• Altitude;
  • Andes;
  • biogeography;
  • Ecuador;
  • foliar nitrogen;
  • latitude;
  • leaf longevity;
  • páramo;
  • SLA;
  • tropical


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

1. Leaf life spans, determined as half-lives, of 16 herbaceous perennial plant species from the aseasonal, tropical Andes at altitudes between 4000 and 4600 m were on average 193±19 days and thus two to three times greater than those of herbaceous perennials from the seasonal northern hemisphere. The duration of lamina expansion was positively correlated with life span. Growth was continuous all year round, but rates of leaf initiation in the high-elevation tropics were lower than in the temperate zone.

2. Significant positive inter-relationships were found between leaf life span and mass-based foliar nitrogen concentration of Andean taxa. The correlation between life span and leaf mass per area (LMA) was marginally significant, while life span and area-based foliar N content were unrelated.

3. Leaf traits of 46 Andean herbaceous perennials (forbs and graminoids), spanning an altitudinal range of 1500 m, indicated that the 16 taxa utilized for growth and leaf life span determinations were representative. For graminoids no altitudinal changes in LMA or foliar N were observed across species. However, in forbs LMA and area-based foliar N increased significantly with altitude. No differences in leaf traits were observed between herbaceous rosette and cushion growth forms.

4. The Andean data set was used in conjunction with data from the temperate, subarctic and arctic zones to evaluate the significance of leaf life spans in herbaceous plants from a biogeographical perspective. Leaf life span declined significantly with increasing seasonality, expressed as latitude and duration of the annual growth period. Because temperature regimes are similar along this gradient during the respective growth periods (Diemer 1996), observed differences should be the result of variation in the day length and/or duration of the growing season.

5. Data on leaf life spans and leaf traits of herbaceous plants from contrasting latitudes were used to evaluate the validity of general inter-relationships proposed by Reich (1993). The high correlation between LMA and life span observed in both data sets indicates that this relationship holds for various growth forms. On the other hand, the apportionment of mass-based foliar N concentration in relation to life span appears to differ between herbaceous and woody growth forms.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The relevance of leaf life spans first discussed by Small (1972), Chabot & Hicks (1982), Mooney & Gulmon (1982), Coley, Bryant & Chapin (1985) and Schulze, Fuchs & Fuchs (1977), particularly in the context of deciduous and evergreen leaf habits, has seen a resurgence recently (cf. Kikuzawa 1991; Diemer, Körner & Prock 1992; Karlsson 1992; Reich, Walters & Ellsworth 1992; Reich 1993). Chabot & Hicks (1982) postulated that leaf life span was dependent on carbon and nutrient status, with selection favouring long-lived leaves in climatically adverse or nutrient-poor habitats. However, until now this hypothesis was difficult to test because comprehensive data sets have become available only recently.

Diemer et al. (1992) examined temperate-zone herbaceous perennials (47°N) and found no differences in functional leaf life span between high-elevation, alpine species and closely related low-elevation taxa. Contrary to predictions by Chabot & Hicks (1982) neither leaf mass per unit area (LMA, inverse of SLA) nor photosynthetic capacity was correlated with leaf life span. A preliminary comparison between temperate zone and subarctic ecotypes of four species also did not show a latitudinal trend in leaf longevity. Prock & Körner (1996) found that functional leaf life span of seven subarctic alpine species (68°N) was similar to that of temperate-zone alpine species. However, functional leaf life span declined drastically in the northernmost, stongly seasonal arctic site at Spitzbergen (79°N).

Based on leaf longevity data of Amazonian tree species (Reich et al. 1991) and published literature, Reich et al. (1992) developed a comprehensive interpretation of leaf life span incorporating leaf habits, growth forms, stands and biomes. In accordance with the predictions of Chabot & Hicks (1982) short-lived leaves exhibited higher area- and mass-based net photosynthetic rates, lower LMA and foliar nitrogen concentrations than did long-lived leaves. Subsequently Reich (1993) re-interpreted the conflicting results of Diemer et al. (1992) on herbaceous perennials, and those of Koike (1988) on deciduous forest trees as subsamples exhibiting pronounced variability, which nevertheless were in accordance with the general trends proposed previously (Reich et al. 1992). He interpreted the divergent results of Diemer et al. (1992) and Reich et al. (1991) as examples of the classic fable of the blind men and the elephant, where examination of a particular life form or biome (‘elephant’s leg and trunk’, i.e. Diemer et al. 1992 and Reich et al. 1991, respectively) may lead to differing interpretations of the same general phenomenon (e.g. ‘the elephant’).

Yet the survey of leaf life spans utilized by Reich et al. (1992) was biased toward woody plants, some of which had life spans in excess of 5 years. Data on leaf life spans, LMA and foliar N of herbs from natural communities, however, existed merely from the temperate zone (Diemer et al. 1992). These have since then been expanded to include graminoids (Aerts & Caluwe 1995) and herbs from the subarctic and arctic (Prock & Körner 1996). Therefore the database up to now was simply too small to evaluate whether herbaceous plants conform to the general theory of leaf life spans (e.g. Chabot & Hicks 1982; Reich et al. 1992), or comprise a separate group. I intend to evaluate the validity of Reich’s (1993) hypothesis using a substantially greater data set, ergo the paraphrased title ‘news from the elephant’s leg’.

Furthermore, Reich et al. (1992) acknowledged the lack of data necessary for a biogeographical interpretation of the variation in leaf life spans, and they have recently expanded their data set to address this topic (Reich, Ellsworth & Walters 1998; Reich, Walters et al. 1998). Here I want to evaluate the effect of seasonality on leaf life span for herbaceous species, based on data obtained from aseasonal, high-elevation Andean plant communities. The specific objectives are: (1) the determination of leaf life spans and growth characteristics of tropical, high-elevation herbaceous perennial plant species, (2) an evaluation of how leaf traits, such as leaf mass per unit area (LMA) or foliar nitrogen concentration, which are related to leaf longevity, vary altitudinally within aseasonal, tropical páramo environments and (3) using these data to construct a latitudinal gradient of leaf life spans of cold, high-elevation environments spanning 0° to 79°N, making use of the data of Diemer et al. (1992), Prock (1994) and Prock & Körner (1996). Because growth period temperature regimes are quite uniform along this gradient (cf. Diemer 1996), it will be possible to evaluate the underlying evolutionary relationships between the life span of assimilatory organs, the phenology of vegetative growth and the duration of the growth period. Furthermore a test will be made of Reich’s (1993) contention that herbs conform to general theory of life spans.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Investigations were carried out primarily at three high elevation páramo sites in central Ecuador. The term páramo refers to high altitude grasslands as well as species-rich nival communities with discontinuous plant cover (sometimes termed ‘superpáramo’, cf. Cuatrecasas 1968) of South America. In fact, all of my study sites are situated above the upper altitudinal limit of continuous grassland vegetation. Most of the high elevation páramo taxa are reminiscent of alpine plants of the northern hemisphere; however, some unique growth forms are present as well, namely expansive cushion plants (e.g. Azorella spp., Plantago rigida, Werneria humilis), tall rosettes with pubescent leaves (Culcitium spp., Cerastium mollisimum) and parallel-splayed rosettes (Werneria nubigena, see also Rauh 1988).

The first site ‘Páramo de la Virgen’ (0° 18′S 78°15′W) is located just above the grassland belt at an altitude of 4060 m a.s.l. (≈ 45 km east of Quito). Soils are composed of an organic-rich layer 20–100 cm thick, interspersed with thin layers of volcanic ash. The vegetation zonation generally follows the relief with shrub communities restricted to leeward exposures or depressions, while herbaceous vegetation composed of cushion plants (P. rigida, W. humilis), bunchgrasses (Calamagrostis intermedia, Carex pichinchensis and others), Puya spp. and mosses predominate on exposed microsites. Plant cover generally exceeds 80% (visual estimate). Owing to the high frequency of precipitation and accumulation of organic matter the soils at ‘Páramo de la Virgen’ are poorly drained. [Note. Nomenclature follows Harling & Sparre (1973 to present) and the collection of the Herbario QCA, Pontifica Universidad Catolica, Quito.]

The second site ‘Guagua Pichincha’ (0° 10′N 78° 35′W) is located at 4510 m a.s.l. on the eastern slope of the volcano Guagua Pichincha roughly 10 km west of Quito. Soils are heterogeneous, consisting of exposed volcanic sands and gravels, or a 5–30 cm thick layer of dark organic-rich matter. The vegetation is composed primarily of herbaceous and rosette growth forms (P. rigida, W. humilis, Hypochoeris sonchoides, Culcitium nivale, Culcitium reflexus), with some graminoid and low shrub species present as well. Plant cover ranges from 80 to 100%.

The third study site ‘Cayambe’ (0° 01′N 78° 01 ′W) is situated at the base of the volcano Cayambe at 4570 m a.s.l. Soils are predominantly basaltic sands with a low organic content. The vegetation is composed of herbaceous perennials and grasses, but cover is merely 50%. Dominant species include C. nivale, Culcitium canescens, P. rigida, Werneria pumila and C. mollisimum. Although this site is only 60 m higher than ‘Guagua Pichincha’ climatic conditions are considerably harsher owing to the proximity of glaciers and frequent precipitation with intermittent snow cover.

Mean soil temperatures (– 5 cm depth) determined over the period November 1994–April 1995 decreased significantly with increasing altitude (daytime mean 8·1 ± 0·1 °C at Páramo del la Virgen to 5·6 ± 0·4 °C at Cayambe), whereas mean daytime quantum flux density (QFD) did not exhibit a distinct altitudinal trend (for detailed descriptions of microclimate and sites see Diemer 1996). Leaf temperatures, which ranged from – 4 to + 35 °C, were also independent of altitude, but dependent on growth form, e.g. maxima, minima and daily amplitudes were greatest in prostate rosette or cushion species (Diemer 1996). [Note. Although the microclimate shows drastic diurnal variations, air temperatures in Quito (2820 m a.s.l.) are quite constant all year round (13·0 ± 0·3 °C, Sarmiento 1986). Precipitation exhibits an annual bimodal distribution, but it increases with altitude and high-elevation soils tend to be moist all year round. Hence I will refer to the growth conditions in the páramos as aseasonal.]


At each of the three sites 10 non-flowering individuals of each of the most abundant herbaceous perennial taxa were chosen randomly and marked with a small wooden stake. Table 1 depicts the species and families examined. In total 17 species (16 herbaceous perennials and one graminoid) were selected.

Table 1.  . Leaf life span, expressed as leaf half-life, duration of lamina expansion and the rate of leaf inititation in high-elevation Andean herbaceous plant species. Site abbreviations: P.V., Páramo de la Virgen; G.P., Guagua Pichincha; C., Cayambe. Values are means and SE of 10 individuals per species, except where indicated by * (here extensive herbivory or disturbance reduced sample size to six to eight individuals) Thumbnail image of

A non-invasive tagging procedure was employed for monitoring initiation, growth and fates of fully sunlit leaves (cf. Diemer et al. 1992). Coloured plastic-coated electrical wire (diameter 0·5–0·2 mm) was placed around leaf petioles. Leaf growth was censused at 2 week intervals from late November 1994 until mid-May 1995 and thereafter once each at mid-June and mid-October 1995 (Páramo de la Virgen and Guagua Pichincha only). Leaves that were initiated within a particular time interval were grouped into a cohort. Based on the criteria outlined in Diemer et al. (1992) and Diemer & Körner (1996) I originally intended to distinguish various developmental stages within the life span of a leaf cohort and to determine the absolute life span (leaf birth until death) of marked leaves. However, owing to the great life span it was possible to determine merely the duration of leaf expansion (i.e. the period from leaf birth to 95% expansion of the leaf lamina) with a sufficiently high level of resolution for all species. The determination of leaf life span was therefore carried out indirectly via the calculation of leaf half-lives, an approach utilized in a number of previous studies (e.g. Sydes 1984; Mitchley 1988; Hodgson & Booth 1993). This procedure involves calculation of mortality rate (λ) and subsequently the half-life (t0·5),

inline image

inline image

where N0 represents the initial number of live leaves per plant and N1 the number of marked leaves alive at the end of the measurement interval (t, cf. Sheil, Burslem & Alder 1995).

In order to allow comparisons with published estimates of leaf life span of herbaceous taxa by Diemer et al. (1992) and Prock & Körner (1996), I determined the systematic difference between mean absolute leaf life span and estimates based on half-lives from a subsample of five temperate zone species, three of which were examined in two localities, published in Diemer & Körner (1996) and also in Diemer et al. (1992). Use of t0·5 underestimated absolute life span on average by 17 ± 9% and variation was greatest in species with short leaf life. The discrepancy declined to < 5% once life spans exceeded 80 days. Thus if leaf life spans of tropical, high elevation species are > 80 days, then methodological differences owing to the two approaches (direct assessment vs half-life) should be negligible.

In addition to the determination of leaf half-life and the duration of leaf expansion, I also determined the rate of leaf initiation. The time interval exceeded 4 months. Leaf initiation rates were expressed as the number of leaves initiated per shoot in a 30 day period (month).


Mature leaves of the páramo taxa were harvested from three to four individuals adjacent to the tagged plants on two or three occasions between December 1994 and May 1995. Following leaf area determination with a customized computer setup [laptop, hand scanner (Scanman EasyTouch, Logitech, Romanel, Switzerland) and image analysis software (AdOculos, DBS, Bremen, Germany)] samples were dried in a microwave oven. Subsequently dry mass (24 h at 80 °C), leaf mass per area (LMA, the inverse of SLA) as well as foliar nitrogen concentrations (CHN analyzer, Leco 900, Leco, St. Joseph, USA) were determined in Basel. For data analysis values obtained from differing dates were pooled.

In order to evaluate altitudinal trends of leaf traits of páramo plants, samples were obtained from an additional 29 herbaceous and graminoid species and processed as above. These data were also used to rank the studied taxa with respect to community variability and to provide a biogeographical comparison with published data on LMA and nitrogen concentrations of high-elevation plants (cf. Körner 1989; Körner et al. 1989). In addition to the primary study sites described above, samples were obtained from scree fields located at the base of Chimborazo (1°28′S 78°49′W) at altitudes between 4800 and 4900 m a.s.l. and on lower elevations en route to Guagua Pichincha, Cayambe and Páramo de la Virgen (3200–4400 m a.s.l.).


Diemer (1996) has shown that growth period temperature regimes of high-elevation tropical sites are quite similar to temperate-zone alpine habitats and can even be extended to the subarctic and arctic zones (see also Larcher 1975). Owing to the prostrate physiognomy canopy-level wind regimes at the various sites, exemplified by leaf temperatures, are also quite similar. Furthermore, growth period frequency distributions of quantum flux density are virtually identical among alpine temperate-zone and páramo sites (Diemer 1996). Hence it is possible to combine leaf life span data from these environments for a latitudinal comparison, which is not biased by temperature gradients. Published data on leaf life span, LMA and foliar N of 16 temperate-zone, alpine species from near Innsbruck, Austria (47°N, Diemer et al. 1992) were utilized. However, I used unpublished data on total, rather than functional, life span, which were utilized by Diemer et al. (1992) in their altitudinal comparison. (Functional life span was defined as the duration between leaf emergence and the advent of leaf senescence, while total life span includes the period of leaf senescence as well.)

Data on total life span and LMA of subarctic and arctic species were compiled from Prock (1994), (see also Prock & Körner, 1996) using the same methodology as Diemer et al. (1992). Subarctic data were obtained from seven species on a mountain near Abisko, Sweden (68°N), while the arctic sample comprising six species originated from Spitzbergen, Norway (79°N). All data sets above comprise estimates from summergreen leaves only, i.e. leaves which complete their lives within a given growth period. In all communities species with wintergreen leaves can be found. However, these leaves persist up to 10 months under snow (cf. Bell & Bliss 1977). Thus estimates of total life spans in wintergreen leaves contain a combination of short periods of net carbon gain (growth period), separated by an extended period of net carbon loss (dormant period), thus confounding altitudinal or latitudinal comparisons.


Simple linear regressions were used to test for inter-relations between leaf life span and leaf traits, altitude, latitude and measures of seasonality. Contrary to the data of Reich et al. (1992) no log10-transformations were necessary to linearize data. However, for comparisons of general trends proposed by Reich et al. (1992) and Reich (1993) I transformed my data accordingly. Differences within and among growth forms and latitudes with respect to leaf traits were analysed by t-tests. Latitudinal differences among leaf life spans were tested using one-way ANOVAs. Values were log-transformed to insure normalized distributions. Where necessary, Tukey’s MRT was used to distinguish among treatment levels (e.g. latitudinal origins).

Altitudinal and latitudinal comparisons of leaf life span and leaf traits are statistically confounded, because species rather than localities were used. Species were utilized owing to the scarcity of study sites and also because I wanted to illustrate the within-site variability among species. Furthermore, regressions based on site means did not affect regression equations, but increased R2 substantially (see also Fig. 2).


Figure 2. . Latitudinal variation in leaf life span of high-elevation herbaceous species comprising a gradient spanning the aseasonal tropics to the arctic. The slope of the indicated relationship is highly significant at the 0·1% level (see text). The duration of permanent snow cover where growth is not possible is indicated by the dashed line.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Leaf half-lives (t0·5) of páramo herbs were on average nearly 6·5 months (mean 193 ± 19 days, Table 1). Variation was considerable, spanning a range of more than 200 days (minimum: Geranium, Senecio 103 days, maximum: Plantago 327 days). In addition, leaf longevity tended to increase with altitude (Páramo de la Virgen: 169 ± 29 days, Guagua Pichincha 218 ± 31 days and Cayambe 197 ± 44 days), but this trend was not statistically significant (P = 0·53, one-way ANOVA).

Leaf expansion lasted 37% of the leaf life span (Table 1). A significant positive correlation was observed between the duration of leaf expansion and leaf half-life (P = 0·02, Table 2). On average, 1·6 leaves were initated within a 30 day period (Table 1). No significant correlation was detected between the initiation rate of leaves and the mean leaf life span (P = 0·17, Table 2) for the 16 species.

Table 2.  . Correlations between leaf life span and various leaf traits of Andean herbaceous plants. Data utilized for linear regressions are presented in Tables 1 and 3. P signifies the highest level of significance of either slope or intercept (two-tailed t-test) Thumbnail image of

Regressions utilizing leaf half-life as the dependent variable indicated that life span was weakly related to LMA (P = 0·06, Table 2). A significant negative relationship was obtained between life span and foliar N concentration (P < 0·01), whereas life span and foliar N content per unit area were unrelated (P = 0·69). Finally, a significant positive relationship existed between the duration of leaf expansion and LMA (P < 0·01, Table 2), which suggests that heavier, ‘thicker’ leaves expand their laminae at a slower rate than lighter, thinner leaves.


The herbaceous species utilized for leaf life span determinations did not differ significantly from other herbaceous taxa included in the general survey (Table 3) with respect to leaf traits. Among the groups mean LMA differed by merely 0·7% (P = 0·96, t-test). Foliar nitrogen concentration tended to be higher in the subsample utilized for the life-span data (+ 11%) and also N per unit area (+ 15%), but these differences were not statistically significant (P = 0·58 and 0·29, respectively, t-test). Because only one graminoid species was utilized for life span studies (Luzula racemosa), no statistical comparison is possible within the graminoid subsample. However, leaf traits of L. racemosa lie within the range of the graminoid sample (Table 3). Thus it can be concluded that the subsample of 17 páramo species utilized for leaf growth and life span measurements is representative for high-elevation herbaceous species in central Ecuador.

Table 3.  . Altitudinal comparison of leaf traits of high-elevation perennial plants from the tropical Andes. Species utilized for leaf growth and life-span determinations designated by bold print. Life-form abbreviations: G, graminoid; FR, rosette forb; FC, cushion forb. Site abbreviations: P.V., Páramo de la Virgen; G.P., Guagua Pichincha; C., Cayambe; Ch., Chimborazo Thumbnail image of

The entire data set, spanning an altitudinal range of 1500 m (e.g. 3400–4900 m a.s.l.), also lends itself for an evaluation of altitudinal trends of LMA and foliar nitrogen contents. Results of these analyses are presented in Table 4. No consistent altitudinal trends in leaf traits were observed for the seven graminoid species. In forbs, on the other hand, LMA increased significantly with altitude (P < 0·05, Fig. 1a), whereas foliar N concentration was not correlated with altitude (P = 0·50). Foliar N per unit area increased with altitude (P < 0·05, Fig. 1b). Regressions utilizing the most abundant taxonomic groups, namely the Asteraceae (n = 14) and Fabaceae (n = 5), did not fully substantiate community trends (Table 4). However, these results are biased, because in some cases a particular species was sampled at more than one altitude and site.

Table 4.  . Altitudinal dependence of leaf traits (LMA and foliar N) of high-elevation graminoids and herbs over an altitudinal range of 1500 m. In addition, altitudinal trends of the most prominent taxonomic groups are tabulated. All results were derived by linear regressions utilizing Table 3 as the data source Thumbnail image of

Figure 1. . Altitudinal variation in (a) leaf mass per area (LMA) and (b) area-based foliar nitrogen content in Ecuadorean forb species spanning an elevational range of 1500 m. The slope of the indicated relationship is significant at the 5% level, n = 44 species and/or sampling locations.

Download figure to PowerPoint

A comparison between the two most prevalent forb growth forms, namely cushion and rosette, yielded no significant difference for any of the variables (P = 0·78–0·96, t-test). It should be noted that leaves of the rosette (growth) forms peculiar to the Andes, namely tall rosettes with pubescent leaves (Culcitium spp., C. mollisimum) and parallel-splayed rosettes (Werneria nubigena) are characterized by high LMA (> 180 g m–2) and foliar N contents (> 200 mmol m–2), even though the results of the rosette:cushion comparison are not affected.


Leaf life span was significantly greater in the high-elevation tropical sites (mean 193 ± 19 days, P < 0·01, Tukey MRT and one-way ANOVA of log-transformed values) than in the northern hemisphere (Fig. 2). In addition, leaf life spans from the high arctic (48 ± 2 days) were significantly lower than in the temperate zone (82 ± 5 days, P < 0·01) and the subarctic (76 ± 2 days, P < 0·05, Tukey MRT). However, not only mean life span but also the coefficient of variation (CV) diminished from the tropics (39%) to the temperate zone (22%) and the strongly seasonal subarctic (7%) and arctic (12%) habitats.

Furthermore a significant relationship between life span and duration of the annual growth period, both expressed in terms of days (R2 = 0·63, P < 0·001, Fig. 3a) and cumulative photoperiod (R2 = 0·62, P < 0·001) exists. Mean duration of the annual growth period was 365 days in the tropics, 120 days in the temperate zone, 100 days in the subarctic and 70 days in the arctic (cf. Diemer 1996; Prock & Körner 1996). For the latter estimate a constant photoperiod (quantum flux density > 30 μmol m–2 s–1) of 12 h in the tropics, 14 h in the temperate zone, 20 h in the subarctic and 22 h in the arctic was assumed. This trend is also consistent within the family of the Asteraceae (Fig. 3b), which comprises roughly one third of the tropical and temperate-zone species. The remaining families or genera are not suitable for these taxonomically-based comparisons owing to small sample sizes.


Figure 3. . Variation of leaf life span of high-elevation herbaceous plants in response to the duration of the annual growth-period. Depicted are (a) all species comprising a latitudinal range of 79° and (b) members of the Asteraceae only. The slopes of the indicated relationships are highly significant at the 0·1% level (see text).

Download figure to PowerPoint

Unfortunately comparative data with respect to leaf traits were available only for the tropics and the temperate zone (Diemer et al. 1992). Comparisons utilizing t-tests indicated that LMA was significantly greater in the páramo (mean 106 ± 11 g m–2, utilizing the subset from Table 1) compared with the Alps (mean 59 ± 4, P < 0·01). The reverse was observed with respect to foliar N concentration; here the mean of temperate-zone high-elevation leaves was 39% higher (3·39 ± 0·22%) than in the tropics (2·44 ± 0·21%, P < 0·01). No difference was observed with respect to area-based foliar N content (P = 0·26). (Use of log-transformed values had no effect on results and thus original values are presented here.)


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Leaf life span of high-elevation plants from the Andes was two to three times greater than in other groups of herbaceous taxa (see above). At high-elevation sites in the Central Alps mean leaf life span of 16 species was 82 ± 5 days, as determined by direct leaf census (Diemer et al. 1992). Mean life span of low-elevation herbs from a British calcareous grassland in April, determined via t0·5, was 107 ± 14 days (Mitchley 1988) and thus 45% lower than the Ecuadorian sample. Sydes (1984) obtained mean half-lives of 65 days in a similar grassland for leaves, which expanded between April and July. Hence, observed differences between life spans of tropical high-elevation herbaceous perennials and those of northern hemisphere species are not a result of the differing methodologies utilized. They represent a real latitudinal phenomenon.

Leaf expansion lasted on average 37% of total leaf life span in páramo herbs (Table 1). In five temperate-zone low- and high-altitude species lamina expansion comprised 25–33% of the total life span (Diemer & Körner 1996). In high-elevation subarctic species (Prock 1994) a similar mean percentage (23%) was obtained. Because a close positive correlation exists between the duration of lamina expansion and the amortization of carbon costs invested by the plant in leaf construction in the temperate-zone (R2 = 0·88, Diemer & Körner 1996), one could speculate that both the rate of expansion of the lamina and the rate of carbon gain are lower in tropical high-elevation plants than in species of higher latitudes.

On average 1·6 leaves month–1 were initiated by the Andean herbs, which is 67% less than the annual mean (3·7 leaves month–1) of three low-elevation British grassland herbs (Sydes 1984). However, leaf initation in Britain was cyclical, with a virtual standstill from November to February, and maximum rates of up to 32 leaves month–1 in June and July (Sydes 1984). In three high-elevation alpine herbs in Austria mean leaf initiation rates during the 4 month growth period were on average 1·4 ± 0·3 leaves month–1 (Diemer 1990), a rate quite similar to the páramo species. However, the latter rate is in fact an underestimate, because in the temperate-zone alpine environment the majority of leaves were initiated during the first half of the growing season, whereas in the aseasonal tropics, leaf initiation appears to be constant and continuous all year round. Based on these trends one could hypothesize that higher rates of leaf initiation tend to accelerate leaf turnover, thus selecting for decreasing leaf life span in seasonal environments. Although there is some supportive experimental evidence from short-term responses to soil nutrient status (Shaver 1981; Schmid, Miao & Bazzaz 1990; Diemer et al. 1992), this relationship has not been evaluated in an evolutionary context for herbaceous plants. Harper (1988) suggests that despite an apparent synchrony of leaf deaths and births the two processes are not causally linked (but see Schmid & Bazzaz 1994). I found no significant relationship between leaf life span and rates of leaf initiation in the páramo species (P = 0·17, Table 2). Thus short life spans are not linked to increased turnover, supporting Harper’s (1988) contention. However, because growth conditions are constant in the aseasonal tropics and shading does not occur in open canopies above 4000 m, there is also no selective advantage for optimizing leaf carbon gain in the tropics, compared with temperate-zone closed canopy stands (cf. Werger & Hirose 1991; Egli 1994; Schmid & Bazzaz 1994).


No significant altitudinal trends in LMA and foliar N were observed in graminoids (Table 2), whereas in forbs LMA and area-based foliar N increased significantly with altitude (Fig. 1a,b). No such trend was observed for mass-based foliar N over the 1·5 km altitudinal range in the Andes. The observed altitudinal increase in LMA, or conversely decrease in SLA, has been shown in a number of inter- and intraspecific altitudinal comparisons of herbs (cf. Baruch 1979; Körner, Allison & Hilscher 1983; Woodward 1983; Körner et al. 1989). Furthermore, Körner (1989) has demonstrated a steep altitudinal increase in area-based foliar N of herbs for a number of mountain ranges, which are substantiated by the Ecuadoran páramo data (Fig. 1b). He goes on to suggest, that foliar N concentration of high-elevation herbaceous plants declines continuously with decreasing latitude (Körner 1989). Although this trend appears to be borne out by the differences observed in the temperate-zone vs tropical data (P < 0·01), the mean foliar N concentration of the Ecuadorean herbs used for life-span determinations (2·44 ± 0·21%), do not differ substantially from Körner’s (1989) collections from NW Argentina (2·58 ± 0·11%, 26°S) or Venezuela (2·19 ± 0·16%, 8°N). Thus it appears that the latitudinal differences in foliar N concentration are smaller than proposed, particularly because Körner (1989) explained the extremely low mean from tropical New Guinea (1·32%) with low levels of solar radiation unique to this mountain range.

A number of the inter-relationships between leaf life span and various leaf traits were tested both for the Andean (Table 2) and the comprehensive (latitudinal) data set (see below). In accordance with the temperate-zone data (Diemer et al. 1992), leaf life span of high-elevation Andean species was negatively correlated to foliar N expressed on a mass basis (P < 0·01) and unrelated to foliar N on an area basis (P = 0·69, Table 2). Contrary to the results of Diemer et al. (1992) LMA was marginally positively related to life span (P = 0·06, respectively P < 0·05 for SLA) in the Andean sample. Aerts & de Caluwe (1995) found that leaf longevity in four Carex spp. was unrelated to LMA, but negatively related to foliar N concentration (also non-significant). Nevertheless these results support the general predictions of Reich et al. (1992) and Chabot & Hicks (1982) which were based on carbon balance considerations: namely that maximum photosynthetic rate declines with increasing life span, along with decreases of mass-based foliar N and SLA. Furthermore, as life span increases, so do construction costs in leaf tissue.


In woody plants a bimodal distribution of long-lived (evergreen) leaf habits can be observed, with peaks in the tropics and at high latitudes (Chabot & Hicks 1982). For trees Kikuzawa (1991) has developed a model based on carbon balance considerations, which predicts a general latitudinal shift from evergreen habits to deciduous habits (fast leaf turnover and short longevity) as the duration of the unfavourable period increases. For herbaceous plants, neither models nor predictions are presently available.

A significant decline in leaf life span with increasing latitude or exent of seasonality was observed in thermally similar high-elevation environments (Table 5). Thus, as the duration of the annual growth period declines, leaves become more short-lived (Figs 3 and 4). Concurrently I also observed a latitudinal decline in the variability of leaf life spans (CV decreases from 38% to 12, respectively 7%). An interpretation of this trend is that external (climatic) constraints, which increase with latitude, have canalized the plasticity of leaf life spans in herbaceous perennial plants. Because growth period temperature regimes are similar over the latitudinal gradient examined in this study (see Diemer 1996), the primary determinants of latitudinal variations in life span must be duration of the growth period and/or photoperiod.

Table 5.  . Correlations between leaf life span and various leaf traits of high-elevation herbaceous plants spanning a latitudinal transect of 79° latitude. Data utilized for Andean data are presented in Tables 1 and 3. Temperate zone data were derived from Diemer et al. (1992) and arctic and subarctic data are by Prock (1994). Units for life span and LMA as defined in the text, unless indicated otherwise Thumbnail image of

Figure 4. . Inter-relationships between (a) log10 specific leaf area (SLA) and log10 leaf life span, and (b) log10 foliar nitrogen concentration and log10 leaf life span of herbaceous plants. The slope of the indicated relationship (dotted line) is highly significant at the 0·1% level (see Table 6). Included as a dashed line is the general relationship proposed by Reich, Ellsworth & Walters (1998) and Reich, Walters et al. (1998).

Download figure to PowerPoint


Overall plasticity

In a reciprocal transplant experiment of low-elevation herbs, leaf life spans of temperate-zone provenances (47°N) transplanted to the subarctic (68°N) increased, whereas leaf life spans of subarctic provenances tended to decline when these plants were grown in the temperate zone (Prock & Körner 1996). However, variation among species was considerable and both photoperiod and temperature varied between the two low-elevation transplant sites. In high-elevation taxa, which grow under similar climatic regimes, no consistent transplant effects were observed (Prock 1994), although the senescence of leaves of temperate-zone alpine provenances transplanted to the subarctic was delayed. However, overall leaf life spans among provenances at their respective latitudinal origins were quite similar, and within the range of year to year variations at a given site (Prock 1994; see also Fig. 2). Field observations and controlled environment experiments of high arctic plants led Bell & Bliss (1977) to conclude that leaf life span has low plasticity in these herbaceous species from extremely seasonal environments. Hence, there is no clear evidence of photoperiodic control of leaf life span in temperate-zone and arctic herbs. This does not, however, preclude the hypothesis that high-elevation tropical plants maintain greater leaf life spans because, contrary to higher latitudes, no selection exists for the termination of growth or leaf persistence within a restricted time period. Clearly experimental work is neccessary to test for the plasticity of leaf life spans.

Carbon gain

A second plausible determinant of leaf life span is lifetime carbon gain or surplus, which are linked to the duration of the growth period. Teleologically speaking a plant has two options, namely (a) to produce few long-lived leaves or (b) to produce several cohorts of short-lived leaves, within a given time period. Option (b) may be more advantageous in low-elevation habitats, where the duration of the growth period is longer and competition for light is severe in closed canopies (cf. Harper 1988; Egli 1994; Schmid & Bazzaz 1994) or herbivory is prevalent (Mooney & Gulmon 1982; Southwood et al. 1986; but see Grubb 1992). In temperate-zone high-elevation plants, which grow under conditions where above-ground competition (shading) is negligible, option (a) is favoured, i.e. fewer cohorts of long-lived leaves are produced (cf. Diemer et al. 1992). Obviously this trait is maintained even under the constant growth conditions of the high-elevation aseasonal tropics.

There is a general consensus (Chabot & Hicks 1982; Harper 1989; Kikuzawa 1991; Reich et al. 1992) that life span should be related to a balance between leaf carbon costs and benefits. Obviously leaves must live long enough to assimilate sufficient carbon to amortize their construction costs and to maintain the heterotrophic plant tissues. Thus it seems plausible that leaf life span could have evolved to optimize a carbon return ratio per unit leaf area. Unfortunately the database to test this hypothesis is very small; leaf carbon balance data are available for a temperate-zone sample comprising four species from various localities and a range of leaf life spans from 58 to 121 days (Diemer & Körner 1996). Leaf life span and lifetime leaf carbon balance per unit area were positively correlated (R2 = 0·74, P < 0·01, Diemer & Körner 1996). Yet carbon return ratios (lifetime carbon gain:costs) ranged from 3:1 to nearly 7:1 and were independent of life span (R2 = 0·02, P = 0·38, Diemer & Körner 1996).

Nevertheless some general hypotheses with respect to carbon gain can be formulated for the extremely long-lived leaves of high-elevation tropical herbaceous plants, which need to be substantiated by gas-exchange data. Published photosynthetic rates of non-woody páramo plants of Venezuela (Baruch 1979; Rada et al. 1992) are appreciably lower than in temperate-zone alpine plants (cf. Körner & Diemer 1987). The low foliar N contents of equatorial páramo species (Table 3), relative to temperate-zone high-elevation taxa (cf. Körner 1989), can be used as indirect evidence of low photosynthetic rates. Furthermore, lifetime dark respiration rates should be higher in tropical herbs, owing to constant 12 h photoperiods, than at higher latitudes, where growth-period photoperiods commonly exceed 14 h (increasing to 24 h in the arctic, i.e. no dark respiration). Hence, low rates of carbon uptake and high carbon costs (dark respiration and high LMA) in tropical Andean herbs, may have led to increased life spans in order to exceed minimal carbon return ratios, relative to species from higher latitudes.

Growth constraints

Finally, aside from carbon gain characteristics and photoperiod, a third selective force has been held responsible for the evolution of longevity under conditions of climatic stress or low nutrient availability, namely morphogenetic constraints (Grime 1979; Chabot & Hicks 1982). There is indeed ample evidence that inherently slow development (cell expansion and mitosis) and growth, high construction costs and increased longevity are characteristic of plants of cold climates (reviews by Billings 1974; Körner & Larcher 1988). Yet, there is no a priori reason, why high-elevation tropical herbs should be less constrained morphologically, than high-elevation plants from higher latitudes. With respect to soil nutrients Marrs et al. (1988) observed that N availability declined with altitude in Costa Rica, which they attributed to high soil moisture contents and thus reduced N mineralization. Unfortunately soil-nutrient data are missing, yet soil-texture and soil-moisture regimes at the highest sites in Ecuador are quite similar to the subarctic and temperate-zone sites used in the latitudinal comparison. However, if lower foliar N concentrations of tropical plants relative to higher latitudes are the result of low soil nitrogen supply, then selection should favour longer leaf life spans.


Do the relationships between life span and LMA, and between life span and foliar N concentration, accord with those reported by Reich et al. (1992) for a larger number of woody plants? To compare herbaceous and woody growth forms, I included available data from low-elevation species from the temperate zone (Diemer et al. 1992) with the latitudinal data set. All data were log-transformed and converted to the units utilized by Reich, Walters & Ellsworth (1992), Reich, Ellsworth & Walters (1998) and Reich, Walters et al. 1998), and therefore I utilized SLA rather than LMA. Results are tabulated in Table 6. As predicted by Reich (1993), a close conformity exists with respect to log10 life span and log10 SLA between the latitudinal herbaceous data set and the data utilized by Reich, Ellsworth & Walters (1998) and Reich, Walters et al. (1998), e.g. those of Reich et al. (1992) and further data, which were substantiated with t-tests of slopes and intercepts of the respective equations (P > 0·15, two-tailed t-tests, Fig. 4a). However, in herbaceous taxa the correlation between log10 life span and log10 foliar N concentration differs significantly, both with respect to slope and intercept, from that for woody plants (Fig. 4B, P < 0·01 and < 0·001, respectively, t-test). This suggests that structural constraints of life-span evolution (reflected in LMA) are independent of life form. On the other hand, differences occur between herbaceous and woody plants with respect to allocation of foliar N (Table 6, Fig. 4b). Among species with particularly short-lived leaves, herbs contain more N per unit mass than woody plants (Fig. 4b). This may reflect inherently higher rates of net photosynthesis of herbaceous plants and/or storage of N in the form of Rubisco. The steeper slope of the relationship also indicates that among taxa with long-lived leaves, herbaceous plants invest proportionally less N per unit mass than trees. Thus, although herbaceous plants conform to the general hypotheses concerning the inter-relationships of leaf life span and leaf traits, there are some characteristics which set them apart from woody plants (e.g. foliar N concentration). Furthermore, their world-wide abundance in thermally similar environments permits an examination of biogeographical trends associated with the evolution of life spans, growth and functioning of leaves, as well as a suite of further hypotheses involving ecology, genetics and population biology.

Table 6.  . Test of the general inter-relationships between leaf life span and specific leaf area (SLA) and mass-based foliar N contents (Nmass) as proposed by Reich, Ellsworth & Walters (1998) and Reich, Walters et al. (1998), utilizing their data and the latitudinal herbaceous data set, which includes low-elevation species as well. Units expressed in accordance with Reich are leaf life span in months, SLA in cm2 g–1 and Nmass in mg g–1. Results of linear regressions of log10 transformed values utilizing log10 leaf life span as independent variable Thumbnail image of


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Thanks are owing to Ch. Körner, who provided an impetus to this study, facilitated the initial contacts and preliminary measurements in Ecuador in 1991 and reviewed the manuscript. S. Prock kindly provided life span data from subarctic and arctic zones, while P. Reich supplied unpublished statistical data for life-form comparisons and commented on the manuscript. P. Grubb and B. Schmid thoroughly reviewed the manuscript, while M. Fischer and D. Birrer contributed advice on use of mortality measures and allometric tests. S. León and B. Øllgard identified and confirmed the páramo plant species. S. Pelaez-Riedl and P. Bockmühl helped with the processing of laboratory samples and N analyses. Funding was provided by the Schweizerischer Nationalfonds (P31–39493·93 to MD).

  1. Present address: Institut für Umweltwissenschaften, Universität Zürich, Winterthurerstrasse 190, CH 8057 Zürich, Switzerland.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Aerts, R. & De Caluwe, H. (1995) Interspecific and intraspecific differences in shoot and leaf lifespan of four Carex species which differ in maximum dry matter production. Oecologia 12, 467477.
  • 2
    Baruch, Z. (1979) Elevational differentiation in Espeletia schultzii (Compositae), a giant rosette plant of the Venezuelan paramos. Ecology 12, 8598.
  • 3
    Bell, K.L. & Bliss, L.C. (1977) Overwinter phenology of plants in a polar semi-desert. Arctic 12, 118121.
  • 4
    Billings, W.D. (1974) Arctic and alpine vegetation: plant adaptations to cold summer climates. Arctic and Alpine Environments (eds J. D. Ives & R. G. Barry), pp. 403–443. Methuen, London.
  • 5
    Chabot, B.F. & Hicks, D.J. (1982) The ecology of leaf life spans. Annual Review of Ecology and Systematics 12, 229259.
  • 6
    Coley, P.D., Bryant, J.P., Chapin, F.S. (1985) Resource availability and plant antiherbivore defense. Science 12, 895899.
  • 7
    Cuatrecasas, J. (1968) Páramo vegetation and its lifeforms. Geo-ecology of the Mountaineous Regions of the Tropical America (ed. C. Troll), pp. 163–186. Colloqium Geographicum 9.
  • 8
    Diemer, M. (1990) Die Kohlenstoffbilanz von Blättern krautiger Pflanzen aus dem Hochgebirge und der Niederung. PhD thesis, University of Innsbruck.
  • 9
    Diemer, M. (1996) Microclimatic convergence of high-elevation tropical páramo and temperate zone alpine environments. Journal of Vegetation Science 12, 821830.
  • 10
    Diemer, M. & Körner, Ch. (1996) Lifetime leaf carbon balances of herbaceous perennial plants from low and high altitudes in the central Alps. Functional Ecology 12, 3343.
  • 11
    Diemer, M., Körner, Ch., Prock, S. (1992) Leaf life spans in wild perennial herbaceous plants: a survey and attempts at a functional interpretation. Oecologia 12, 1016.
  • 12
    Egli, P. (1994) An individual-based analysis of annual shoot and stand development in the invading species Solidago altissima L. PhD thesis, University of Basel.
  • 13
    Grime, J.P. (1979) Plant Strategies and Vegetation Processes. Wiley, Chichester.
  • 14
    Grubb, P.J. (1992) A positive distrust in simplicity—lessons from plant defences and from competition among plants and among animals. Journal of Ecology 12, 585610.
  • 15
    Harling, G. & Sparre, B. (1973) – present) Flora of Ecuador. Opera Botanica Series B.
  • 16
    Harper, J.L. (1988) Canopies as populations. Plant Canopies: Their Growth, Form and Function (eds G. Russell, B. Marshall & P. G. Jarvis), pp. 105–128. SEB Symposium 31. Cambridge University Press, Cambridge.
  • 17
    Harper, J.L. (1989) The value of a leaf. Oecologia 12, 5358.
  • 18
    Hodgson, J.P. & Booth, R.E. (1993) Longevity. Methods in Comparative Plant Ecology (eds G. A. F. Hendry & J. P. Grime), pp. 115–117. Chapman & Hall, London.
  • 19
    Jurik, T.W. & Chabot, B.F. (1986) Leaf dynamics and profitability in wild strawberries. Oecologia 12, 296304.
  • 20
    Karlsson, P.S. (1992) Leaf longevity in evergreen shrubs: variation within and among European species. Oecologia 12, 346349.
  • 21
    Kikuzawa, K. (1991) A cost–benefit analysis of leaf habit and leaf longevity of trees and their geographical pattern. American Naturalist 12, 12501263.
  • 22
    Koike, T. (1988) Leaf structure and photosynthetic performance as related to the forest succession of deciduous broad-leaved trees. Plant Species Biology 12, 7787.
  • 23
    Körner, Ch. (1989) The nutritional status of plants from high altitudes. Oecologia 12, 379391.
  • 24
    Körner, Ch. & Diemer, M. (1987) In situ photosynthetic responses to light, temperature and carbon dioxide in herbaceous plants from low and high altitude. Functional Ecology 12, 179194.
  • 25
    Körner, Ch. & Larcher, W. (1988) Plant life in cold climates. Plants and Temperature (eds S. Long & F. I. Woodward), pp. 25–57. SEB Symposium 42. CBL, Cambridge.
  • 26
    Körner, Ch., Allison, A., Hilscher, H. (1983) Altitudinal variation of leaf diffusive conductance and leaf anatomy in heliophytes of montane New Guinea and their interrelation with microclimate. Flora 12, 91135.
  • 27
    Körner, Ch., Neumayer, M., Pelaez Menendez-Riedl, S., Smeets-Scheel, A. (1989) Functional morphology of mountain plants. Flora 12, 353383.
  • 28
    Larcher, W. (1975) Pflanzenökologische Beobachtungen in der Páramostufe der venezolanischen Anden. Anzeiger der math.-naturw. Klasse der Österreichischen Akademie der Wissenschaften 12, 194213.
  • 29
    Marrs, R.H., Proctor, J., Heaney, A., Mountford, M.D. (1988) Changes in soil nitrogen-mineralization and nitrification along an altitudinal transect in tropical rain forest in Costa Rica. Journal of Ecology 12, 466482.
  • 30
    Mitchley, J. (1988) Control of relative abundance of perennials in chalk grassland in Southern England. III. Shoot phenology. Journal of Ecology 12, 607616.
  • 31
    Mooney, H.A. & Gulmon, S.L. (1982) Constraints on leaf structure and function in reference to herbivory. Bioscience 12, 198206.
  • 32
    Prock, S. (1994) Vergleichende ökologische Unter-suchungen zur Phänologie, Blattlebensdauer und Bio-masse-allokation von krautigen Tal- und Gebirgs-pflanzen aus der Alpenregion, der Subarktis und der Arktis. PhD thesis, University of Innsbruck.
  • 33
    Prock, S. & Körner, Ch. (1996) A cross-continental comparison of phenology, leaf dynamics and dry matter allocation in arctic and temperate zone herbaceous plants from contrasting altitudes. Ecological Bulletins 12, 93103.
  • 34
    Rada, F., Gonzalez, J., Azocar, A., Briceño, B., Jaimez R. (1992) Net photosynthesis–leaf temperature relations in plant species with different height along an altitudinal gradient. Acta Oecologica 12, 535542.
  • 35
    Rauh, W. (1988) Tropische Hochgebirgspflanzen. Springer, Berlin.
  • 36
    Reich, P.B. (1993) Reconciling apparent discrepancies among alpine, temperate and tropical studies relating leaf life-span, structure and function: ‘the blind men and the elephant retold’. Functional Ecology 12, 721725.
  • 37
    Reich, P.B., Uhl, C., Walters, M.B., Ellsworth, D.S. (1991) Leaf lifespan as a determinant of leaf structure and function among 23 tree species in Amazonian forest communities. Oecologia 12, 1624.
  • 38
    Reich, P.B., Walters, M.B., Ellsworth, D.S. (1992) Leaf life-span in relation to leaf, plant and stand processes in diverse ecosystems. Ecological Monographs 12, 365392.
  • 39
    Reich, P.B., Ellsworth, D.S., Walters, M.B. (1998) Specific leaf area regulates photosynthesis-N relations: global evidence from within and across species and functional groups. Functional Ecology (submitted manuscript).
  • 40
    Reich, P.B., Walters, M.B., Ellsworth, D.S., Vose, J., Volin, J., Gresham, C., Bowman, W. (1998) Relationships of leaf dark respiration to leaf N, SLA, and life-span: a test across biomes and functional groups. Oecologia (submitted manuscript).
  • 41
    Sarmiento, G. (1986) Ecological features of climate in high tropical mountains. High Altitude Tropical Biogegraphy (eds F. Vuilleumier & M. Monasterio), pp. 11–45. Oxford University Press, Oxford.
  • 42
    Schmid, B. & Bazzaz, F.A. (1994) Crown construction, leaf dynamics, and carbon gain in two perennials with contrasting architecture. Ecological Monographs 12, 177203.
  • 43
    Schmid, B., Miao, S.L., Bazzaz, F.A. (1990) Effects of simulated root herbivory and fertilizer application on growth and biomass allocation in the clonal perennial Solidago canadensis. Oecologia 12, 915.
  • 44
    Schulze, E.D., Fuchs, M., Fuchs, M.I. (1977) Spatial distribution of photosynthetic capacity and performance in a mountain spruce forest of Northern Germany. Oecologia 12, 239248.
  • 45
    Shaver, G.R. (1981) Mineral nutrition and leaf longevity in an evergreen shrub, Ledum palustre ssp. decumbens. Oecologia 12, 362365.
  • 46
    Sheil, D., Burslem, D.F.R.P., Alder, D. (1995) The interpretation and misinterpretation of mortality rate measures. Journal of Ecology 12, 331333.
  • 47
    Small, E. (1972) Photosynthetic rates in relation to nitrogen recycling as an adaptation to nutrient deficiency in peat bog plants. Canadian Journal of Botany 12, 22272233.
  • 48
    Southwood, T.R.E., Brown, V.K., Reader, P.M. (1986) Leaf palatability, life expectancy and herbivore damage. Oecologia 12, 544548.
  • 49
    Sydes, C.L. (1984) A comparative study of leaf demography in limestone grassland. Journal of Ecology 12, 331345.
  • 50
    Werger, W.J.A. & Hirose, T. (1991) Leaf nitrogen distribution and whole-canopy photosynthetic carbon gain in herbaceous stands. Vegetatio 12, 1120.
  • 51
    Woodward, F.I. (1983) The significance of interspecific differences in specific leaf area to the growth of selected herbaceous species from different altitudes. New Phytologist 12, 313323.