Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2004 Jun;93(6):671–680. doi: 10.1093/aob/mch094

Seasonal Branch Nutrient Dynamics in Two Mediterranean Woody Shrubs with Contrasted Phenology

RUBÉN MILLA 1,*, M MAESTRO‐MARTÍNEZ 1, G MONTSERRAT‐MARTÍ 1
PMCID: PMC4242299  PMID: 15072979

Abstract

• Background and aims Mediterranean woody plants have a wide variety of phenological strategies. Some authors have classified the Mediterranean phanaerophytes into two broad phenological categories: phenophase‐overlappers (that overlap resource‐demanding activities in a short period of the year) and phenophase‐sequencers (that protract resource‐demanding activities throughout the year). In this work the impact of both phenological strategies on leaf nutrient accumulation and retranslocation dynamics at the level of leaves and branches was evaluated. Phenophase‐overlappers were expected to accumulate nutrients in leaves throughout most of the year and withdraw them efficiently in a short period. Phenophase‐sequencers were expected to withdraw nutrients progressively throughout the year, without long accumulation periods.

• Methods To test this hypothesis, variations in phenology and leaf NPK in the crown of a phenophase‐overlapper Cistus laurifolius and a phenophase‐sequencer Bupleurum fruticosum were monitored monthly during 2 years.

• Key Results Changes in nutrient concentration at the leaf level were not clearly related with the different phenologies. Nitrogen and phosphorous resorption efficiencies were lower in the phenophase‐overlapper, and accumulation–retranslocation seasonality was similar in both species. Changes in the branch nutrient pool agreed with the hypothesis that the phenophase‐overlapper accumulated nutrients from summer until the bud burst of the following spring, recovering a large nutrient pool during massive leaf shedding. The phenophase‐sequencer did not accumulate nutrients from autumn until early spring, achieving lower nutrient recovery during spring leaf shedding.

• Conclusions It is concluded that phenological demands influence branch nutrient cycling. This effect is easier to detect by assessing changes in the branch nutrient pool rather than changes in the leaf nutrient concentration.

Key words: Phenology, Mediterranean climate, phenophase‐sequencers, phenophase‐overlappers, nutrients, resorption efficiency, Cistus laurifolius, Bupleurum fruticosum

INTRODUCTION

The seasonal arrangement of life cycle events (phenophases) is important for the survival and reproductive success of plants (Mooney, 1983; Rathcke and Lacey, 1985). This is especially the case in strongly seasonal climates like the Mediterranean‐type, where two unfavourable seasons (summer drought and winter cold) alternate with two favourable periods (spring and early autumn) (Mitrakos, 1980). Phenological adjustment to contrasted seasonality has been considered a symptom of proper acclimation (Mooney and Dunn, 1970; Orshan, 1989). However, a wide variety of phenological patterns have been identified among Mediterranean woody perennials (e.g. Orshan, 1989; De Lillis and Fontanella, 1992; Cabezudo et al., 1993; Pérez Latorre and Cabezudo, 2002).

To simplify the observed diversity, some authors have classified species according to two types: phenophase‐overlappers (species that tend to concentrate their phenological activity in a short, favourable period, allowing overlapping among phenophases) and phenophase‐sequencers (species that tend to protract their activity throughout the year, avoiding overlapping) (Mooney and Kummerow, 1981; Mooney, 1983; Castro‐Díez and Montserrat‐Martí, 1998). Both strategies represent the extremes of a trade‐off between the advantages of growing in the optimal period and the disadvantages of facing intra‐plant competition between different resource sinks.

Phenological events like vegetative growth, flowering or fruiting are resource‐demanding processes (Oliveira et al., 1996; Bazzaz and Grace, 1997; Rosecrance et al., 1998; Orgeas et al., 2002), and may compete for resource investments with other sinks, e.g. formation of reserves (Chapin et al., 1990). In seasonal climates, storage has been considered crucial to uncouple resource availability from resource use (Bloom et al., 1985; Millard, 1996). If part of the nutrient demands for growth is supplied by reserves, the dynamics of nutrient storage organs may be expected to be strongly affected by the phenological calendar of the plant. However, as far as is known, the impact of the above‐defined phenological strategies on nutrient storage dynamics has not been explored.

To examine the interaction between phenology and nutrient use, an important nutrient storage structure should be selected for study. Some seminal papers advocate the function of leaves as nutrient storage organs in perennial evergreens living in nutrient‐poor environments (Loveless, 1961, 1962; Monk, 1966; Small, 1972; Reader, 1978). Although recent research has highlighted that perennial leaves are multifunctional organs that connect the economy of different plant resources (Harper, 1989; Kikuzawa, 1995; Aerts, 1996; Givnish, 2002), the nutritional significance of the evergreen habit cannot be ignored (Aerts and Chapin, 2000, and references therein). Therefore, leaves were selected as nutrient storage organs and the impact of plant phenology evaluated on leaf nutrient accumulation and retranslocation dynamics.

It was expected that the nutrient demands of growing shoots would compete with nutrient accumulation in old leaves. In this way, leaves from phenophase‐overlappers would behave as nutrient storage organs throughout most of the year, sharply withdrawing nutrients in the short growth period; whereas leaves of the phenophase‐sequencers should not exhibit long accumulation periods (hypothesis 1). As a consequence, higher nutrient levels in the leaves of phenophase‐overlappers was expected before the onset of leaf senescence (as compared with phenophase‐sequencers), to make the nutrient resorption process more efficient (hypothesis 2).

To test these hypotheses, two Mediterranean shrubs, Cistus laurifolius and Bupleurum fruticosum, with contrasting phenology were selected. Their nutrient dynamics in leaves and branches, and their phenology were monitored over two consecutive years.

MATERIALS AND METHODS

Study species

Cistus laurifolius L. and Bupleurum fruticosum L. are evergreen monoecious shrubs, 1·5–2·5 m high, with a similar leaf area and leaf habit (see Table 1). On the basis of previous field knowledge of the species, we selected C. laurifolius as a phenophase‐overlapper, with lateral inflorescences that burst in late spring simultaneously with the peak in vegetative growth. Bupleurum fruticosum was selected as a phenophase‐sequencer, with terminal inflorescences that appear at the end of the vegetative shoot, therefore avoiding overlap between vegetative and reproductive shoot growth.

Table 1.

Leaf traits and nutrient resorption parameters in the studied species

n C. laurifolius B. fruticosum
Leaf traits
 La (mm2) 25 785·53 (24·58)a 648·13 (25·01)b
 Lma (g m–2) 25 215·62 (3·69)a 135·06 (2·21)b
 Lt (mm) 25 0·48 (0·01)a 0·31 (0·00)b
 Ld (kg m–3) 25 447·97 (9·90)a 432·69 (4·60)a
 LLMEAN (years) 8 1·06 (0·03)a 0·99 (0·12)a
 LLMAX (years) 8 1·17 (0·00)a 1·58 (0·11)b
Nutrient resorption parameters
 N‐reff (ratio) 25 0·61 (0·01)a 0·64 (0·03)b
 P‐reff (ratio) 25 0·62 (0·01)a 0·65 (0·00)b
 K‐reff (ratio) 25 0·54 (0·01)a 0·34 (0·03)b
 Nsen (g m–2) 25 1·29(0·01)a 1·00 (0·02)b
 Psen (g m–2) 25 0·13 (0·00)a 0·05 (0·00)b
 Ksen (g m–2) 25 0·60 (0·01)a 1·49 (0·03)b
 N‐rPOOLBR (mg branch–1) 8 178·08 (12·42)a 25·18 (3·32)b
 P‐rPOOLBR (mg branch–1) 8 13·73 (0·96)a 0·91 (0·11)b
 K‐rPOOLBR (mg branch–1) 8 62·57 (4·36)a –0·93 (0·20)b
Open in a new tab

Values are mean (s.e.m.).

Different letters in the same raw mean significant differences at P = 0·01.

Study area and study period

The study was carried out from January 1999 to December 2000. One natural population was selected per species in the western part of the pre‐Pyrenean mountain range foothills, north‐east Spain. The B. fruticosum site (Zaragoza, Orés, monte El Fragal, 760 m, UTM 30TXM6682) was situated on the south‐south‐west slope of a limestone hill. The population of C. laurifolius (Zaragoza, Luesia, Arba river watershed, 780 m, UTM 30TXM6397) was 18 km away from the B. fruticosum site, on a west‐orientated slope of a hill made up of calcareous conglomerates.

At both sites, three horizon A soil samples were taken before the growing season. Soil pH, organic matter, total nitrogen (N), phosphorus (P) (Olsen) and potassium (K) (ammonium acetate), were measured in every sample (Table 2). Taken together, the nutrient contents were quite low in both plots, as compared with other Mediterranean sites (Specht, 1969; Rapp et al., 1999).

Table 2.

Soil chemical characteristics of the study sites

pH O.M. (%) Total N (%) P Olsen (ppm) Total K (ppm)
C. laurifolius site (n = 3) 7·69 (0·06) 3·88 (0·20) 0·23 (0·02) 1·53 (0·86) 152 (18·00)
B. fruticosum site (n = 3) 8·48 (0·12) 3·16 (1·11) 0·08 (0·04) 1·42 (0·27) 142 (24·2)
Open in a new tab

Values are mean (SEM).

Data from the nearest weather station (Ayerbe, 580 m) were used to describe temperature and rainfall patterns at the study sites. The climate is typically Mediterranean with a cold winter due to the continental character of the area. The weather in 1999 was quite wet (818 mm), with considerable summer precipitation, and almost no precipitation during winter, mean annual temperature being 12·7 °C. In the year 2000, rainfall was normal (687 mm) and mostly in spring and autumn, and the mean annual temperature was similar to that of 1999 (12·2 °C).

Phenological method

To obtain a reliable phenological diagram the semi‐quantitative method described in Montserrat‐Martí and Pérez‐Rontomé (2002) was followed. This procedure is based in Orshan’s qualitative method (Orshan, 1989) modified to quantify the frequency of phenophases within the population on three levels (I, phenophase in >25 % of the plants; II, phenophase in 5–25 %; III, phenophase in <5 %). Ten adult individuals were monitored on a monthly basis to assess the intensity of dolichoblast vegetative growth (DVG), flower bud formation (FBF), flowering (FLO), fruit setting (FS), seed dispersal (SD) and leaf shedding of dolichoblasts (LSD). Cistus laurifolius bears long (dolichoblast) and short (brachyblast) shoots, so two additional phenophases were assessed in this species: brachyblast vegetative growth (BVG) and leaf shedding of brachyblasts (LSB). In addition to the data from the ten individuals, more information from the rest of the population was used to draw the phenological diagrams. The degree of overlapping between vegetative and reproductive phenophases was quantified in each of the ten marked individuals using two indexes: the Phenophase Sequence Index (PSI) (Castro‐Díez and Montserrat‐Martí, 1998) with values ranging from 0·33 to 1:

PSI = t (DVG + FBF + FLO)/[t(DVG) + t(FBF) + t(FLO)]

and the Phenophase Sequence Index of Fructification (PSIF) (Montserrat‐Martí and Pérez‐Rontomé, 2002) with values ranging from 0·5 to 1:

PSIF = t(DVG + FS)/[t(DVG) + t (FS)]

t being the number of months required to complete the phenophase(s) in parentheses. High index values indicate a protracted arrangement of phenophases and lower ones indicate a higher degree of overlap.

Leaf survivorship and leaf longevity

Leaf demography was monitored during 2000 to obtain leaf birth and shedding calendars. Ten well‐developed 2‐year‐old branches were randomly selected at the mid‐crown of the marked individuals. Green, senescent and dry leaves were carefully counted and recorded in a branch drawing for every sampling date. Length of current‐year shoots was also recorded, and used to estimate the elongation period of dolichoblasts and brachyblasts.

Leaf demography monitoring data were further used (a) to calculate mean (LLMEAN) and maximum (LLMAX) leaf longevities (in years) [an average leafing date for each species was assumed, and the number of years from that leafing date until 50 % (LLMEAN) and 100 % (LLMAX) of the leaf cohort fell was counted]; and (b) to compute leaf nutrient content of the whole branch (see below).

Sampling of plant material

During every field visit, 15 well‐developed 2‐year‐old, sun‐orientated branches from 15 individuals per species were randomly collected at the mid‐crown. Current‐year and 1‐year‐old (when present) leaves were separated in the laboratory, and each fraction was pooled to get a unique composite sample per leaf cohort and species. In the period of maximum leaf abscission, fully senesced leaves (senescent leaves easily detached with a gentle touch), were also harvested. Leaves were oven‐dried to a constant weight at 60 °C. The N concentration was measured with an elemental analyser (Elementar varioMAX N/CN, Hanau, Germany), P concentration was determined by vanado‐molybdate colorimetry (Allen et al., 1976), and K content was measured with a flame photometer.

To determine leaf morphological parameters, 25 additional leaves per cohort and species were also harvested every sampling date. Leaf area (La) was measured with a Delta‐T Image Analysis System (Delta‐T Devices Ltd, Cambridge, UK). Leaf thickness (Lt) was measured with a caliper, avoiding the main leaf veins. Leaves were oven‐dried to a constant weight at 60 °C and their leaf mass per area (Lma) was calculated by dividing their dry weight by La. Leaf density (Ld) was determined by multiplying Lma × Lt–1. Average values of mature leaves are presented in Table 1.

Nutrient calculations

Leaf nutrient data were calculated on the basis of mass (mg g–1) and area (g m–2). Leaf nutrient content of the whole branch (NBR, mg) was computed according to the following formula:

NBR = nLwNw

where n is the number of leaves per marked branch of leaf demography, Lw is the average dry weight of a leaf, and Nw is the mass‐based concentration of N, P or K. NBR was calculated monthly from January to December 2000.

Nutrient resorption from senescing leaves was estimated at both leaf and branch levels. Leaf level resorption efficiencies (reff, ratio) were computed as:

reff = (Nmax – Nsen)/Nmax

where Nmax is the maximum nutrient content (on an area basis) before the onset of senescence and Nsen is the nutrient content in senesced leaves.

Using the nutrient content of the whole branch (NBR) values described above, the amount of nutrients recovered before leaf abscission per 2‐year‐old branch unit (rPOOLBR, mg) was calculated as:

rPOOLBR = NMAXBR – NSENBR

where NMAXBR is the maximum leaf nutrient pool per branch before the onset of senescence, and NSENBR is the amount of nutrients shed from the branch due to leaf abscission.

Statistics

t‐Tests were performed to assess differences between the two species regarding leaf traits and retranslocation parameters. A two‐way ANOVA was used to compare PSI and PSIF values between species and years. Phenophase sequence indexes and K‐reff values were arcsin[sqrt(x)] transformed before comparisons to achieve normality and homoscedasticity. Analysis were performed using SPSS 11.0 software.

RESULTS

Phenological patterns and phenophase overlapping

The phenological diagrams of both species are shown in Fig. 1. Cistus laurifolius exhibited BVG almost all year round. Although the long BVG period apparently discards C. laurifolius as a phenophase‐overlapper, some considerations must be taken into account. First, brachyblasts in this species are lateral sylleptic branches of dolichoblasts, and their elongation period is synchronous with that of dolichoblasts (see Fig. 1); thus, although brachyblasts contribute significantly to the leaf biomass, they produce most of their leaves at the same period that dolichoblasts do. Second, brachyblasts show a remarkable heteroblasty, producing big leaves during spring with very small ones (most of them hypsophyll‐like) emerging during summer and autumn. Third, the frequency of shoots carrying out BVG in an individual crown out from the spring period was quite low, although BVG was detected in every one of the ten marked individuals almost all year round, our phenological method did not allow us to assess the frequency of the phenophases in the crown. Therefore, we considered that the DVG period comprised nearly all of the vegetative biomass production by C. laurifolius, and when comparing both species we did not consider BVG.

graphic file with name mch094f1.jpg

Open in a new tab

Fig. 1. Phenophase diagrams of C. laurifolius and B. fruticosum. DVG, dolichoblast vegetative growth; BVG, brachyblast vegetative growth; FBF, flower bud formation; FLO, flowering; FS, fruit setting; SD, seed dispersal; LSD, leaf shedding of dolichoblasts; LSB, leaf shedding of brachyblasts. The levels of frequency in the population are indicated by double line (level I); single line (level II); and broken line (level III). The lengths of current‐year shoots per one 2‐year‐old branch per individual (ten individuals) are plotted below the phenophase diagrams; in C. laurifolius, dolichoblasts and brachyblasts are plotted separately. See Materials and Methods for more details.

The main resource‐demanding phenophases (DVG, FBF, FLO and FS) started earlier and finished later in B. fruticosum. Both vegetative and reproductive activities were more protracted in this species, maintaining important growth activities from January–February to mid‐November. In C. laurifolius almost all the branch growth took place in a short period between the end of the spring and the beginning of the summer. The LSD period was much longer in B. fruticosum. Differences between mean and maximum leaf longevities also underlined the protracted leaf‐shedding pattern in this species.

PSI and PSIF values are presented in Table 3. Both species differed strongly in the degree of overlap between vegetative and reproductive growth (PSI:Fspecies = 359·9, P < 0·001; PSIF:Fspecies = 5294·7, P < 0·001). Cistus laurifolius grew its vegetative and reproductive dolicoblasts at the same time, therefore its PSI was much lower than B. fruticosum. The latter delayed the development of inflorescence until the end of vegetative shoot growth. Fruit growth was also very fast in C. laurifolius, where fecundation and ovary growth started during vegetative development, and the fruit matured only 15 d after the arrest of dolichoblast growth. On the other hand, B. fruticosum postponed the beginning of fruit growth until August, completely avoiding overlap with vegetative growth.

Table 3.

Mean (s.e.m.) values of PSI and PSIF for B. fruticosum and C. laurifolius in 1999 and 2000

P SI P SIF
1999 C. laurifolius 0·50 (0·00)a 0·67 (0·00)a
B. fruticosum 0·97 (0·03)b 1·00 (0·00)b
2000 C. laurifolius 0·50 (0·00)a 0·68 (0·04)a
B .fruticosum 0·98 (0·02)b 1·00 (0·00)b
Open in a new tab

Different letters in the same column indicate a significant (P < 0·01) F value in a two‐way ANOVA (species × year) analysis for the species factor.

Leaf emergence and shedding patterns

In Fig. 2, leaf birth and shedding percentages per sample date are plotted. Leaf emergence started earlier in B. fruticosum, but stopped when reproductive growth was initiated (see Fig. 1). Cistus laurifolius exhibited a more sequential leaf emergence; however, 75 % of the total leaf production was accomplished during FBF, FLO and FS (from early May to late July). In the later species, there was a remarkable (20 %) leaf emergence during late summer and early autumn; as noted in the previous section, these brachyblast late‐emerging leaves are very small in comparison with those borne during spring and early summer and do not contribute significantly to the leaf biomass pool per branch.

graphic file with name mch094f2.jpg

Open in a new tab

Fig. 2. Leaf birth and shedding curves. Data are sampling‐date percentages relative to the total amount of emerged or shed leaves during the whole year. Values are mean ± s.e.m.; n = 10 branches per species.

Leaf fall was more gradual in B. fruticosum, this species sheds overwintering 1‐year‐old leaves during spring, and some of the current‐year ones during summer and autumn (Fig. 2), although occasional longer leaf longevities (nearly 2 years) were also observed within the population. In contrast, C. laurifolius shed its leaves during late‐spring and early‐summer; although most of the shed leaves were 1‐year‐old, during July some of the fallen leaves were current‐year ones, mainly small leaves from the lower nodes of the expanding dolichoblasts, and probably shed because of stronger sink‐strength of the upper leaves in the growing shoot.

Seasonality of nutrient content variations and nutrient resorption from senescing leaves

Figure 3 shows the variations in leaf nutrient content at both organ and branch levels. Nutrient data are plotted on a leaf age scale. Nutrient patterns at the leaf level (Fig. 3A) differed between area (g m–2) and mass (mg g–1) basis for N and P. Regarding N and P variations, three important facts should be highlighted. First, the maximum nutrient content before spring resorption occurred earlier in B. fruticosum (December–January) than in C. laurifolius (May–July for N and March for P). Second, both species kept N and P mass‐based levels constant, or slightly decreased, from the end of leaf expansion (July–August) until the maximum in late winter–early spring. Third, area‐based series tended to enhance decreasing or constant tendencies in mass‐based series. As a consequence, area‐based N and P data showed constant or increasing tendencies over the leaf life‐span. Changes in K were very different between species. An early K maximum was detected in B. fruticosum during summer, after which it was exported continuously until the next spring. In C. laurifolius there was no summer peak and the K maximum was the following spring, prior to leaf senescence and shedding.

graphic file with name mch094f3.jpg

Open in a new tab

Fig. 3. Leaf NPK dynamics along the life span of the leaves of C. laurifolius (circles) and B. fruticosum (triangles). (A) Graphs of leaf nutrient concentration on mass (mg g–1) and area (g m–2) basis; (B) graphs of leaf nutrient pools per branch unit (mg branch–1). In part B, brachyblast and dolichoblast leaves were summed for C. laurifolius.

The branch patterns plotted in Fig. 3B reflect the nutrient pool in a leaf cohort from bud‐burst to the shedding date of the last leaf. These patterns are dependent on leaf nutrient concentration (Fig. 3A), biomass gain in individual leaves (Fig. 4), and leaf emergence and shedding dynamics (Fig. 2). The patterns were quite different between species. The nutrient pools in B. fruticosum reached their maximum 3 months after bud‐burst, then progressively decayed, without autumn or winter accumulation. This pattern was remarkably sharper for K pools, probably because of the early K maximum in leaf concentration. In contrast, the pattern in C. laurifolius exhibited a drop in the nutrient pools from bud‐burst until the end of the summer. This drop was probably due to early leaf shedding (see above); however, it cannot be discounted that some nutrients on recently borne leaves could be diverted towards nutrient‐demanding structures such as inflorescences, thus operating resorption could account for that drop as well. From this summer minimum, nutrient pools increased until the beginning of the final massive leaf shedding the following spring. Thus, nutrient accumulation in C. laurifolius extended for 7–8 months.

graphic file with name mch094f4.jpg

Open in a new tab

Fig. 4. Leaf mass per area (Lma) along the life‐span of the 1999 leaf cohort in the studied species. Values are mean ± SEM; n = 25 leaves per species

The N‐reff and P‐reff values in Table 1 were similar to the average ratios compiled by Aerts (1996) and corrected by van Heerwaarden et al. (2003) for perennials, being slightly higher for B. fruticosum. The K‐reff values are lower than N‐reff and P‐reff, both for C. laurifolius and B. fruticosum, the latter being markedly lower because most of the resorption was carried out from the previous summer until winter (see Fig. 3). The rPOOLBR values are also shown in Table 1. The amount of N, P and K recovered and available for spring growth or storage was much higher in C. laurifolius branches. Both a higher nutrient accumulation during autumn–winter and a higher leaf survival until the following spring could account for the higher nutrient pool before spring leaf senescence (Fig. 3), that was partly recovered (54–62 %) before leaf shedding.

DISCUSSION

Annual patterns of nutrient depletion and accumulation in leaves as affected by plant phenology

The seasonal variations in N and P concentrations in the leaves of B. fruticosum and C. laurifolius resemble those published for other Mediterranean woody plants (Mooney and Rundel, 1979; Oliveira et al., 1996; Fernández‐Escobar et al., 1999; Cherbuy et al., 2001). Three phases can be identified during the life‐span of Mediterranean evergreen leaves: (1) an initial nutrient drop phase during dolichoblast growth and leaf mass gain (sharper in mass‐based series); (2) a steady period with gentle variations towards accumulation or resorption during mature leaf life‐span; and (3) a final drop before leaf shedding. This general sequence can be modified by growth activities and climate variability (Fernández‐Escobar et al., 1999). The drop in the first phase is probably due to dilution of nutrients during leaf expansion; however, as some recently borne leaves are shed during this phase (see Results for C. laurifolius) and there is resorption associated with sequential leaf development (Jonasson and Chapin, 1985), there could be remobilization of nutrients from early‐season to late‐season emerged leaves or to reproductive buds within the growing shoot. Growth demands may have induced the final drop in leaf nutrients in the species considered. This drop began earlier in B. fruticosum (phenophase‐sequencer). Since nutrients in recently borne leaves in perennial species are mainly supplied by internal reserves (Millard, 1994; Mediavilla and Escudero, 2003), the earlier drop is probably related to the earlier bud‐burst in B. fruticosum. It is quite remarkable that nutrients, calculated on an area basis, showed increasing tendencies when mass‐based series showed steady tendencies, or steady tendencies when mass‐based series decreased. This was specially the case in N and P for C. laurifolius. As mass gain per unit leaf area grew until nearly leaf senescence (see Fig. 4) in C. laurifolius, nutrients were accumulated in the leaf without drastic changes in nutrient concentration. The K dynamics differed between species, especially at the dilution phase. K was accumulated in the leaves of B. fruticosum during the first months of the leaf life‐span. Since K is a strong osmolyte (Hsiao and Läuchli, 1986; Gucci et al., 1997), an increase in leaf K could be related with (a) a need for osmotic adjustment, which could be part of a water‐saving strategy (sensu Lo Gullo and Salleo, 1988) to avoid water stress during summer drought; or (b) a need for water‐lifting towards developing infloresences. The elongation of reproductive stems in B. fruticosum is carried out during the summer drought; thus, reproductive shoots would probably be supplied with water by osmotic water lifting (Boyer, 1988; Hsiao and Xu, 2000; Zimmermann et al., 2002), and K is likely to be partly responsible for the osmotic gradient. In contrast, C.laurifolius exhibits maxima of branch K contents prior to leaf fall and thus concurrent with the following season’s vegetative and reproductive growth. Although soil water availability is moderate during that period, the establishment of a strong osmotic gradient in the soil–root–shoot system is probably facilitating water supply for the simultaneous growth activities of C. laurifolius during late spring and early summer. Nevertheless, more research is needed to understand the interesting differences in K dynamics between phenological types.

The variations in leaf nutrients at the branch level suggested a strong relationship between nutrient dynamics and phenological activity. As stated by hypothesis 1, the leaves of C. laurifolius (phenophase‐overlapper) accumulated nutrients from late‐summer until the following season’s bud‐burst, when there was no phenological activity (apart from the slow BVG). The maintenance and/or accumulation of nutrient contents in the whole branch can be triggered by three processes: (1) increasing nutrient concentration within leaf biomass (leaf level process); (2) increasing leaf mass per area (leaf level process); (3) maintaining and/or enhancing the number of leaves per branch (branch level process). It was observed that there were neither significant mass‐based N and P enhancements nor significant leaf birth from November until May, when the exponential gain of nutrients per branch unit was attained. Therefore, the accumulation of N and P in C. laurifolius branches was probably due to increasing Lma throughout the life‐span of the leaf (Fig. 4), and to the absence of noticeable leaf shedding out of the June–July period (Fig. 3).

In contrast, the leaves of B. fruticosum (phenophase‐sequencer) progressively depleted nutrients from the beginning of summer until late‐autumn. Following the reasoning for C. laurifolius dynamics, the progressive leaf shedding together with the earlier arrest of mass gain per unit leaf area are likely to be responsible for this pattern, at least for N and P (Figs 3 and 4). The drop in nutrient levels was concurrent with flower bud formation, flowering and fruit growth; therefore, the ability to mobilize internal reserves in sub‐optimal seasons could help phenophase‐sequencers develop resource‐demanding phenophases during periods like the Mediterranean summer (Castro‐Díez and Montserrat‐Martí, 1998).

In both species, changes in leaf demography are likely to have more of an effect on the final N and P pool amounts than changes in nutrient concentration. Some authors have pointed out that shoot growth implies leaf shedding costs within leaf‐exchangers (Addicott and Lyon, 1973; Osborne, 1973); i.e. the resource‐demanding activities of growing organs can be strong enough to induce senescence in source leaves (Gil‐Nam, 1997). This seems to be the case in our study: B. fruticosum (phenophase‐sequencer) carries out resource‐demanding activities during most of the year, which could imply sink‐driven leaf shedding costs through gradual abscission, while C. laurifolius (phenophase‐ overlapper) concentrates its resource‐demanding activities at the end of spring, avoiding sink‐driven leaf shedding costs during the rest of the year. However, alternative factors such as drought‐induced leaf shedding (del Arco et al., 1991; Borchert et al., 2002) can also contribute to explain leaf shedding patterns in ecosystems existing under seasonal drought.

The patterns in K branch‐level contents seem to be under a more complex control. Leaf mass gain, leaf nutrient concentration variations and leaf demography dynamics have some influence on the final KBR pattern. As pointed out for K leaf‐level dynamics, a need for osmotic adjustment or water‐lifting during periods of high water demands could be responsible for the sharply different K patterns. Strong water demands could be due to (a) reproductive development during summer drought (B. fruticosum) or (b) phenophase overlapping during late spring (C. laurifolius). However, more specific research on K dynamics is needed to understand them properly.

The seasonal variability in the availability of soil nutrients is driven by climatic factors (nutrient release) and root uptake (Silla and Escudero, 2003). The main pulses of nutrient release in Mediterranean soils are detected in early spring and early autumn (Serrasoles et al., 1999). Cistus laurifolius neither exhibited remarkable phenological activity in early spring, nor in early autumn. During these periods there was a net nutrient accumulation in the leaves (Fig. 3B), suggesting a storage use of the soil availability pulses. In contrast, there was phenological activity in B. fruticosum in both periods, thus probably making use of the soil pulses for current growth.

Effects of the phenological strategy on the resorption of nutrients from senescing leaves

The N and P resorption efficiencies were lower for C. laurifolius. Hypothesis 2 states that the N and P contents of phenophase‐overlappers will be higher before senescence to make the resorption process more efficient. In fact, C. laurifolius started senescence with a higher nutrient content on an area basis (see Fig. 3A), but was not able to withdraw nutrients more efficiently than B. fruticosum, shedding a much more nutrient‐rich litter. Therefore, the results presented here do not support hypothesis 2. Controls on the nutrient resorption process during leaf senescence are complex; several authors have found that different environmental and internal plant factors can explain part of the inter‐ and intraspecific variability of resorption values (see Aerts and Chapin, 2000). For instance, nutrient status (Nmax in this article) has been shown to enhance reff (Lajtha, 1987; Nambiar and Fife, 1987), to diminish reff (Boerner, 1984), or even to have no effect upon reff (Staaf, 1982; Chapin and Shaver, 1989; Schlesinger et al., 1989). However, Chapin and Moilanen (1991) point out that the pool of nutrients reabsorbed and shed is always high when the nutrient status is high. Although C. laurifolius did not perform N and P resorption more efficiently than B. fruticosum, N and P contents in the litter of the later species were lower, and N and P pools withdrawn per gram of senescent leaf were lower as well (see Table 1), in accordance with Chapin and Moilanen (1991). The K‐reff was different between species because of the progressive K withdrawal from B. fruticosum leaves. Nutrient retranslocation before senescence decreases nutrient resorption efficiency ratios (Pasche et al., 2002). This is a bias that could account for the absence of a clear predictor of resorption efficiency. It is noticeable that nutrient resorption efficiencies computed for the whole life‐span of the leaves, instead of just during spring senescence, would produce significantly higher efficiencies in the case of K resorption for B. fruticosum, but lower ones in cases such as N and P for C. laurifolius.

The branch‐level nutrient recovery in C. laurifolius was higher since there were more senescent leaves in spring, as opposed to a more efficient resorption process. Its internal nutrient availability could help to support the high resource demands of simultaneous vegetative and reproductive growth (i.e. to be a phenophase‐overlapper). This was also the case in Bistorta bistortoroides, an alpine herb with a short growing season (Jaeger and Monson, 1992). Facing different resource‐demanding activities at the same time has an additional cost due to intra‐plant competition among growing organs differing in sink strength (Ho, 1992; Mediavilla and Escudero, 2003). It has been suggested that sink strength and sink type (vegetative or reproductive) could regulate nutrient withdrawal from senescing leaves (Chapin and Moilanen, 1991; Jonasson, 1995), the carbon fluxes, i.e. phloem loading and unloading processes, being controlled by carbon demands of growing structures. If this hypothesis worked for C. laurifolius, not only nutrients, but also carbon demands could be partially satisfied by the large pool of leaves from the previous year that are depleted of resources during current‐year vegetative and reproductive growth. The possession of large amounts of nutrient reserves can alleviate internal competition between reproductive and vegetative growth, and thus allow overlap of phenophases (Jaeger and Monson, 1992). In contrast, B. fruticosum has to build abundant current‐year leaf biomass prior to reproductive development; those current‐year leaves are partly used to supply carbon and nutrient requirements of subsequent reproductive growth. In addition, C. laurifolius produced three times more leaf biomass per branch in spring than B. fruticosum (data not shown). Therefore, not only the degree of phenophase‐overlapping, but also the growth rate, can be influenced by the way plants use the nutrient stores accumulated in old leaves. This result supports the view of earlier studies (Monk, 1966; Cherbuy et al., 2001) that emphasize the role of old leaves as nutrient sources for further growth in evergreen species, and suggests that a high amount of resource reserves could be a prerequisite for concurrent activities.

In conclusion, it has been shown that nutrient pools in old leaves can strongly depend on plant growth activities throughout the year. Moreover, the results suggest that the nutritional costs of phenological activity are more readily detectable by assessing variations in nutrient pools in branches than in leaves. The nutrient patterns at the branch level were consistent with hypotheses 1 and 2. The phenophase‐overlapper showed non‐stop nutrient accumulation within the branch, sharply decreasing nutrient reserves during massive spring growth. The phenophase‐sequencer progressively shed leaves (and thus nutrients), throughout the year, probably as a consequence of sink‐induced leaf senescence.

ACKNOWLEDGEMENTS

We thank Carmen Pérez‐Rontomé for her help in data collection and analysis, as well as Pilar Castro‐Díez, María Herrero, Tadaki Hirose and two anonymous referees for their valuable comments on earlier versions of the manuscript. This study was supported by (1) the ‘Comisión Interministerial de Ciencia y Tecnología’ (Spanish government) project REN 2000‐0163‐P4‐05, REN2002‐02635/GLO and the Thematic Network GLOBIMED (REN 2001‐4841‐E/GLO), and (2) the ‘Gobierno de Aragón’ project P‐024/2001 and financial support to R.M. The English was edited by Morris Villarroel.

Received: 26 November 2003; Returned for revision: 28 January 2004; Accepted: 12 February 2004; Published electronically: 8 April 2004

References

  1. AddicottFT, Lyon JL.1973. Physiological ecology of abscission. In: Kozlowski TT, ed. Shedding of plant parts New York: Academic Press, 85–124. [Google Scholar]
  2. AertsR.1996. Nutrient resorption from senescing leaves of perennials: are there general patterns? Journal of Ecology 84: 597–608. [Google Scholar]
  3. AertsR, Chapin III FS.2000. The mineral nutrition of wild plants revisited: a re‐evaluation of processes and patterns. Advances in Ecological Research 30: 1–67. [Google Scholar]
  4. AertsR.2002. The role of various types of mycorrhizal fungi in nutrient cycling and plant competition. In: van der Heijden MGA, Sanders IR, eds. Mycorrhizal ecology Berlin: Springer‐Verlag, 117–134. [Google Scholar]
  5. AllenSE, Grimsban HM, Parkinson JA, Quarmby C, Roberts JD.1976. Chemical analysis. In: Chapman SB, ed. Methods in plant ecology Oxford: Blackwell, 411–466. [Google Scholar]
  6. BazzazFA, Grace J.1997.Plant resource allocation. San Diego: Academic Press. [Google Scholar]
  7. BloomAJ, Chapin III FS, Mooney HA.1985. Resource limitation in plants: an economic analogy. Annual Review of Ecology and Systematics 16: 363–392. [Google Scholar]
  8. BoernerREJ.1984. Foliar nutrient dynamics and nutrient use efficiency of four deciduous tree species in relation to site fertility. Journal of Applied Ecology 21: 1029–1040. [Google Scholar]
  9. BorchertR, Rivera G, Hagnauer W.2002. Modification of vegetative phenology in a tropical semi‐deciduous forest by abnormal drought and rain. Biotropica 34: 27–29. [Google Scholar]
  10. BoyerJS.1988. Cell enlargement and growth‐induced water potentials. Physiologia Plantarum 73: 311–316. [Google Scholar]
  11. CabezudoB, Pérez Latorre AV, Navarro T, Nieto Caldera JM.1993. Estudio fenomorfológico en la vegetación del sur de España. II. Alcornocales Mesomediterráneos (Montes de Málaga, Málaga). Acta Botanica Malacitana 18: 179–188. [Google Scholar]
  12. Castro‐DíezP, Montserrat‐Martí G.1998. Phenological pattern of fifteen Mediterranean phanaerohytes from Quercus ilex communities of NE‐Spain. Plant Ecology 139: 103–112. [Google Scholar]
  13. ChapinFS, Moilanen L.1991. Nutritional controls over nitrogen and phosphorus resorption from Alaskan birch leaves. Ecology 72: 709–715. [Google Scholar]
  14. ChapinFS, Shaver GR.1989. Lack of latitudinal variations in graminoid storage reserves. Ecology 70: 269–272. [Google Scholar]
  15. ChapinFS, Schulze ED, Mooney HA.1990. The ecology and economics of storage in plants. Annual Review of Ecology and Systematics 21: 423–447. [Google Scholar]
  16. CherbuyB, Joffre R, Gillon D, Rambal S.2001. Internal remobilization of carbohydrates, lipids, nitrogen and phosphorus in the Mediterranean evergreen oak Quercus ilex Tree Physiology 21: 9–17. [DOI] [PubMed] [Google Scholar]
  17. CorreiaOA, Martins‐Louçao AC, Catarino F. M.1992. Comparative phenology and seasonal foliar nitrogen variations in mediterranean species in Portugal. Ecologia Mediterranea 18: 7–18. [Google Scholar]
  18. delArcoJM, Escudero A, Vega‐Garrido M.1991. Effects of site characteristics on nitrogen retranslocation from senescing leaves. Ecology 72: 701–708. [Google Scholar]
  19. DeLillisM, Fontanella A.1992. Comparative phenology and growth in different species of the Mediterranean maquis of central Italy. Vegetatio 99/100: 83–96. [Google Scholar]
  20. Fernández‐EscobarR, Moreno R, García‐Creus M.1999. Seasonal changes of mineral nutrients in olive leaves during the alternate‐bearing cycle. Scientia Horticulturaee 82: 25–45. [Google Scholar]
  21. Gil‐NamH.1997. Molecular genetic analysis of leaf senescence. Current Opinion in Biotechnology 8: 200–207. [DOI] [PubMed] [Google Scholar]
  22. GivnishTJ.2002. Adaptive significance of evergreen vs. deciduous leaves: solving the triple paradox. Silva Fennica 36: 703–743. [Google Scholar]
  23. GucciR, Lombardini L, Tattini M.1997. Analysis of leaf water relations in leaves of two olive (Olea europaea) cultivars differing in tolerance to salinity. Tree Physiology 17: 13–21. [DOI] [PubMed] [Google Scholar]
  24. HarperJL.1989. The value of a leaf. Oecologia 80: 53–58. [DOI] [PubMed] [Google Scholar]
  25. HoLC.1992. Fruit growth and sink strength. In: Marshall C, Grace J, eds. Fruit and seed production. Aspects of development, environmental physiology and ecology New York: Cambridge University Press, 101–124. [Google Scholar]
  26. HsiaoTC, Läuchli A.1986. Role of potassium in plant‐water relations. In: Tinker B, Läuchli A, eds. Advances in plant nutrition New York: Praeger Scientific, 281–312. [Google Scholar]
  27. HsiaoTC, Xu LK.2000. Sensitivity of growth of roots versus leaves to water stress: biophysical analysis and relation to water transport. Journal of Experimental Botany 51: 1595–1616. [DOI] [PubMed] [Google Scholar]
  28. JaegerCH, Monson RK.1992. Adaptive significance of nitrogen storage in Bistorta bistortoides, an alpine herb. Oecologia 92: 578–585. [DOI] [PubMed] [Google Scholar]
  29. JonassonS.1995. Resource allocation in relation to leaf retention time of the wintergreen Rhododendron lapponicum Ecology 76: 475–485 [Google Scholar]
  30. JonassonS, Chapin III FS.1985. Significance of sequential leaf development for nutrient balance of the cotton sedge, Eriophorum vaginatum L. Oecologia 67: 511–518. [DOI] [PubMed] [Google Scholar]
  31. KikuzawaK.1995. Leaf phenology as an optimal strategy for carbon gain in plants. Canadian Journal of Botany 73: 158–163. [Google Scholar]
  32. LajthaK 1987. Nutrient resorption efficiency and the response to phosphorous fertilization in the desert shrub Larrea tridentata (DC.) Cov. Biogeochemistry 4: 265–276. [Google Scholar]
  33. LoGulloMA, Salleo S.1988. Different strategies of drought resistance in three Mediterranean sclerophyllous trees growing in the same environmental conditions. New Phytologist 108: 267–276. [DOI] [PubMed] [Google Scholar]
  34. López‐GonzálezG.2001.Los árboles y arbustos de la Península Ibérica e Islas Baleares. Madrid: Mundi‐Prensa. [Google Scholar]
  35. LovelessAR.1961. A nutritional interpretation of sclerophylly based on differences in chemical composition of sclerophyllous and mesophytic leaves. Annals of Botany 25: 168–184. [Google Scholar]
  36. LovelessAR.1962. Further evidence to support a nutritional interpretation of sclerophylly. Annals of Botany 26: 551–561. [Google Scholar]
  37. MediavillaS, Escudero A.2003. Relative growth rate of leaf biomasa and leaf nitrogen content in several mediterranean woody species. Plant Ecology 168: 321–332. [Google Scholar]
  38. MillardP.1994. Measurement of the remobilization of nitrogen for spring leaf growth of trees under field conditions. Tree Physiology 14: 1049–1054. [DOI] [PubMed] [Google Scholar]
  39. MillardP.1996. Ecophysiology of the internal cycling of nitrogen for tree growth. Journal of Plant Nutrition and Soil Sciences 159: 1–10. [Google Scholar]
  40. MitrakosKA.1980. A theory for Mediterranean plant life. Acta Oecologica 1: 245–252. [Google Scholar]
  41. MonkCD.1966. An ecological significance of evergreenness. Ecology 47: 504–505. [Google Scholar]
  42. Montserrat‐MartíG, Pérez‐Rontomé C.2002. Fruit growth dynamics and their effects on the phenological pattern of native Pistacia populations in NE Spain. Flora 197: 161–174. [Google Scholar]
  43. MooneyHA.1983. Carbon‐gaining capacity and allocation patterns of Mediterranean climate plants. In: Kruger FJ, Mitchell DT, Jarvis JUM, eds. Mediterranean‐type ecosystems. The role of nutrients. Berlin: Springer‐Verlag, 103–119. [Google Scholar]
  44. MooneyHA, Dunn EL.1970. Convergent evolution of Mediterranean‐climate evergreen sclerophyll shrubs. Evolution 24: 292–303. [DOI] [PubMed] [Google Scholar]
  45. MooneyHA, Kummerow J.1981. Phenological development of plants in Mediterranean‐climate regions. In: di Castri F, Goodall DW, Specht RL, eds. Mediterranean‐type shrublands Amsterdam: Elsevier, 303–307. [Google Scholar]
  46. MooneyHA, Rundel PW.1979. Nutrient relations of the evergreen shrub, Adenostoma fasciculatum, in the California chaparral. Botanical Gazette 140: 109–113 [Google Scholar]
  47. NambiarEKS, Fife DN.1987. Growth and nutrient retranslocation in needles of radiata pine in relation to nitrogen supply. Annals of Botany 60: 147–156. [Google Scholar]
  48. OliveiraG, Martins‐Louçao MA, Correia O, Catarino F.1996. Nutrient dynamics in crown tissues of cork‐oak (Quercus suber L.). Trees – Structure and Function 10: 247–254. [Google Scholar]
  49. OrgeasJ, Ourcival JM, Bonin G.2002. Seasonal and spatial patterns of foliar nutrients in cork oak (Quercus suber L.) growing on siliceous soils in Provence (France). Plant Ecology 164: 201–211 [Google Scholar]
  50. OrshanG.1989.Plant pheno‐morphological studies in Mediterranean type ecosystems. Dordrecht: Kluwer. [Google Scholar]
  51. OsborneDJ.1973. Internal factors regulating abscission. In: Kozlowski TT, ed. Shedding of plant parts New York: Academic Press, 125–147. [Google Scholar]
  52. PascheF, Pornon A, Lamaze T.2002. Do mature leaves provide a net source of nitrogen supporting shoot growth in Rhododendron L.? New Phytologist 154: 99–105. [Google Scholar]
  53. PérezLatorreAV, Cabezudo B.2002. Use of monocharacteristic growth forms and phenological phases to describe and differentiate plant communities in Mediterranean‐type ecosystems. Plant Ecology 161: 231–249. [Google Scholar]
  54. RappM, Santa Regina I, Rico M, Gallego HA.1999. Biomass, nutrient content, litterfall and nutrient return to the soil in Mediterranean oak forests. Forest Ecology and Management 119: 39–49. [Google Scholar]
  55. RathckeBJ, Lacey EP.1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics 16: 179–214. [Google Scholar]
  56. ReaderRJ.1978. Contribution of overwintering leaves to the growth of three broad‐leaved, evergreen shrubs belonging to the Ericaceae family. Canadian Journal of Botany 56: 1248–1261. [Google Scholar]
  57. RosecranceRC, Weinbaum SA, Brown PH.1998. Alternate bearing affects nitrogen, phosphorous, potasium and starch storage pools in mature pistachio trees. Annals of Botany 82: 463–470. [Google Scholar]
  58. SantaReginaI, Rico M, Rapp M, Gallego HA.1997. Seasonal variation in nutrient concentration in leaves and branches of Quercus pyrenaica Journal of Vegetation Science 8: 651–654. [Google Scholar]
  59. SchlesingerWH, de Lucia EH, Billings WD.1989. Nutrient use efficiency of woody plants on contrasting soils in the western Great Basin, Nevada. Ecology 70: 105–113. [Google Scholar]
  60. SerrasolesI, Diego V, Bonilla, D.1999. Soil nitrogen dynamics. In: Rodá F, Retana, J, Gracia CA, Bellot J, eds. Ecology of Mediterranean evergreen oak forests Berlin: Springer‐Verlag, 223–236. [Google Scholar]
  61. SillaF, Escudero A.2003. Uptake, demand and internal cycling of nitrogen in saplings of Mediterranean Quercus species. Oecologia 136: 28–36. [DOI] [PubMed] [Google Scholar]
  62. SmallE.1972. Photosynthetic rates in relation to nitrogen recycling as an adaptation to nutrient deficiency in peat bog plants. Canadian Journal of Botany 50: 2227–2233. [Google Scholar]
  63. SpechtRL.1969. A comparison of the sclerophyllous vegetation characteristic of the Mediterranean type climates in France, California, and Southern Australia. Australian Journal of Botany 17: 293–308. [Google Scholar]
  64. StaafH.1982. Plant nutrient changes in beech leaves during senescence as influenced by site characteristics. Acta Oecologica. Oecologia Plantarum 3: 161–170. [Google Scholar]
  65. van HeerwaardenLM, Toet S, Aerts R.2003. Current measures of nutrient resorption efficiency lead to a substantial underestimation of real resorption efficiency: facts and solutions. Oikos 101: 664–669. [Google Scholar]
  66. ZimmermannU, Schneider H, Wegner LH, Wagner HJ, Szimtenings M, Haase A, Bentrup FW.2002. What are the driving forces for water lifting in the xylem conduit? Physiologia Plantarum 114: 327–335. [DOI] [PubMed] [Google Scholar]

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

RESOURCES