Leaf life span, dynamics and construction cost of species from Mediterranean old-fields differing in successional status

Authors

  • Marie-Laure Navas,

    Corresponding author
    1. Centre d’Ecologie Fonctionnelle et Evolutive (C.N.R.S.), 1919, Route de Mende, 34293 Montpellier Cedex 5, France;
    2. E.N.S.A.-M., Département ‘Sciences pour la Protection des Plantes et Ecologie’ 2 Place Viala, 34060 Montpellier Cedex 1, France,
      Author for correspondence: M-L Navas Tel: (+) 33 4 99 61 24 57 Fax: (+) 33 4 67 54 59 77 Email: navas@ensam.inra.fr
    Search for more papers by this author
  • Béatrice Ducout,

    1. Centre d’Ecologie Fonctionnelle et Evolutive (C.N.R.S.), 1919, Route de Mende, 34293 Montpellier Cedex 5, France;
    Search for more papers by this author
  • Catherine Roumet,

    1. Centre d’Ecologie Fonctionnelle et Evolutive (C.N.R.S.), 1919, Route de Mende, 34293 Montpellier Cedex 5, France;
    Search for more papers by this author
  • Jean Richarte,

    1. E.N.S.A.-M., Département ‘Sciences pour la Protection des Plantes et Ecologie’ 2 Place Viala, 34060 Montpellier Cedex 1, France,
    Search for more papers by this author
  • Joël Garnier,

    1. Les Agros, Route de Seillans, 83 830 Bargemon, France
    Search for more papers by this author
  • Eric Garnier

    1. Centre d’Ecologie Fonctionnelle et Evolutive (C.N.R.S.), 1919, Route de Mende, 34293 Montpellier Cedex 5, France;
    Search for more papers by this author

Author for correspondence: M-L Navas Tel: (+) 33 4 99 61 24 57 Fax: (+) 33 4 67 54 59 77 Email: navas@ensam.inra.fr

Summary

  • • Variations in leaf life span (LLS), construction cost (CC) and dynamics patterns (periods of leaf production, tp, and loss, tL, time lag separating the end of leaf production and the beginning of leaf loss, t) were investigated in species differing in successional status and life forms. We tested how those traits varied along the succession and how these were interrelated. A new graphical framework is proposed to assess the influence of dynamics traits on LLS.
  • • The study was conducted on 42 species of contrasted life forms, typical of various stages of secondary succession, under the Mediterranean climate of southern France.
  • • LLS increased along the succession, tp was shorter and t longer in species from the later stage, without significant change in CC or tL. Herbaceous species, mostly of early successional status, had short-lived, low-CC leaves, produced and lost continuously. Woody species, of later successional status, had long-lived leaves, with slightly higher CC than herbs. LLS and CC or payback time were weakly correlated.
  • • Variations in LLS and leaf dynamics along the succession were related to changes in plant stature and growth potential of species, captured by leaf traits. Whether this is the consequence of a decrease in frequency of disturbance or of a change in the level of resources remains an open question.

Introduction

Leaf life span (LLS) is a functional trait involved in a fundamental trade-off in plant functioning between a rapid production of biomass and an efficient conservation of nutrients. A short LLS tends to be associated with fast rates of carbon fixation (Chabot & Hicks, 1982; Reich et al., 1992) and relative growth (Reich et al., 1992; Ryser & Urbas, 2000), whereas a long LLS is the single factor contributing most to a high residence time of nutrients within the plant (Aerts, 1995; Garnier & Aronson, 1998; Eckstein et al., 1999). LLS is thus expected to vary with the characteristics of natural habitats: in particular, fast-growing species with short LLS tend to occupy nutrient-rich sites, while slow-growing species with long LLS tend to be favoured in nutrient-poor habitats (Small, 1972; Berendse & Aerts, 1987; Reich et al., 1992; Aerts, 1995). There also seems to be a systematic trend in LLS along disturbance gradients (Ryser & Urbas, 2000). Several studies conducted in temperate or tropical successional gradients have shown that species that develop rapidly following a disturbance usually have shorter LLS than species occurring later in the successional sequence (Shukla & Ramakrishnan, 1984; Koike, 1988; Hegarty, 1990; Reich et al., 1995; Bazzaz, 1996). In most temperate environments, there is a gradual shift of growth and/or life forms during secondary succession, from annuals to herbaceous perennials and woody species (Escarréet al., 1983; Inouye et al., 1987; Prach et al., 1997). Here, we will test whether LLS increases with the successional status of species in the Mediterranean region of southern France as observed in other climatic regions, and how this relates to the differences in LLS among life forms. In the successional seres studied, the following sequence of life forms (sensuRaunkiaer, 1934) is usually identified (Escarréet al., 1983; Tatoni et al., 1994): therophytes dominate early after land abandonment, hemicryptophytes become the prominent growth form between 10 and 15 yr and 60 yr after abandonment, chamaephytes occur mostly in intermediate successional stages and phanerophytes dominate after 60 yr of abandonment. Based on a simple model involving the construction costs of perennating organs, Kikuzawa & Ackerly (1999) have predicted that LLS should increase from annuals to herbaceous perennials to woody species. This hypothesis will be tested here, using the whole range of life forms found in the Mediterranean systems studied.

The parameters underlying LLS variations can be approached from two different, complementary, perspectives. Ultimately, LLS can be considered to result from a balance between lifetime carbon gain and construction and maintenance costs (Chabot & Hicks, 1982; Mooney & Gulmon, 1982; Kikuzawa, 1991; Diemer & Körner, 1996), while proximately, LLS results from the balance between the production and loss rates of leaves (Jurik & Chabot, 1986; Harper, 1989). In the first approach, it has been postulated that leaves with a high construction cost (leaf CC, the amount of glucose required to synthesize one gram of leaf biomass, Penning de Vries et al., 1974; Williams et al., 1987) or a large payback time (ratio of construction costs to photosynthetic rate of a leaf) should have a long LLS, to repay for their high initial carbon investment (Mooney & Gulmon, 1982; Williams et al., 1989; Kikuzawa, 1991; Sobrado, 1991). Reviewing the available data on deciduous and evergreen woody species, Poorter & Villar (1997) found only very slight, nonsignificant differences in leaf CC between the two types of species, while in a recent study on 162 woody species, Villar & Merino (2001) showed that leaf CC was slightly lower (c. 6%) in deciduous than in semi-deciduous and evergreen species, which had identical leaf CC. More generally, leaf CC tends to be slightly lower in herbaceous than in woody species (Poorter & Villar, 1997; Baruch & Goldstein, 1999), but whether this is associated with differences in LLS remains an open question. On the other hand, a strong positive relationship between LLS and leaf payback time was found by Sobrado (1991) on tropical woody species, confirming previous results by Williams et al. (1989). Here, we test whether leaf CC varies among species differing in successional status and life forms, and whether changes in leaf CC or CC : Amax relate to the putative differences in LLS among those.

In the second approach, the focus is on the dynamics of leaf production, with an aim to understand how leaf emergence and loss influence the age structure and the LLS of leaves in the canopy (Kikuzawa, 1983; Jurik & Chabot, 1986; Kikuzawa, 1988; Harper, 1989; Kikuzawa, 1995). Patterns of leaf birth and death are variously taken into account in the calculation of LLS (cf. Southwood et al., 1986; Kikuzawa, 1988; Williams et al., 1989; Ackerly, 1999; Craine et al., 1999), but to our knowledge, their relative importance on the determination of LLS has not been explicitly assessed. Some empirical relationships have been shown, in the case of woody species differing in successional status (Kikuzawa, 1983; Lechowicz, 1984; Kikuzawa, 1988; Koike, 1988). Trees of early successional environments generally exhibit continuous production and fall of short-lived leaves, trees from mature forests are characterized by flushy production and loss of long-lived leaves, whereas shrubs from deciduous forest understorey produce long-lived leaves simultaneously in early spring or in both spring and fall. Such differences in patterns of leaf emergence and loss are considered to be adaptive to resource availability (Kikuzawa, 1988). Here, building on the approach proposed by Jow et al. (1980) and Williams et al. (1989), we present a general model, which shows that LLS actually depends on three parameters: the periods of leaf production and loss, the time lag between the end of the production period and the beginning of the period of leaf death. The relative importance of these parameters in the determination of LLS is assessed, and we test whether a trend comparable with that observed in woody species differing in successional habit is detected on the range of species spanned in the present study.

Finally, the continuous monitoring of leaf births and deaths required to test the LLS model proposed allowed us to test whether yearly averages of LLS (as monitored by, e.g. Sydes, 1984; Diemer et al., 1992) can be properly estimated from the LLS assessed for leaves produced during the peak production period (as done by, e.g. Craine et al., 1999; Ryser & Urbas, 2000). This makes it all the more important that recent broad-scale comparisons involving LLS are based on the compilation of data from various experiments where LLS has been assessed in somewhat different ways (Reich et al., 1992; Diemer, 1998; Eckstein et al., 1999).

This study was conducted in Mediterranean old-fields over a 4-yr period on 42 species differing in successional status. These species represented a broad range of life forms, from annuals (therophytes) to long-lived trees (evergreen phanerophytes). They also displayed a range of leaf phenology from species producing and loosing leaves more or less continuously to species where these are produced and lost as seasonal flushes.

A model of leaf life span

A graphical framework is proposed to understand how the periods of leaf production and loss influence the LLS of species. Although LLS have been previously calculated from leaf dynamics pattern, to our knowledge this is the first model proposed to compare species with very different leaf longevity and highly contrasted patterns of leaf production and fall. Indeed, distinct methods have been previously used to estimate leaf longevity of herbaceous (e.g. Craine et al., 1999) or woody species (e.g. Jow et al., 1980; Kikuzawa, 1983). Furthermore, our framework applies to yearly average values of LLS (i.e. the whole set of leaves produced during 1 yr), and thus integrates variations in LLS, which may occur during the growing season, in contrast with other models based on changes in size of leaves (e.g. Craine et al., 1999). For each species, the cumulated number of leaves produced or lost is plotted against time. As a first approximation, these data are fitted with linear regressions (see Material and Methods section) for periods of actual leaf production or loss (Fig. 1). For a set of leaves produced, LLS is equal to the area delimited by the two regression lines linking the cumulated number of leaves produced or lost to time, divided by the number of leaves (Fig. 1).

Figure 1.

Theoretical cases showing leaf dynamics and leaf life span (LLS) of species characterized by (a) separated periods of leaf production and loss, (b) similar and overlapping periods of leaf production and loss and (c) longer period of leaf loss than of leaf production. Dotted lines represent the end of the period of leaf production [BE], the beginning [CF] and end [DG] of period of leaf loss. Solid lines link the cumulated number of leaves produced or lost to time, [AE] and [CG], respectively. tp is the period of leaf production (solid arrow), tL is the period of leaf loss (dashed arrow) and t is the time lag between the end of leaf production and the beginning of leaf loss (dashed and dotted arrow). In (a), dashed lines represent LLS of individual leaves belonging to a same cohort but with different dates of loss: LLS decreases from La to Lb and Lc.

A general expression of LLS can be written as (see Appendix 1 for details):

LLS = [(1/2) * (tp + tL)] + t(Eqn 1)

where tp is the period of leaf production, tL is the period of leaf loss and t is the time lag between the end of leaf production and the beginning of leaf loss. t is positive when the periods of leaf production and loss do not overlap (Fig. 1a) and negative otherwise (Fig. 1b,c). These three parameters were determined for the species studied here, and their relative impact on LLS was assessed.

Materials and Methods

Study sites

The study was conducted between September 1997 and June 2001, at two sites in southern France. The complete successional sequence, from recent field abandonment to mature forest (see Species section), could be recognized at each site. The first study site, Cazarils (43°46′ N, 3°42′ E), is located 20 km northwest of Montpellier and 32 km from the Mediterranean coast. It is typical of the French Mediterranean shrublands on limestone rocks (i.e. garrigue). A complete description can be found in Aronson et al. (1998). The second site, Les Agros (43°25′ N, 6°35′ E), is located 220 km east of Montpellier and 27 km from the coast on the southern edge of a large limestone plateau. The vegetation is herbaceous, with numerous shrubs, surrounded by a mixed Pinus halepensis – Quercus ilex woodland. More details can be found in Garnier et al. (2001).

At both sites, the climate is Mediterranean subhumid with cool to cold winters (Daget, 1977), and is characterised by a marked summer drought, frequent frosts in winter and unpredictability of precipitation in timing and amount, with generally frequent heavy rainfall events in autumn. Mean annual temperature is slightly lower in Cazarils (12.8°C) than in Les Agros (13.6°C). Rainfall is c. 1100 mm at both sites.

Species

In Cazarils, 21 species were selected in September 1997 and 14 in September 1998, in les Agros, 10 species were selected in October 1998 (Table 1). This corresponded to a total of 42 different species because both Brachypodium phoenicoides and Teuchrium chamaedrys were studied at both sites in 1998 and Brachypodium phoenicoides in 1997 and 1998 at Cazarils. These are common species typical of widespread plant associations found on calcareous soils in Mediterranean southern France: the Brachypodietum phœnicoidis (xeric grassland), and a mix of Quercetum galloprovinciale and Querceto-Buxetum (mixed evergreen-deciduous oak forest) (Braun-Blanquet et al., 1952).

Table 1.  List of species (nomenclature follows Tutin et al., 1968–80, 1993) their successional status and leaf life span (LLS), construction costs (CC), period of leaf production (tp), r2 of linear relationship linking the cumulated number of leaves produced to time, period of leaf loss (tL) and r2 of linear relationship linking the cumulated number of loss leaves to time and the time lag (t)
SpeciesFamilySucc. statusLife formSampl. site/dateLLS (d) ± SELeaf CC (g glucose g−1) ± setp Period of leaf production (d)r2 for produced leavestL Period of leaf loss (d)r2 for lost leavest Time lag (d)
  1. Life forms according to Raunkiaer (1934) were taken from Braun-Blanquet et al. (1952), de Bolòs et al. (1993) and completed with personal observations. Therophytes (Th.), Hemicryptophytes (He.) and Geophytes (Ge.) were grouped into herbaceous species (H), chamæphytes (Ch.), deciduous phanerophytes (DPh.) and evergreen phanerophytes (EPh.) were grouped into woody species (W). Sampling site and date: C97: Cazarils 1997, C98: Cazarils 1998, B98: Bargemon 1998. –: not available.

Acer monspessulanumAceraceae5DPh.B98174 ± 4.41.52 ± 0.02 950.861810.94 −50
Aegilops geniculataGramineae1Th.C98 64 ± 2.51.28 ± 0.022450.912230.99−247
Avenula bromoidesGramineae3He.C97 78 ± 3.81.27 ± 0.033360.932720.94−278
Brachypodium distachyonGramineae1Th.C97 66 ± 3.01.40 ± 0.012370.932740.87−166
Brachypodium phoenicoidesGramineae3He.C97113 ± 4.01.44 ± 0.013650.963300.92−265
Brachypodium phoenicoidesGramineae3He.C98155 ± 9.71.40 ± 0.023650.985400.87−308
Brachypodium phoenicoidesGramineae3He.B98156 ± 8.61.36 ± 0.013650.963770.97−178
Bromus erectusGramineae3He.C97 85 ± 4.01.39 ± 0.013650.962720.96−237
Bromus lanceolatusGramineae1Th.C97 65 ± 3.11.40 ± 0.012160.962450.91−188
Bupleurum rigidumUmbelliferae5He.C98128 ± 8.41.47 ± 0.011950.733260.87 −24
Buxus sempervirensBuxaceae5Eph.C97488 ± 28.61.58 ± 0.011590.528320.90−148
Calamintha nepetaLabiatae2Ch.C97106 ± 3.21.58 ± 0.013650.973730.89−300
Carex halleranaCyperaceae4He.C98128 ± 5.21.41 ± 0.013420.943080.96−199
Catananche coeruleaCompositae3He.C98 90 ± 2.71.39 ± 0.023650.964180.94−338
Crataegus monogynaRosaceae4DPh.B98147 ± 7.71.53 ± 0.03 680.701530.44 −50
Dactylis glomerataGramineae2He.C97 59 ± 3.91.39 ± 0.023650.942650.97−343
Daucus carotaUmbelliferae2He.C97 83 ± 3.41.44 ± 0.013490.973580.98−321
Dorycnium hirsutumLeguminosae3Ch.C97100 ± 7.81.64 ± 0.022020.922670.96 −96
Dorycnium pentaphyllumLeguminosae3Ch.B98170 ± 7.11.58 ± 0.023250.802490.72−219
Echinops ritroCompositae3He.C98 96 ± 3.61.39 ± 0.023650.963840.97−365
Eryngium campestreUmbelliferae2Ge.C98 75 ± 4.51.33 ± 0.01 850.921260.78−109
Helianthemum nummulariumCistaceae4Ch.C97 80 ± 3.21.42 ± 0.033360.942720.92−278
Juniperus oxycedrusCupressaceae4EPh.C97552 ± 11.51.44 ± 0.013010.888500.93 −76
Lavandula latifoliaLabiatae2Ch.C97127 ± 4.81.66 ± 0.012450.802390.97 −49
Medicago minimaLeguminosae1Th.C97 70 ± 2.71.47 ± 0.012580.982740.92−230
Phillyrea latifoliaOleaceae5EPh.C98802 ± 27.91.63 ± 0.01 287910.87 448
Phleum pratenseGramineae2He.C97 62 ± 2.21.50 ± 0.022370.962310.96−209
Pistacia terebinthusAnacardiaceae5DPh.C97173 ± 3.21.61 ± 0.04 341080.96  78
Plantago lanceolaPlantaginaceae2He.B98136 ± 3.31.36 ± 0.073650.955330.93−271
Potentilla crantziiRosaceae3He.C98118 ± 4.11.42 ± 0.013420.974170.96−308
Prunus mahalebRosaceae4DPh.C97131 ± 5.91.52 ± 0.012100.892140.88−128
Prunus spinosaRosaceae4DPh.C98160 ± 5.91.59 ± 0.011560.522170.97 −50
Psoralea bituminosaLeguminosae3He.B98162 ± 5.91.64 ± 0.023650.985330.96−244
Quercus ilexFagaceae5EPh.C97692 ± 22.41.43 ± 0.03 420.868060.69 371
Quercus pubescensFagaceae5DPh.C97206 ± 9.01.49 ± 0.02 282140.70 −38
Rosa micranthaRosaceae4DPh.C98102 ± 7.11.53 ± 0.031750.672170.98−163
Rubia peregrinaRubiaceae4Ch.B98144 ± 7.01.28 ± 0.023650.927040.84−339
Rubus sp.Rosaceae4DPh.B98203 ± 9.01.49 ± 0.023650.976700.95−277
Sanguisorba minorRosaceae2He.C98 93 ± 3.01.39 ± 0.013420.943860.91−338
Teucrium chamaedrysLabiatae4Ch.C98158 ± 5.31.55 ± 0,013420.924730.95−169
Teucrium chamaedrysLabiatae4Ch.B98217 ± 6.71.55 ± 0.013650.945750.95−178
Thymus vulgarisLabiatae2Ch.C97179 ± 6.61.57 ± 0.013650.904820.90−141
Viburnum tinusCaprifoliaceae5EPh.B98369 ± 10.71.62 ± 0.033650.887070.94−244
Viola scotophyllaViolaceae5He.C98112 ± 5.41.35 ± 0.013420.924830.89−338
Xeranthemum inapertumCompositae1Th.C97 56 ± 1.71.56 ± 0.012870.932650.91−259

The selected species belong to 18 botanical families, and span the whole range of life forms described by Raunkiaer (1934): there were five herbaceous annuals (therophytes), 16 herbaceous perennials (one geophyte and 15 hemicryptophytes) and 21 woody perennials (eight chamaephytes, eight deciduous phanerophytes and five evergreen phanerophytes) (Table 1). These species were attributed a successional stage indicator value on a scale from one to five, according to the time during postcultural succession when they most commonly occur (Braun-Blanquet et al., 1952; Escarréet al., 1983, M. Debussche, pers. comm.) (Table 1). These five stages correspond to communities differing markedly in floristic (and faunistic) composition: stage one (0–10 yr after abandonment) is dominated by therophytes, while stage two (10–15 yr after abandonment) is dominated by hemicryptophytes and some chamaephytes from the family Labiatae. The third stage (beginning 15–25 yr after abandonment) is dominated by Brachypodium phoenicoides, a tussock perennial grass. In stage four (c. 60 yr after abandonment), small woody species become the prominent growth form represented. Finally, phanerophytes dominate the fifth stage (more than 100 yr after abandonment), while a number of perennial herbaceous species and shrubs remain in the understorey. The comparatively large number of herbaceous perennial species sampled corresponds to their actual persistence as a major component of the vegetation throughout the succession (cf. Escarréet al., 1983).

LLS and leaf dynamics traits

Measurements Measurements were conducted on 10 different healthy individuals, except in phanerophytes for which two branches were selected from each of five individuals. At the first census, a well-lit stem was selected and marked with a coloured ring or twine. The last mature leaf of that stem was marked with ink. During the following censuses, which took place every 2 (fall and spring) to 5 wk (winter and summer), the last mature leaf was marked, the number of leaves produced since the previous census was noted and the total number of living leaves remaining on the stem above the first marked leaf was recorded. A leaf was considered to be dead when it had turned completely brown or had fallen.

All leaves that appeared between two successive censuses were considered to belong to the same cohort. A maximum of 13 cohorts were identified for each species during the first year of observation. For a given species, the observations were conducted until all leaves forming those cohorts were dead: this occurred before the end of the first year of observation for annuals, between the first and the second years for most herbaceous perennials and chamaephytes, and after more than 3 yr for some phanerophytes (c. 40 censuses).

Calculations of LLS and leaf dynamics parameters To have a sufficient number of leaves per cohort, cohorts of leaves were pooled for all the individuals of a given species. All leaves appearing between censuses i (time ti) and i + 1 (time ti+1) were supposed to have been produced at (ti + ti+1)/2 (calculated time of leaf birth). Similarly, all leaves disappearing between censuses i and i + 1 were supposed to have been lost at (ti + ti+1)/2 (calculated time of leaf death).

The life span of each leaf per cohort was the difference between the calculated times of leaf birth and death. This procedure took into account the differences in behaviour of leaves within and among cohorts (cf. Fig. 1a).

Average LLS was calculated per species, using either the whole set of leaves produced during 1 yr (yearly average value of LLS) or the spring cohort with the largest number of leaves. The exact time when this latter cohort was produced differed among species because of differences in phenology. Standard errors of mean were calculated per species and not per individual plant for each species because of the low number of leaves produced per cohort by each single plant.

For each species, the period of leaf production, tp, was the difference between the dates of production of the last and first cohorts. The period of leaf loss, tL, was assessed for the whole set of leaves produced during the year following the beginning of the experiment, as the difference between the dates of loss of the last surviving and that of the first dying leaves. The time lag between the end of leaf production and the beginning of leaf loss, t, was the difference between the dates of production of the last cohort and of loss of the first dying leaf (Fig. 1). Because these dynamic traits were assessed from linear adjustments, no simple method exists to calculate their standard errors of means.

Testing the model of LLS To assess the validity of the LLS model proposed above, the relationships between the cumulated number of leaves produced or lost and time were fitted with linear regressions applied for whole periods of active leaf production or loss. For leaf production, these relationships were found to be highly significant (r2 > 0.80) for 36 species (Table 1). For leaf loss, the same level of significance was found for 39 species. Weak relationships (r2 ≈ 0.50–0.70) were found when leaves were produced (Buxus sempervirens, Prunus spinosa, Rosa micrantha … ) or lost (Crataegus monogyna, Dorycnium pentaphyllum) as seasonal flushes. The quality of those relationships was not improved when thermal time was used instead of calendar time (not shown). No regression was calculated for species forming or loosing one to two cohorts of leaves (Phillyrea latifolia, Pistacia terebinthus and Quercus pubescens). As a consequence, those species were not included in the test of the model defined by Eqn. 1. For each species but these particular ones, equations of regressions linking the cumulated number of leaves produced or lost to time were used to estimate the dates of production of the last cohort and of loss of the first dying leaf, respectively. An estimate of t was the difference between those two estimated dates.

Values calculated with Eqn. 1 were then compared with the yearly average values calculated with the actual dates of birth and death of leaves, as described in the previous section.

Construction costs and estimate of payback time

Leaf construction costs (leaf CC) of each species were evaluated on five samples of leaf blades harvested in May 98 or 99, corresponding to the period when LLS was determined (cf. Table 1). For herbaceous and small woody species, samples were taken from plants in full light (i.e. not under tree cover) while for tall woody species these were taken from the outer part of the canopy. Samples were oven-dried at 60°C for at least 2 d, weighed and ground individually. Construction costs, expressed in g glucose g−1, were calculated with the method proposed by Vertregt & Vries (1987) and modified by Poorter (1994):

CC = [−1.041 + 5.077 * C/(1000 − Min)] * [(1000 − Min)/1000] + [5.325 * Norg/1000](Eqn 2)

where C, Min and Norg are, respectively, the concentrations of carbon, minerals and organic nitrogen (mg g−1).

The concentrations of carbon (C) and total nitrogen (N) were determined with an elemental analyser (Carlo Erba Instruments, model EA 1108, Milan, Italy). We assumed that Norg = N, because nitrate accumulation has been reported to be negligible both in legume (Gebauer et al., 1988) and nonlegume species when leaf N was lower than 29 mg g−1 (Garnier & Freijsen, 1994), a condition always met here (cf. Garnier et al., 2001).

The mineral concentration of each sample was calculated as proposed by Poorter & Villar (1997):

Min = Ash − AA * 30 + Nw(Eqn 3)

where Ash was ash content (in mg g−1), AA * 30 the ash alkalinity (in meq g−1) multiplied by the mass of carbonate (g −1) and Nw the nitrate content (mg g−1), assumed to be negligible (see above). Ash content was determined after combustion in a muffle furnace at 550°C for 6 h. Thereafter, ash alkalinity was determined acidimetrically (Poorter & Villar, 1997).

Payback time was estimated by calculating the ratio of leaf CC and maximum photosynthetic rate (Amax). Maximum photosynthetic rate was estimated from Specific Leaf Area (SLA) and Leaf Nitrogen Concentration (LNC) values taken from Garnier et al. (2001), and using the multiple regression model proposed by Reich et al. (1997): Log10(Amax) = 0.82 Log10(SLA) + 0.88 Log10(LNC) – 0.76 (r2 = 0.80, n = 109, P < 0.001).

Treatment of data

As no significant effect of sampling site or date was found on LLS for species collected at both sites or dates (F < 0.21, P > 0.67), analyses were performed on the whole set of data. Variations in LLS, leaf CC, tp, tL and t between successional stages and among life forms were tested with one-way analyses of variance (anova), on transformed variables when appropriate. Post-hoc tests (Student Newmans Keuls) were performed when required. To test whether changes in leaf traits were likely to be the sole consequence of changes in life form along the succession, the same analysis was performed separately on hemicryptophytes and chamaephytes, the two life forms occurring in more than two successional stages (Table 1). Similarly, to test whether changes in leaf traits along the succession or among life forms hold for a given phylum, the same analysis was performed separately on Gramineae and Rosaceae, the two families occurring in more than two successional stages (Table 1). Relationships between traits were tested with linear regressions, using Type II when required. A multiple regression was conducted to compare the relative influence of tp, tL and t on LLS.

Statistical analyses were performed with the STATISTICA package (Version 5.1, StatSoft, Inc., Tulsa, OK, USA).

Results

Methods to assess LLS

The two average values of LLS, calculated either with leaves produced during spring or with all the leaves produced during 1 yr were significantly related (Fig. 2). LLS of species with short or medium longevity (< 200 d) was underestimated when assessed from the spring cohort. This was particularly true for therophytes whose LLS, calculated with the whole set of leaves, was lower than 80 d. In what follows, yearly averages of LLS will be used.

Figure 2.

Relationship between life spans calculated for all leaves produced over 1 yr and for the cohort with the highest number of leaves produced in spring. Solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. The solid line is the regression line: log(y) = 0.49 + 0.786 * log(x), r2 = 0.89, ***, P < 0.001. The dotted line is y=x.

Trait variation and successional status

LLS varied from 55 (Xeranthemum inapertum) to 802 d (Phillyrea latifolia), leaf CC varied from 1.27 (Avenula bromoides) to 1.66 g glucose g−1 (Lavandula latifolia), period of leaf production varied from 28 (Phillyrea latifolia) to 365 d (several herbaceous species), period of leaf loss varied from 108 (Pistacia terebinthus) to 850 d (Juniperus oxycedrus), the time lag between the end of leaf production and the beginning of leaf loss varied from −365 (Echinops ritro) to 448 d (Phillyrea latifolia) (Figs 3 and 4, Table 1).

Figure 3.

(a) Leaf life span (LLS) and (b) construction cost (CC) of species differing in successional status. Solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. Open symbols denote average values per stage. Results of one-way anova testing for the effect of successional status are shown with their level of significance (ns, non-significant; ***, P < 0.001). Different letters correspond to significant differences.

Figure 4.

(a) Period of leaf production, tp (b) period of leaf loss, tL, and (c) time lag between the end of leaf production and the beginning of leaf loss, t, for species differing in successional status. Solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. Open symbols are average values per stage. Results of one-way anova testing for the effect of successional stage are shown with their level of significance (ns, non-significant; **, P < 0.01, ***, P < 0.001). Different letters correspond to significant differences.

LLS varied substantially among species differing in successional status (Fig. 3a). LLS were significantly shorter in species from the youngest stage, longer in those from the oldest stage, with intermediate values for species characteristics of fields of medium age. Leaf CC was slightly, but not significantly higher in species from the two oldest stages, where most phanerophytes occurred (Fig. 3b). Despite large variability among species, the period of leaf production was the shortest in the oldest stage, where deciduous trees producing leaves as a single flush were found (Fig. 4a). The period of leaf loss increased slightly but non-significantly along the succession (Fig. 4b: note the very contrasted periods of leaf loss of deciduous and evergreen woody species of the two oldest stages). The time lag between the end of leaf production and the beginning of leaf loss was significantly longer in species from the oldest stage consisting mostly of trees with positive t-values; most species from earlier stages showed overlapping periods of leaf production and loss (Fig. 4c).

No significant change in any trait was found along the succession for the two life forms occurring in more than two successional stages (P > 0.25: F < 2.49 for the hemicryptophytes and F < 3.58 for the chamæphytes).

LLS and leaf CC differed significantly among life forms (Table 2). On average, woody species had longer LLS than herbaceous species, therophytes had the shortest LLS, whereas evergreen phanerophytes had the longest LLS. Hemicryptophytes, chamæphytes and deciduous phanerophytes showed intermediate values of LLS. Leaf CC of herbaceous species were slightly lower than that of woody species, with the largest range of variation displayed by chamæphytes (Fig. 3b). The period of leaf production was significantly shorter for trees than for other life forms, with shortest values for deciduous trees. The period of leaf loss was the shortest for deciduous trees and therophytes and the longest for evergreen trees, with intermediate values for the two other life forms. The time lag between the end of leaf production and the beginning of leaf loss was significantly the longest for evergreen trees because of distinct periods of leaf production and loss for most of species. For grasses, significant decrease in LLS (F = 10.80, P < 0.001) and increase in tp (F = 11.65, P < 0.01) were recorded along the succession whereas an increase in CC was found for Rosaceae (F = 8.89, P < 0.03). These changes were partly due to differences in life forms among stages: the increase in tp recorded for grasses over the first three stages of succession was explained by the replacement of therophytes, occurring only in the first stage, by hemicryptophytes (F = 14.65, P < 0.001); differences in CC of Rosaceae were due to the replacement of hemicryptophytes, occurring in the second and third stages by phanerophytes found only in the fourth stage (F = 20.11, P < 0.01). Furthermore, t was significantly lower for herbaceous than woody Rosaceae (F = 7.14, P < 0.05).

Table 2.  Variations in leaf life span (LLS), construction cost (CC), periods of leaf production (tp) and loss (tL) and time lag between the end of leaf production and the beginning of leaf loss (t) among life forms, tested by one-way anova
 LLS (d)Leaf CC (g glucose g−1)tp (d)tL (d)t (d)
  1. Significant differences among life forms according to Student Newmans Keuls post-hoc test are given by different letters. Eryngium campestre, the unique geophyte included in this study has been excluded from analyses. ns, non-significant; ***, P < 0.001; **, P < 0.01; D, deciduous; E, evergreen.

F 22.3***5.3** 11.1*** 13.2***   6.3***
Therophytes 64.3 d1.42 c249 ab256 dc−218 b
Hemicryptophytes109 c1.41 bc339 a378 b−268 b
Chamaephytes142 b1.54 ab323 a404 bc−196 b
D. Phanerophytes162 b1.54 a141 c247 d −85 b
E. Phanerophytes580 a1.54 abc179 bc797 a  70 a

Relationships between LLS, leaf CC and dynamics

LLS was weakly, positively related to leaf CC, whether evergreens were included in the analysis or not (Fig. 5a). A strongly significant positive relationship was found between the ratio of CC and Amax and LLS (Fig. 5b). However, it was nonsignificant when evergreens were removed from the analysis, for which significantly larger payback time was found than for other life forms (F = 11.8, P < 0.001).

Figure 5.

Relationships between leaf life span (LLS) and (a) construction cost (CC) (b) the ratio of construction cost and Amax for species differing in successional status: solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. r2 of linear regressions are shown with level of significance (ns, non-significant; ***, P < 0.001; *, P < 0.05). Equation of significant regression line calculated with data for all species (solid line) (a) log(y) = 0.62 + 1.03 * log(x), r2 = 0.15 (b) log(y) = 1.93 + 15.61 * log(x), r2 = 0.41. Underlined r2 is for the regression calculated without evergreen phanerophytes.

No relationship was found between LLS and the period of leaf production (Fig. 6a). By contrast, LLS was significantly related to the period of leaf loss and to the time lag between leaf production and loss, but the relationships remained only marginally significant when evergreens were excluded from analyses (Fig. 6b,c). A negative relationship was found between t and tp (r2 = 0.64, P < 0.001), the species with a long period of leaf production being characterized by overlapping periods of leaf production and loss. The multiple regression using tp, tL and t to fit LLS data was highly significant (r2 = 0.93), with similarly significant influence of tL and t and no effect of tp.

Figure 6.

Relationships between leaf life span (LLS) and (a) period of leaf production, tp (b) period of leaf loss, tL, and (c) time lag between the end of leaf production and the beginning of leaf loss, t, for species differing in successional status: solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. r2 of linear regressions are shown with level of significance (ns, non-significant; ***, P < 0.001; *, P < 0.05). Equation of significant regression line (solid line) (b) log(y) = 1.73 + 0.0011 * log(x), r2 = 0.55 (c) log(y) = 2.31 + 0.00106 * log(x), r2 = 0.41. Underlined r2 is for the regression calculated without evergreen phanerophytes.

There was a highly significant relationship between LLS values calculated with actual dates of birth and death and LLS estimated with Eqn. 1 (Fig. 7). However, LLS was slightly underestimated by the model (average 20%) and a deviation larger than 40% was recorded for 11 species because of inaccurate estimates of t (six species), tp (two species) and tL (one species). LLS of Dactylis glomerata and Echinops ritro were not accurately adjusted by our model. This may be related to the fact that these species are the only ones which begin to loose leaves as soon as some have been produced. The differences between recorded and estimated tp, tL and t for Brachypodium phoenicoides and Teuchrium chamaedrys were similar among sites and/or dates (data not shown).

Figure 7.

Relationship between simulated leaf life span (LLS), calculated with Eqn. 1, and observed LLS for species differing in successional status. Solid circle, therophyte; solid square, hemicryptophyte; solid inverted triangle, chamæphyte; solid triangle, deciduous phanerophyte; solid diamond, evergreen phanerophyte. The solid line is the regression line: y=–12.3 + 0.94x, r2 = 0.86. The dotted line is for y=x.

Discussion

Estimating LLS

The renewed emphasis on the importance of LLS for plant functioning (e.g. Reich et al., 1992; Aerts, 1995) has made it necessary to assess how differences in experimental procedures may affect the determination of this trait. In our study, monitoring only the leaves appeared during the peak production period underestimates LLS of species with short-lived leaves (< 200 d), but had little or no influence on the other species (Fig. 2). This was due to differences in the pattern of leaf production: those species with short yearly average LLS tend to produce leaves continuously (cf. Fig. 6a), and the LLS of their spring leaves was usually shorter than that of leaves produced in other seasons. The impact of these seasonal differences in LLS is important because spring leaves may represent only half of the total production of leaves in these species (data not shown, but see Kyparissis et al., 1997). By contrast, species with relatively long LLS tend to produce leaves in a spring flush (cf. Figure 6b), which is then highly representative of the average population of leaves. In that case, no seasonal differences in LLS are recorded. Caution should thus be taken when comparing the LLS of either species that produce leaves continuously or species that differ substantially in leaf emergence pattern. For doing such comparisons, we recommend that LLS were evaluated for a set of leaves representing more than three-quarters of the whole annual production.

Variation in LLS and leaf CC

LLS varied between 56 and 162 d in herbaceous species, and between 80 and 802 d in woody species (Fig. 3a, Table 1). Our sample did not include evergreen trees with very long-lived leaves (e.g. Pinus), which explains why the range for woody species is substantially narrower than that observed by Diemer (1998a) reviewing data on Mediterranean shrubs and trees. For herbaceous species, our data seem to be the first available to date in a Mediterranean environment. In hemicryptophytes, LLS was 109 d on average (Table 2), which is longer than for species from zones with a shorter growth period than Mediterranean France (56 d to 82 d: Sydes, 1984; Diemer et al., 1992; Craine et al., 1999; Ryser & Urbas, 2000) and shorter than for species from a-seasonal Andes (193 d: Diemer, 1998b) where growth occurs all year long. Overall, these results support the hypothesis of a positive association between LLS and length of growing season, previously formulated for regions where growth is seasonally constrained by temperature (Prock & Körner, 1996; Diemer, 1998b; Ryser & Urbas, 2000). Our data extend it to the Mediterranean climate, where growth is constrained by both temperature and water stress. This absence of differences between climates is probably the consequence of the small effect of drought on LLS: a similar range of LLS was found by Reich et al. (1999) for American sites differing in dryness, whereas 10% lower in LLS was documented for Australian species from arid places when compared with species from wetter places (Wright et al., 2002).

As recorded under other climatic conditions, LLS increased along the successional gradient (Fig. 3a). A number of potential environmental constraints on plants have been identified in successional habitats (Grime, 1979; Tilman, 1990). The following discussion will focus only on those likely to affect the traits measured. First, secondary succession may be considered to be characterized by a double gradient of decreasing disturbance frequency and intensity, and of increasing light competition resulting from the accumulation of standing biomass. To these gradients corresponds a gradient of life forms (Escarréet al., 1983; Inouye et al., 1987; Prach et al., 1997), from low to high stature plants. The increase in LLS along the succession actually corresponds to this gradient, with LLS increasing in the order: therophytes < hemicryptophytes < chamaephytes = deciduous phanerophytes < evergreen phanerophytes, the largest difference been found between the two classes of phanerophytes, even though evergreen trees with very long-lived leaves were not included in species set (Table 2, see also Kikuzawa & Ackerly, 1999). To account for this, Kikuzawa & Ackerly (1999) recently proposed a cost-benefit model in which the optimum LLS of a species depends not only on the construction costs of the leaf itself, but also on those structures (shoots in particular) supporting the leaf. A prediction of this model is thus that LLS should increase with plant height or mass. This was indeed found for the species studied here, whose height varied from 0.02 to 0.05 m (therophytes of the first stage) to more than 7 m (some phanerophytes of the latest stage) (E. Garnier et al., unpublished data): Log(LLS, d) = 0.28 * Log(plant height, m) + 5.27 (r2 = 0.50, P < 0.001). Changes in LLS along disturbance (Ryser & Urbas, 2000) or successional gradients (Shukla & Ramakrishnan, 1984; Hegarty, 1990; Reich et al., 1995; Bazzaz, 1996) have also been recorded within particular life forms, but to what extent this also relates to changes in plant stature or mass cannot be assessed with the data provided in these studies.

Second, secondary succession may involve changes in availability of soil resources either in absolute amounts or relative to light availability (discussed in, e.g. Tilman, 1990; Bazzaz, 1996). As LLS has a strong impact on nutrient conservation within the plant (Aerts, 1995; Garnier & Aronson, 1998; Eckstein et al., 1999), the increase in LLS with successional status of the species could be interpreted as a response to a decrease in nutrient availability as succession proceeds. Such a decrease could be due to the progressive development of an imbalance between nutrient supply by the soil and nutrient demand by the vegetation as a large standing biomass develops during the course of succession. Whether this actually holds in our particular situation remains to be established.

Tilman (1990) has further postulated that several of the underlying environmental constraints of successional habitats (including those discussed above) should lead to trade-offs in maximal growth rates of the species. High rates of resource capture and relative growth are indeed prominent plant traits characterizing the ‘ruderal’ syndrome associated with disturbed environments (Grime, 1979) and early stages of succession (Gleeson & Tilman, 1994). These high rates are assumed to be adaptive in such an environment, where rapid growth is both possible because of low competitive interactions with an established vegetation, and necessary to produce rapidly numerous seeds leading to a high colonization potential. These rates, together with a number of correlated traits, usually decline along successional gradients (Bazzaz, 1979; Gleeson & Tilman, 1994; Reich et al., 1995; Bazzaz, 1996) as the frequency of disturbance decreases and/or the absolute or relative levels of resources changes. This is the case for specific leaf area (SLA, the ratio of leaf area to leaf mass) in particular (Reich et al., 1995), a trait negatively related to LLS (Reich et al., 1997; Diemer, 1998b; Wright & Westoby, 2002). Such a decrease in SLA has been found for the species studied here (data from Garnier et al., 2001), and a negative association between SLA and LLS was observed as well (Log(SLA) =−1.19 * Log(LLS) + 3.51, r2 = 0.46, P < 0.001). The increase in LLS could thus be the direct consequence of the decrease in the growth potential of species along the succession.

Leaf CC were in the range found for many species (e.g. Poorter & Villar, 1997; Baruch & Goldstein, 1999; Villar & Merino, 2001), and were slightly lower in herbaceous than in woody species (−8%: Table 2), as found previously (Poorter & Villar, 1997; Baruch & Goldstein, 1999). This was confirmed comparing herbs and woody species within a family. Contrary to Villar & Merino (2001), we did not find higher leaf CC in evergreen than in deciduous species, maybe because evergreens in our sample had relatively short LLS. Differences among species were so small that changes in leaf CC along the succession were not significant (Fig. 3b). To our knowledge, the only other study where leaf CC of species differing in successional status have been assessed is that conducted on seven Piper species by Williams et al. (1989) in a tropical forest. These authors found higher leaf CC (and shorter LLS) in early successional species compared with late successional ones, but these differences were related to the 100-fold difference in the mean irradiance experienced by the seven species. These results are thus not comparable with ours, obtained on well-lit leaves for all species.

Leaf CC were found to be only weakly related lo LLS (Fig. 5a), which confirms the conclusions put forward by Sobrado (1991) and Villar & Merino (2001) for woody species. As already found by Williams et al. (1989) and Sobrado (1991) for tropical trees, leaf longevity was positively related to leaf payback time, calculated here as leaf CC : Amax (Fig. 5b). However, the relationship became hardly significant when evergreens were removed from the analysis. This weak relationship may indicate that the variation in Amax does not reflect the actual variation of daily carbon gain in the wide spectrum of species studied here as it does for tropical trees. Kikuzawa (1991) proposed that optimal LLS depends on the (b * leaf CC : Amax) ratio where b is the leaf age when photosynthesis becomes zero. This ratio could not be calculated here because b was not available for our species. Following this author, we may suggest that LLS of all species but evergreens is only controlled by b because leaf CC : Amax did not significantly differ among those species (Fig. 5b) and the decline in leaf photosynthesis with age has been shown to be faster in species with short LLS (Kitajima et al., 1997). In evergreens, both parameters could well be of importance.

Whatever the actual form and variables involved in the relationship between LLS and leaf CC, our data further confirm that the impact of differences in leaf CC among species on LLS is likely to be relatively low (Kikuzawa & Ackerly, 1999; Villar & Merino, 2001).

Influence of leaf dynamics on LLS

Mediterranean flora is characterized by a large variability in leaf phenology among life forms (Floret et al., 1989, M.-L. Navas, unpub. data), explaining why it is an appropriate model to test for relationships between leaf dynamics and leaf functioning. The only consistent pattern of variation in leaf dynamics parameters among species differing in successional status was found for the period of leaf loss (Fig. 4b): tL increased regularly, but not significantly, with successional status, in parallel with LLS (compare Figs 4b and 3a). The period of leaf production and the time lag between the end of leaf production and beginning of leaf loss show an abrupt and significant change for species of the latest stage (Fig. 4a,b). Species prominent in the first four stages display a large variability in phenology: herbs have a long period of leaf production, overlapping with a period of leaf loss, whereas most chamaephytes produce leaves in the autumn and/or spring and loose them next spring or summer. By contrast, most phanerophytes prominent in the oldest stage have a flushy spring production of leaves, separated from the period of leaf loss occurring the next autumn for deciduous species and more than 1 yr after production for evergreens. Therefore, two groups of species can be recognized along the successional gradient; these groups been also found within Rosaceae. Restricting the comparison to stages 4 and 5 where most woody species occur, this result is in line with studies showing that trees occurring early in successional environments exhibit continuous leafing and produce short-lived leaves, while trees forming mature forests tend to produce long-lived leaves in flushes (Kikuzawa, 1983; Lechowicz, 1984; Kikuzawa, 1988; Koike, 1988). These differences in leaf dynamics and structure have been related to changes in resource availability during succession (Kikuzawa, 1983, 1988), but whether this applies to our particular situation is unknown (see previous section).

LLS depended on the period of leaf loss and the time lag between the end of leaf production and the beginning of leaf loss (Fig. 5) The lack of relationship between LLS and leaf production period, already found by Kikuzawa (1988) was probably due to the negative link between t and tp. Using our framework, there were only two species, starting to lose leaves as soon as they were produced, for which the quality of fit between measured and estimated LLS was low (Fig. 7). In that case, a delay in the recording of the beginning or end of periods of leaf production or loss, due to the observation of cohorts instead of individual leaves, could have resulted in false estimations of tp and tL. These results demonstrate that our framework can be used for estimating LLS of a large range of species.

The quality of prediction given by our framework has to be compared with results from models previously proposed. LLS has previously been assumed to depend on leaf loss in species forming a single cohort (Southwood et al., 1986) or in species where the rate of leaf loss is similar to the rate of leaf production, resulting in a constant number of leaves per plant or branch (e.g. Jow et al., 1980; Ackerly, 1999; Duru & Ducrocq, 2000). In these cases, LLS can be properly estimated by dividing the number of leaves by the birth (or mortality) rate (Southwood et al., 1986; Ackerly, 1999). In terms of the model presented here, this means that LLS can be approximated by the sum of tp and t (see Appendix 2). To test whether this two-parameter model holds for the species studied here, we compared the LLS observed with that estimated by this sum. The relationship is significant when all species are included or for woody species, but not for herbaceous species or when evergreens are excluded from the analysis (Table 3). In another approach, LLS was assumed to depend both on birth and death rates of leaves, with a further hypothesis that these were also constant, but not necessary equal (Jow et al., 1980; Williams et al., 1989; Craine et al., 1999). Williams et al. (1989) calculated LLS as the difference between the number of leaves recorded at a date ti after the beginning of the experiment divided by the death rate, and the number of leaves that appeared until ti divided by the birth rate. Using our model, this means that LLS is equal to the difference between tL and tp (cf. Appendix 2). Table 3 shows that the LLS of our species are significantly predicted with this second model when all species are considered. The quality of the prediction decreases when evergreens are excluded, while accurate estimates of LLS are found for both herbaceous and woody species. The model proposed here (Eqn. 1) gave the best estimates of LLS for all sets of species (Table 3, Fig. 7). Our data also show that when species displaying a large range of phenological patterns are compared, the period of leaf loss and the time lag parameter to a lesser extent, have a stronger impact on LLS than the period of leaf production.

Table 3. r2 of relationships between observed leaf life span (LLS) and estimates provided by different models proposed by Southwood et al. (1986) (model 1), Williams et al. (1989) (model 2) and the present study (eqn. 1) (model 3). Derivations of LLS from models are (I) Model 1: LLS = tp + t; (2) Model 2: LLS = tL– tp; (3) Model 3: LLS = (1/2)  * (t p + tL) – t where tp, tL and t are the periods of leaf production and loss, and time lag between the end of leaf production and the beginning of leaf loss. See Appendices 1 and 2 for more details on calculation
SpeciesnModel 1 Model 2 Model 3 
  1. n= number of species. The total number of species is not equal to the sum of individual values for life forms because it includes a geophyte, Eryngium campestre. ns, nonsignificant, *: P < 0.05, **: P < 0.01, ***: P < 0.001.

All species450.40***0.70***0.86***
All species but evergreens400.06Ns0.49***0.77***
Herbaceous230.01Ns0.59***0.79***
Woody220.31**0.66***0.83***
Evergreen phanerophytes 50.13Ns0.00ns0.17ns

Conclusion

The model proposed here showed that differences in LLS depended largely on the period of leaf fall and the time lag between the end of leaf production and the beginning of leaf fall. Short LLS of species from early successional stages was related to long overlapping periods of leaf production and loss, and the long LLS of species from late successional stages was related to opposite traits. LLS was only weakly related to leaf construction costs and to payback time when evergreens were not taken into account. Differences in leaf life span and dynamics were related to changes in plant stature and growth potential of species from different successional stages.

Acknowledgements

Many thanks to James Aronson and Edouard Le Floc’h for the setting of favourable scientific conditions at the Cazarils site, to Sandrine Debain and Gérard Laurent for technical help, and to Rafael Villar and two anonymous reviewers for their comments on previous versions of the manuscript.

Appendix

Appendix 1. Derivation of leaf life span from leaf dynamics patterns

For a set of leaves produced, LLS is equal to the area delimited by the two regression lines linking the cumulated number of leaves produced or lost to time, respectively, divided by the number of leaves (cf. Fig. 1).

If the periods of leaf production and loss do not overlap (cf. Fig. 1a), then:

LLS = (A[ABE] + A[BCFE] + A[CGF]) * (1/N)((A1-1))

where A[ABE], A[BCFE] and A[CGF] are the area of polygons delimited by the points between brackets and N is the total number of leaf produced by that species during the period of census.

If the periods of leaf production and loss overlap (Fig. 1b,c), LLS is equal to:

LLS = (A[ABE] − A[CBH] + A[EHG]) * (1/N)((A1-2))

where A[ABE], A[CBH] and A[EHG] represent the area of triangles limited by the points into brackets.

Areas of polygons can be calculated with tp, tL, t and N as the lengths of [AB], [CD], [BC] and [BE], respectively. tp is the period of leaf production, tL is the period of leaf loss and t is the time lag between the end of leaf production and the beginning of leaf loss (see material and methods section). Therefore (A1-1) gives

LLS = (0.5 * tp * N + t * N + 0.5 * tL * N) * (1/N)((A1-3))

which can be simplified as

LLS = (1/2) * (tp + tL) + t((A1-4))

(A1-2) gives:

LLS = [0.5 * tp * N − (t2 * N)/(2 * tL) + (N/2tL) * (tL − t)2] * (1/N)((A1-5))

which can be simplified to

LLS = (1/2) * (tp + tL) − t((A1-6))

A single equation (A1-4) can be used for both kinds of species assigning positive value to t when periods of leaf production and loss are separated and negative value when they overlap.

Appendix 2. Derivation of leaf life span from leaf dynamics patterns

Model 1: Southwood et al. (1986)

In this model, LLS is calculated as the ratio between Ni, the total number of leaves present at time i, and the death rate of leaves. Ni is equal to the difference between the cumulated number of leaves produced and lost since the beginning of the censuses. Assuming that birth and death rates are constant over time, the cumulated number of leaves produced and lost can be written as linear functions of time. The relationship between Ni and time can thus be written as:

Ni = (b * ti) − d * (ti − (tp + t))((A3-1))

where b and d are the rates of birth and death of leaves, respectively, tp is the period of leaf production and t is the time lag between the end of leaf production and the beginning of leaf loss (see material and methods section). Therefore:

LLS = Ni/d = d(b/d) − (1) * ti + tp + t((A3-2))

Assuming similar leaf birth and death rates as in Southwood et al. (1986), LLS can be approximated by:

LLS = tp + t((A3-3))

Model 2: Williams et al. (1989)

In this model, LLS is calculated as:

LLS = (N0 + Ni−1,i)/d − Ni−1,i/b((A3-4))

where N0 is the number of leaves recorded at the beginning of the experiment, Ni−1,i is the number of leaves produced between ti−1 and ti. Assuming that birth and death rates are constant over time as done by Williams et al. (1989), it can be shown that:

Ni−1,i/d = tL((A3-5))
Ni−1,i/b = tp((A3-6))

If censuses begin when N0 equals zero as done in the present experiment, LLS can be estimated as

LLS = tL − tp((A3-7))

Ancillary

Advertisement