlargement, wall thickening and xylem formation (Table 3). Similarly, longer periods spent for cell enlargement were associated with longer periods spent for completing cell walls, and, as a consequence, a longer overall duration of xylogenesis. The slopes of the RMA regressions between the durations of differentiation and cell production ranged between 0·99 and 1·02, with 95 % confidence intervals always including 1 (Table 3).

Fig. 4.

Fig. 4.

Open in a new tab

Correlation matrix between the durations of cell enlargement, cell wall thickening and xylogenesis and the radial number of cells produced annually by the cambium of nine conifer species of Europe and Canada. The enclosed vertical bars represent the mean and 95 % confidence interval based on the Fisher's z-transformation of the bootstrapped Spearman correlation coefficient calculated by 10 000 replications. The correlation is considered significant when both confidence intervals are either higher or lower than zero. Note the log10 scale on the axis for the number of cells.

Table 3.

Intercepts and slopes resulting from the RMA regressions comparing the durations of cell differentiation and the radial number of cells produced annually by the cambium of nine conifer species of Europe and Canada

Dependent variable Independent variable Intercept Slope CI of intercept CI of slope R2
Duration of cell enlargement Log10 radial number of cells –70·04 102·37 –108·91 to –32·93 76·25 to 129·51 0·32
Duration of cell wall thickening Log10 radial number of cells –44·52 102·03 –83·61 to –6·25 75·25 to 129·25 0·31
Duration of xylogenesis Log10 radial number of cells –33·23 104·89 –77·33 to 9·85 74·72 to 135·58 0·24
Duration of cell wall thickening Duration of cell enlargement 25·28 0·99 9·09 to 41·32 0·80 to 1·19 0·65
Duration of xylogenesis Duration of cell enlargement 38·53 1·02 23·55 to 53·12 0·84 to 1·21 0·66
Duration of xylogenesis Duration of cell wall thickening 12·53 1·02 0·93 to 23·39 0·92 to 1·13 0·91
Open in a new tab

CI indicates the bootstrapped 95 % confidence interval of the estimated parameters.

The number of cells is transformed using the common logarithm.

DISCUSSION

Linear patterns of phenology

This study regarded a meta-analysis that correlated the phenological timings of xylem and cell production in order to verify the existence of common relationships across nine conifer species living at different latitudes and altitudes in Europe and Canada. The dynamics of xylogenesis were surprisingly homogeneous between the conifer species: despite the wide geographical range considered, the number of studies included and the variability among trees, the dates of successive phenological timings significantly converged towards positive correlations. This result was somehow partially intuitive: time is directional and asymmetric, with effects that can clearly propagate only from preceding towards subsequent events, and not vice versa (Forrest et al., 2010). Thus, a certain phenological phase can take place only after the previous one has occurred. However, more interestingly, regressions showed a novel and unexpected aspect of the relationship between the phenological timings: the models were linear within the range analysed by our data, with several slopes showing values very close to 1. Indeed, it seems that, on average, shifts of one phenological phase lead to comparable shifts of the successive phases. Accordingly, it could be deduced that the average time required for differentiating the first cells of the tree ring is independent of the date of growth resumption, and is estimated as 15 and 17 d for cell enlargement and wall thickening, respectively. Similarly, the last latewood tracheids spend approx. 40 d to complete thickening and lignification of their cell walls. In addition, the durations of cell differentiation and the overall period of xylogenesis are linearly proportional according to the factor 1. These results considered the most general pattern between phenological timings, although dispersions from the average were obviously observed, mainly for the end of differentiation, which indicated the occurrence of other factors affecting the phenology within and between species.

Non-linear patterns of cell production

Correlations and regressions were performed with the variable cell production transformed into the common logarithm. Consequently, the relationships between the phenological timings and cell production were distinctly non-linear, and involved an exponential pattern. Earlier spring phenology and later conclusion of cell differentiation were associated with more than proportional increases in the number of cells. As a result, small increases in the duration of xylogenesis could correspond to a substantial increase in cell production, both within and across species. Higher cell production implies more derivatives to be differentiated, and greater volumes of cell walls to be thickened, the process of wall thickening being longer and more demanding in resources than enlargement (Fig. 2). This necessarily results in higher forest productivity and involves greater carbon assimilation by trees. Similarly, a wider xylem area results in a greater amount of water supplied for producing and maintaining leaves, with a consequent positive feedback of increased potential productivity of plants. These findings suggest that, in temperature-limited ecosystems, the environmental conditions leading to increases in the period of growth could generate disproportionate responses in terms of tree growth.

The divergent patterns observed between phenological timings and cell production could be related to the different mechanism regulating the physiological processes. At the latitudes where the study was carried out, a phenological phase such as cambial resumption is an event triggered by temperature, according to either a gradual influence or a threshold effect (Rossi et al., 2007; Seo et al., 2008; Swidrak et al., 2011). In contrast, xylem growth is a gradual accumulation of cells resulting from an integration of several mitotic processes lasting from a few weeks to several months. Earlier cambial resumptions are expected in sites or in species living where warmer spring temperatures occur. Thus, on the one hand, this allows more time for cambial cells to divide and produce new tracheids. On the other, the more favourable thermal conditions speed up cell production once dormancy has been broken, leading to higher rates of cell division being achieved. The latter mechanism explains most of the variability in cell production (Rathgeber et al., 2011; Cuny et al., 2012). As a consequence, the exponential relationships observed between phenological timings and cell production could be the cumulative result of the length of the period of cambial activity and the rate of cell division. Moreover, the relationship between the length of the growing period and cell production could be complicated by the allocations of carbon observed within the different tree compartments. Trees in colder climates allocate higher proportions of their resources for the metabolism of below-ground organs, thus shifting the growth from stem towards roots (Gower et al., 2001). The observed non-linear patterns could therefore be in part explained because the shorter the growing period is, the higher the proportion of resource allocation to the roots.

The analysis compared species from different ecosystems, with a gradient of phenological timings representing long-term conditions of different growing seasons, here identified by the duration of xylogenesis, and defined by alternation of thermally favourable and unfavourable climate conditions during the year (i.e. summer vs. winter). It could be assumed that the observed reduction in the duration of xylogenesis from 220 to 50 d corresponds to a reduction in the temperatures of the sites. In most sites investigated for this study, the soil is frozen or covered by snow in winter, with temperatures close to 0 °C. At these temperatures, metabolism slows down as much as possible because only minimal physiological activity is allowed, and root activity and water uptake are inhibited (Decker et al., 2003; Körner, 2003). Thermal conditions play a crucial role in decomposing soil organic matter and mineralizing soil nutrients, which affect the availability of nutrients for trees, mainly in the colder environments (Jarvis and Linder, 2000). The acquisition of both mobile and immobile soil resources increases with longer growing seasons (Nord and Lynch, 2009). Besides, snow cover, wintertime nutrient cycling and microbial processes strongly influence the annual transfer of nutrients in plants of cold ecosystems (Groffman et al., 2009; Lupi et al., 2013). Thus, the capture of nutrients from the soil, which are less available in colder sites and for a shorter period of time, could be an important constraint to growth. If the hypothesis of the effect of nutrient availability on cell production is confirmed, the adaptation of the phenological timings and xylem cell production could be asynchronous under a changing climate, at least at the highest latitudes or altitudes: shifts of cell production could be delayed in respect to cambial phenology, and might occur only when the biogeochemical processes (e.g. decomposition and mineralization) are adjusted to the new environmental conditions.

Linking onset and ending of xylogenesis

The diverse correlations observed between the phenological timings revealed the structure of the process of xylem formation and the close interconnection of all differentiation phases, from cell production to maturation. Our findings clearly demonstrated that the phenological phases are not episodic events that occur independently of each other, as also observed by Rossi et al. (2012). Although there is no evidence of a causal mechanism in which a phenological phase necessarily affects or determines the successive one just because it occurs first, the observed correlations suggest that either such a causal link exists or, alternatively, the phenological timings are all related to the same internal or environmental drivers.

Earlier cellular differentiation was correlated with higher annual cell production, which in turn was associated with later maturation of terminal latewood tracheids. In contrast, the onset and ending of cell differentiation were not directly related (the correlations between the dates of the first differentiating cells and those of ending of cell enlargement and lignification were not significant; Fig. 3). It should be noted that onset and ending of xylogenesis involve different parts of the xylem, the first cells of earlywood and the last cells of latewood, respectively. Nevertheless, both phenological events showed significant correlations with total xylem cell production, which suggested the presence of possible common drivers of these phenological events. The results of this study definitively quantify the well-known phenomenon recorded along latitudinal and altitudinal gradients of temperate and cold climates in which reductions in wood production are associated with shorter periods of growth, regardless of species and biome (Rossi et al., 2008; Moser et al., 2010; Lupi et al., 2012). The relationship between duration and amount of growth has been observed both in nature and in manipulation experiments. Cooling treatments reduced the width of tree rings and shortened the period of cambial activity (Gričar et al., 2007). Similarly, roots with larger tree rings (i.e. a higher number of tracheids) than the stem showed later endings of cell wall lignification and thus longer periods of xylogenesis (Thibeault-Martel et al., 2008).

Sparks and Menzel (2002) reported that earlier phenological events are more variable and change faster than later events. In the last decades, most of the reported lengthening of the growing season for both primary and secondary meristems is associated with an earlier onset rather than a delayed ending (Chmielewski and Rötzer, 2001; Boulouf Lugo et al., 2012). An earlier spring phenology is more closely linked to more favourable conditions for photosynthesis than a longer autumnal growing season (Chuine, 2010). Our analysis across species confirmed these hypotheses by showing a greater variability in the spring onset than in the summer–autumn ending of differentiation. The coefficients of variation, dimensionless numbers independent of the mean of the distributions, indicated that the dispersion of the phenological timings gradually reduced from 11·7 for the first enlarging cell to 7·7 for the ending of cell wall lignification (data not shown). If the hypothesis of a chain of intercorrelated phenological events based on causality during xylem formation is confirmed, this would indicate that the ending of cell differentiation represents more than just a direct response of trees to the transitory weather conditions. Ending of cell maturation may represent a conclusive event resulting directly or not from the previous phenological timings and integrating the weather signal over longer periods of time.

Concluding remarks and implications

Cambial activity is an intriguing biological system that in temperate and cold climates generates an annual sequence of phenological processes resulting in an accumulation of xylem cells representing the secondary growth. The phenological timings and xylem cell production of conifers of the Northern hemisphere are correlated. During the growing season and for the ranges observed in this study, the trees adjust their phenological timings according to linear patterns. Thus, shifts of one phenological phase are associated with synchronous and comparable shifts of the successive phases. In contrast, xylem cell production is related to the phenological phases by means of non-linear relationships. Lengthening of the growing season is associated with a more than proportional increase in the number of xylem cells. Thus, modifications of the environmental conditions that affect plant phenology could potentially engender disproportionate responses in secondary growth. Marginal consequences of environmental changes on the length of the growing season could be substantial in terms of cell production and carbon uptake by trees, and consequently forest productivity.

ACKNOWLEDGEMENTS

The results presented in this work were based on projects funded by the Austrian Science Fund (FWF P19563-B16 and P22280-B16), Conseil de Recherches en Sciences Naturelles et en Génie du Canada, Consortium de Recherche sur la Forêt Boréale Commerciale, Fonds de Recherche sur la Nature et les Technologies du Québec, Forêt d'enseignement et de recherche Simoncouche, Slovenian Research Agency (P4-0015 and P4-0107), Swiss National Foundation (INTEGRAL 121859) and UNIFORALL (CPDA110234). The co-operation among authors was supported by the EU COST Action FP1106. The authors thank A. Garside for checking the English text.

LITERATURE CITED

  1. Abe H, Funada R, Ohtani J, Fukazawa K. Changes in the arrangement of cellulose microfibrils associated with the cessation of cell expansion in tracheids. Trees. 1997;11:328–332. [Google Scholar]
  2. Anderson DM, Mauk EM, Wahl ER, et al. Global warming in an independent record of the past 130 years. Geophysical Research Letters. 2013;40 in press. [Google Scholar]
  3. Boulouf Lugo J, Deslauriers A, Rossi S. Duration of xylogenesis in black spruce lengthened between 1950 and 2010. Annals of Botany. 2012;110:1099–1108. doi: 10.1093/aob/mcs175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Burkett VR, Wilcox DA, Stottlemyer R, et al. Nonlinear dynamics in ecosystem response to climatic change: case studies and policy implications. Ecological Complexity. 2005;2:357–394. [Google Scholar]
  5. Chernick MR. Bootstrap methods: a guide for practitioners and researchers. Hoboken, NJ: John Wiley & Sons; 2008. [Google Scholar]
  6. Chmielewski F-M, Rötzer T. Response of tree phenology to climate change across Europe. Agricultural and Forest Meteorology. 2001;108:101–112. [Google Scholar]
  7. Chuine I. Why does phenology drive species distribution? Philosophical Transactions of the Royal Society B: Biological Sciences. 2010;365:3149–3160. doi: 10.1098/rstb.2010.0142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cuny HE, Rathgeber CBK, Lebourgeois F, Fortin M, Fournier M. Life strategies in intra-annual dynamics of wood formation: example of three conifer species in a temperate forest in north-east France. Tree Physiology. 2012;32:612–625. doi: 10.1093/treephys/tps039. [DOI] [PubMed] [Google Scholar]
  9. Decker KLM, Wang D, Waite C, Scherbatskoy T. Snow removal and ambient air temperature effects on forest soil temperatures in northern Vermont. Soil Science Society of America Journal. 2003;67:1234–1243. [Google Scholar]
  10. Deslauriers A, Rossi S, Anfodillo T, Saracino A. Cambium phenology, wood formation and temperature thresholds in two contrasting years at high altitude in Southern Italy. Tree Physiology. 2008;28:863–871. doi: 10.1093/treephys/28.6.863. [DOI] [PubMed] [Google Scholar]
  11. Diez JM, Ibáñez I, Miller-Rushing AJ, et al. Forecasting phenology: from species variability to community patterns. Ecology Letters. 2012;15:545–553. doi: 10.1111/j.1461-0248.2012.01765.x. [DOI] [PubMed] [Google Scholar]
  12. Forrest J, Miller-Rushing AJ. Toward a synthetic understanding of the role of phenology in ecology and evolution. Philosophical Transactions of the Royal Society B: Biological Sciences. 2010;365:3101–3112. doi: 10.1098/rstb.2010.0145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gower ST, Krankina O, Olson RJ, Apps M, Linder S, Wang C. Net primary production and carbon allocation patterns of boreal forest ecosystems. Ecological Applications. 2001;11:1395–1411. [Google Scholar]
  14. Gričar J, Čufar K, Oven P, Schmitt U. Differentiation of terminal latewood tracheids in silver fir trees during autumn. Annals of Botany. 2005;95:959–965. doi: 10.1093/aob/mci112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gričar J, Zupančič M, Čufar K, Oven P. Regular cambial activity and xylem and phloem formation in locally heated and cooled stem portions of Norway spruce. Wood Science and Technology. 2007;41:463–475. [Google Scholar]
  16. Groffman PM, Hardy JP, Fisk MC, Fahey TJ, Driscoll CT. Climate variation and soil carbon and nitrogen cycling processes in a northern hardwood forest. Ecosystems. 2009;12:927–943. [Google Scholar]
  17. Gruber A, Strobl S, Veit B, Oberhuber W. Impact of drought on the temporal dynamics of wood formation in Pinus sylvestris. Tree Physiology. 2010;30:490–501. doi: 10.1093/treephys/tpq003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jarvis P, Linder S. Constraints to growth of boreal forests. Nature. 2000;405:904–905. doi: 10.1038/35016154. [DOI] [PubMed] [Google Scholar]
  19. Johnson DM, Buntgen U, Frank DC, et al. Climatic warming disrupts recurrent Alpine insect outbreaks. Proceedings of the National Academy of Sciences, USA. 2010;107:20576–20581. doi: 10.1073/pnas.1010270107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Körner C. Carbon limitation in trees. Journal of Ecology. 2003;91:4–17. [Google Scholar]
  21. Körner C. Alpine treelines: functional ecology of the global high elevation tree limits. Basel: Springer; 2012. [Google Scholar]
  22. Körner C, Basler D. Phenology under global warming. Science. 2010;327:1461–1462. doi: 10.1126/science.1186473. [DOI] [PubMed] [Google Scholar]
  23. Lupi C, Morin H, Deslauriers A, Rossi S. Xylogenesis in black spruce: does soil temperature matter? Tree Physiology. 2012;32:74–82. doi: 10.1093/treephys/tpr132. [DOI] [PubMed] [Google Scholar]
  24. Lupi C, Morin H, Deslauriers A, Rossi S, Houle D. Role of soil nitrogen for the conifers of the boreal forest: a critical review. International Journal of Plant and Soil Science. 2013;2:155–189. [Google Scholar]
  25. Menzel A, Sparks TH, Estrella N, et al. European phenological response to climate change matches the warming pattern. Global Change Biology. 2006;12:1969–1976. [Google Scholar]
  26. Moser L, Fonti P, Buentgen U, et al. Timing and duration of European larch growing season along altitudinal gradients in the Swiss Alps. Tree Physiology. 2010;30:225–233. doi: 10.1093/treephys/tpp108. [DOI] [PubMed] [Google Scholar]
  27. Nord EA, Lynch JP. Plant phenology: a critical controller of soil resource acquisition. Journal of Experimental Biology. 2009;60:1927–1937. doi: 10.1093/jxb/erp018. [DOI] [PubMed] [Google Scholar]
  28. Porter JR, Semenov MA. Crop responses to climatic variation. Philosophical Transactions of the Royal Society B: Biological Sciences. 2005;360:2021–2035. doi: 10.1098/rstb.2005.1752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Prislan P, Koch G, Cufar K, Gricar J, Schmitt U. Topochemical investigations of cell walls in developing xylem of beech (Fagus sylvatica L.) Holzforschung. 2009;63:482–490. [Google Scholar]
  30. Rathgeber CBK, Rossi S, Bontemps J-D. Tree size influences cambial activity in a mature silver fir plantation. Annals of Botany. 2011;108:429–438. doi: 10.1093/aob/mcr168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Rossi S, Anfodillo T, Menardi R. Trephor: a new tool for sampling microcores from tree stems. IAWA Journal. 2006a;27:89–97. [Google Scholar]
  32. Rossi S, Deslauriers A, Anfodillo T. Assessment of cambial activity and xylogenesis by microsampling tree species: an example at the Alpine timberline. IAWA Journal. 2006b;27:383–394. [Google Scholar]
  33. Rossi S, Deslauriers A, Anfodillo T, Carraro V. Evidence of threshold temperatures for xylogenesis in conifers at high altitude. Oecologia. 2007;152:1–12. doi: 10.1007/s00442-006-0625-7. [DOI] [PubMed] [Google Scholar]
  34. Rossi S, Deslauriers A, Gričar J, et al. Critical temperatures for xylogenesis in conifers of cold climates. Global Ecology and Biogeography. 2008;17:696–707. [Google Scholar]
  35. Rossi S, Morin H, Deslauriers A. Causes and correlations in cambium phenology: towards an integrated framework of xylogenesis. Journal of Experimental Botany. 2012;63:2117–2126. doi: 10.1093/jxb/err423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Seo J-W, Eckstein D, Jalkanen R, Rickebusch S, Schmitt U. Estimating the onset of cambial activity in Scots pine in northern Finland by means of the heat-sum approach. Tree Physiology. 2008;28:105–112. doi: 10.1093/treephys/28.1.105. [DOI] [PubMed] [Google Scholar]
  37. Smith RJ. Use and misuse of the reduced major axis for line-fitting. American Journal of Physical Anthropology. 2009;140:476–486. doi: 10.1002/ajpa.21090. [DOI] [PubMed] [Google Scholar]
  38. Sparks TH, Menzel A. Observed changes in seasons: an overview. International Journal of Climatology. 2002;22:1715–1725. [Google Scholar]
  39. Sparks TH, Jeffree EP, Jeffree CE. An examination of the relationship between flowering times and temperature at the national scale using long-term phenological records from the UK. International Journal of Biometeorology. 2000;44:82–87. doi: 10.1007/s004840000049. [DOI] [PubMed] [Google Scholar]
  40. Sparks TH, Jaroszewicz B, Krawczyk M, Tryjanowski P. Advancing phenology in Europe's last lowland primeval forest: non-linear temperature response. Climate Research. 2009;39:221–226. [Google Scholar]
  41. Stenseth NC, Mysterud A. Climate, changing phenology, and other life history traits: nonlinearity and match–mismatch to the environment. Proceedings of the National Academy of Sciences, USA. 2002;99:13379–13381. doi: 10.1073/pnas.212519399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Swidrak I, Gruber A, Kofler W, Oberhuber W. Effects of environmental conditions on onset of xylem growth in Pinus sylvestris under drought. Tree Physiology. 2011;31:483–493. doi: 10.1093/treephys/tpr034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Thibeault-Martel M, Krause C, Morin H, Rossi S. Cambial activity and intra-annual xylem formation in roots and stems of Abies balsamea and Picea mariana. Annals of Botany. 2008;102:667–674. doi: 10.1093/aob/mcn146. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES