CC BY
21
================ СИСТЕМНОЕ ИЗУЧЕНИЕ АРИДНЫХ ТЕРРИТОРИЙ===============
УДК 581.132+581.5
СТРУКТУРНО-ФУНКЦИОНАЛЬНЫЕ ОСОБЕННОСТИ АДАПТАЦИИ РАСТЕНИЙ ГОБИ К АРИДИЗАЦИИ КЛИМАТА
© 2004. Иванов Л.А.1, Ронжина Д.А.1, Иванова Л.А.2, Чечулин М.Л.2, Белоусов И.В.2,
Гунин П.Д.2, Пьянков В.И.2
1 Ботанический сад-институт, УрО РАН, 620144 Екатеринбург, ул. 8 марта, 202
2 Уральский государственный университет, 620083 Екатеринбург, ул. Ленина, 51
3 Институт экологии и эволюции им. А.Н. Северцова РАН, Москва, Ленинский пр-т, 33
Проведен комплексный анализ структурно-функциональных параметров 108 видов растений из разных типов степных и пустынных сообществ Гоби. Изучены параметры биомассы и ее распределение по органам, морфо-функциональные показатели листьев (площадь, толщина, плотность листа), количественные показатели фотосинтетических тканей (число и размеры клеток и хлоропластов, площадь поверхности мезофилла и внутренняя диффузионная проводимость листа для СО2), химический состав (содержание углерода, азота, минеральных веществ, органических кислот) и конструкционная цена листьев.
Выявлены основные механизмы адаптации растений Гоби к аридному стрессу на разных уровнях организации. По сравнению с растениями бореальной зоны (Средний Урал) у степных и пустынных видов Монголии увеличена доля подземных органов в общей массе (в 1.5-2 раза), снижены доля стеблей (в 1.5-2 раза) и отношение площади листьев к массе целого растения (в 1.5-2 раза). На уровне фотосинтетических тканей адаптация растений к аридному стрессу выражалась в увеличении внутренней ассимиляционной поверхности листьев (площади поверхности мезофилла) в 3-4 раза. Степные и пустынные растения Монголии имели значительные энергетические затраты на построение единицы массы листа (1.41 - 1.56 г глюкозы на 1 г сухого веса).
Выявлено структурное и биохимическое разнообразие растений Гоби, связанное с их дигрессионной активностью. Проведен сравнительный анализ трех групп растений: растения с низкой дигрессионной активностью (N-виды), растения с высокой дигрессионной активностью (D-виды) и рудеральные растения (R-виды) (Gunin, Vostokova, 1989) . Показано, что в ряду N^D^R значительно снижалась доля подземных органов в массе растения и возрастала доля генеративных органов и стеблей. У R-видов была в 2-3 раза больше развита внутрилистовая ассимиляционная поверхность, чем у D- и N-видов, в результате чего рудеральные растения имели вдвое большую проводимость мезофилла для диффузии СО2. Кроме того, R-виды имели меньшее количество углерода и растворимых углеводов в сухой массе листа и большее количество минеральных веществ.
Продемонстрировано, что изменения в распределении биомассы, структуре фотосинтетического аппарата и химическом составе листьев у растений с разной дигрессионной активностью (N, D, R) были сходны с тенденциями для растений с различными типами экологических стратегий Грайма (s -стресс-толеранты, c - конкуренты и r-рудералы. Соответствие между двумя классификациями также проявлялось в сходстве характеристик местообитаний растений и жизненных форм. N-виды, как и s-стратеги, устойчивы к экологическому стрессу (аридность, засоленность) и не выносят антропогенное нарушение. D-виды, как с-стратеги, являются многолетниками с высокой конкурентоспособностью. Большинство R-видов растений Гоби, как и r-стратеги по Грайму, -однолетние растения с высокой скоростью роста, быстро занимают освободившиеся нарушенные территории.
Сделано заключение, что классификация растений Гоби по типам экологических стратегий отражает адаптацию как к аридному стрессу, так и к пастбищной дигрессии. Идентификация типа экологической стратегии позволяет определить функциональную активность вида в сообществе и предсказать реакцию вида при увеличении антропогенного стресса или усилении процессов опустынивания в экосистемах Гоби.
Предложен подход для определения функциональной активности видов растений Монголии, основанный на комплексном анализе количественных показателей на уровне целого растения (индексы распределения биомассы), на уровне анатомии листа (величина внутренней
ассимиляционной поверхности) и биохимического состава растительных тканей (содержание углерода, азота, конструкционная цена растительного материала). Используя эти параметры, можно идентифицировать основные функциональные типы растений экосистем Гоби и место каждого исследованного вида в сукцессионном ряду. Процентное соотношение видов с разной функциональной активностью в растительном сообществе позволит определить уровень дигрессионной нарушенности экосистемы и ее возможные изменения при глобальных климатических и антропогенных воздействиях.
STRUCTURAL AND FUNCTIONAL BASIS OF ADAPTATION OF GOBI
PLANTS TO DESERTIFICATION
© 2004. L.A. Ivanov1, D.A. Ronzhina1, L.A. Ivanova2, I.V.Belousov2, M.L. Chechulin2, P.D. Gunin3, V.I. Pyankov2
1 Botanical Garden-Institute, Ural Division, RAS, 620144 Yekaterinburg, 8 Marta 202;
2 Urals State University, 620083 Yekaterinburg, Lenina 51, 3 Severtsov Institute of Ecology and Evolution, RAS, Moscow, Leninsky prospect 33;
Introduction
The modern tendency of environmental changes can be defined as a climate aridization and an expansion of droughty areas (Verstraete, Schwartz, 1991; Pyankov, Mokronosov, 1993). The alteration of temperature and water regime can cause a shift of botanical-geographical zones, a displacement of plant species areas and a changing of interspecies competitive relations in communities (Pyankov, Mokronosov, 1993).
Global fluctuations connected with warming or cold spell of the climate already occurred during the history of the Earth. But they had slow rates (thousands - tens thousands years) and left an opportunity for evolutionary transformations of the vegetation. The modern climatic alterations are relatively fast (tens -hundreds of years) (Verstraete, Schwartz, 1991) and do not allow ecosystems to adapt to them. In this connection the probable way of community adaptation will consist in a change of plant species areas and in a loss of the most sensitive species in ecosystems. Therefore it is necessary to find out the objective approach for evaluation of vegetation state and forecasting of its changes. Recently, as a potential tool for this goals the concept of plant functional types (PFTs) has attached much attention. The most researchers agree that PFT are characterized by similar ecological features and have a similar reaction on environmental conditions changes (Box, 1996; Steffen, 1996; Lavorel et al. 1997). The identification of PFT depends on botanically-geographycal zones, scales, and purpose of investigations.
Many recent empirical studies have dedicated revealing of plant functional diversity in grasslands (Tilman, Knops, 1997; Bullock et al., 2001; Diaz et al., 2001; Mclntyre, Lavorel, 2001; Vesk, Westoby 2001). Tilman, Knops (1997) identified functional groups on the basis of morphological and physiological differences: legumes, C3- and C4-grasses, forbs, woody plants. They showed that functional diversity had greater impact on ecosystem processes than species diversity. At the same time such classification is not convenient in all cases, because taxonomic criterion such family (grasses and legumes) or simple morphologic criterion such life form (forbs, woody plants) are too common, formal characteristics. These signs do not permit to predict species reaction on anthropogenic influence, for example, grazing.
Some researchers divide pasture plants into groups grazing increasers or decreasers (Vesk, Westoby, 2001). But this classification does not allow to make single-meaning description of plants response to grazing, because many species may act as increasers and also as decreasers, and this depends on climatic or succession conditions (Vesk, Westoby, 2001). Many authors agreed that two functional groups correspond to opposite points of grazing gradient: ruderal species increasing by grazing intensification and palatable grasses increasing by grazing absence (Fernandez-Gimenez, Allen-Diaz, 2000; Gunin, Vostokova, 1989; Nechaeva, 1979). In this case the question retains how to characterize a great number of other species.
The following important problem is the revealing of criteria on which it is possible to identify PFT. Functional activity of species is caused by its physiological peculiarities. Some studies have showed possibilities to use the growth, morphological and physiological parameters for determination of functional features of plant species in grasslands. Plant height is good predictor of grazing response (Diaz et al., 2001), biomass rate reflects competitive ability of plants. N.N. Slemnev (1996) showed possibility to use
physiological parameters such transpiration rate, leaf water content, biomass increase for determination of a role of Mongolian species in steppe and desert communities. At the same time physiological process rate is very labile and reflects the adaptation of a separate plant function to the current environmental conditions. Therefore it is difficult to use it for estimation of the species "climatic opportunities" (Pyankov, Mokronosov, 1993), i.e. features characterizing long-term adaptation to whole complex of environmental conditions. The analysis of stable features and structures connected with the basic physiological functions is necessary.
Morphological and anatomical features of leaves such SLA, leaf area and toughness may be used for evaluation of grazing response (Diaz et al., 2001) and plant ecology strategy (Westoby, 1998). For some species it is showed the correlation between SLA (specific leaf area, i.e. leaf area per mass) and relative growth rates (Garnier, 1992; Lambers, Poorter, 1992). It was proposed also connection between SLA and ecological strategy types (Westoby, 1998). However presence or absence of this connection depends on PFT and leaf anatomy. For example it was shown strong positive correlation between SLA and photosynthetic tissues development in ruderal species, and absence of this correlation in stress-tolerant species of boreal zone (Pyankov et al., 1998). Plants with different types of ecological strategies in boreal zone did not differ on SLA (Pyankov et al., 1998; Pyankov, Ivanov, 2000). SLA was a comparatively poor predictor of grazing response in study in grasslands of Argentina (Diaz et al., 2001).
Thus, it is difficult to estimate functional peculiarities of species by lonely parameter and is necessary to use a complex of structural and functional parameters on different levels of plant organization.
V.I Pyankov developed a special system of plant characteristics that permits to reveal climatic and ecological patterns of plant distribution (Pyankov, Mokronosov, 1993; Pyankov at al., 1998; Pyankov, Ivanov, 2000; Pyankov et al. 2001a). This system include studying of different levels of plants organization: a) at the level of the whole plant - determination of total biomass and its allocation; b) at the level of photosynthetic organ - content of main chemical compounds; c) on the level of certain photosynthetic tissues and chloroplasts - mesophyll anatomy, mesophyll conductance for CO2-transport, chloroplast number and sizes. The total biomass and biomass distribution among different organs of plants can be used for identification of the types of ecological strategies (Pyankov, Ivanov, 2000). Elucidation of the plants strategies enables to estimate the potential resistance to unfavorable natural and anthropogenic factors. Carbon and nitrogen contents in leaves of plants were used for calculation of carbon cost, i.e. the amount of energy necessary for construction of plant mass unit (Pyankov et al., 2001b). A strong correlation between carbon cost and ecological strategies and functional activity of plants were shown. The quantitative characteristic of leaf mesophyll is the complex of parameters (more then 20) of photosynthetic tissues structure. It was shown (Pyankov et al., 1998) that the structure of the photosynthetic apparatus grows out long-term adaptation of species to environmental conditions and reflects functional features of plant species.
The complex analysis including all criteria named above can be used for the identification of natural plants groups - plant functional types. The basic purpose of the paper is to determine functional activity of Mongolian dominate species in different types of ecosystems for evaluation of the plant species resistance to ecological and anthropogenic stress.
Plant materials and methods
The investigations were done in summers of 2000-2003, in Mongolia. The studied places located in range from 480 to 430 N longitude and from 990 to 1070 E latitudes at elevations from 750 to 2000 m. Plants material was collected in July-August, mainly at the stage of flowering. We investigated 108 plant species from 25 families occurring in different types of steppe and desert: Darchan (steppe), Ulan-Bator (mountain steppe), Unzhul-somon (dry plate steppe), Bulgan-somon area (desert steppe, mountain dry steppe), Bajan-Dzag valley (sandy and clay desert), Gurvantes (mountain desert steppe), Echin-Gol (oasis).
Whole plant mass and biomass allocation. Five to ten well-developed intact plants of each species growing under typical conditions for this species were used for analysis. The plants were cleaned from soil, dissected into separate organs, dried and weighted. To study the biomass allocation, we estimated the leaf area (by the system of image analysis "Siams Mesoplant", Open company "SIAMS"); the total weight; and the weights of leaves, stems, underground organs, and generative organs. These data were used for calculating integral morphological indices, namely, the leaf mass ratio (LMR); underground organs mass ratio (UMR); stem mass ratio (SMR); generative organs mass ratio (GMR), leaf area ratio (LAR).
Chemical composition. Methods for determination of leaf biochemical composition were described in our paper (Pyankov et al., 2001a). Leaf samples pooled from 10-15 plants of each species fixed at 125oC and
dried at 75oC in thermostat oven. The total nitrogen and carbon content of leaves was measured with a Carlo Erba 1106 automatic CHN analyzer (Italy). The total content of mineral substances was calculated as the sum of ash and NO3 minus a correction for ash alkalinity (Poorter, 1994). The NO3 concentration was measured with SF-46 spectrophotometer in aqueous extracts with a 5% solution of salicylic acid in H2SO4 and in 4% NaOH. The energy cost of a plant leaf weight unit (construction cost, CC), in grams of glucose per gram of leaf dry weight, was calculated by method of Poorter (1994) and described in (Pyankov et al., 2001b).
Table 1. Comparison of Mongolian plants (n=51) with boreal plants (n=45) on biomass allocation parameters, leaf mesophyll structure, concentrations of main chemical substances in the leaves and significance of distinction between these groups. ns, not significant; *, p<0.05; **, p<0.01; ***, p<0.001. P refers to data for palisade cells. The values include 50th percentiles of species (from the 25th to 75th percentiles of the distribution). (1)Data are adduced from Pyankov, Ivanov, 2000. (2)Data are adduced from Pyankov et al., 1998. (3)Data are adduced from Pyankov et al., 2001a, 2000b.
Parameter Mongolia, Gobi Boreal zone, Criterion
Middle Ural Mann-Whitney
Total weight, g 1-4 1-7(1) ns
Underground organ weight, g 0.5-2.0 0.5-2.0(1) ns
Stem weight, g 0.5-1.0 0.8-2.0(1) ns
Leaf weight, g 0.5-1.0 0.5-1.5(1) ns
Generative organ weight, g 0.1-0.4 0.1-0.5(1) ns
Leaf area ratio, cm2/g 0.2-0.4 0.4-0.9(1) ***
Underground organ mass ratio, % 25-55 10-40(1) ***
Stem mass ratio, % 15-25 25-45(1) ***
Leaf mass ratio, % 20-35 20-30(1) ns
Generative organ mass ratio, % 6-16 6-15(1) ns
Leaf area, cm2 0.3 - 1.4 5.0 - 45.0(2) **
Leaf thickness, mkm 340-800 150-250(2) ***
Specific leaf mass, mg/cm2 800-1220 290-550(2) ***
Cell volume (P), 103 mkm3 4-17 5-15(2) ns
Number of chloroplasts per cell 20-60 20-50(2) ns
Cell number, 103/cm2 600-1700 200-700(2) ***
Chloroplast number, 106/cm2 25-50 6-20(2) ***
Chloroplast volume, mkm3 20-40 30-50(2) *
Achl/A, cm2/cm2 10-20 4-10(2) ***
Ames/A, cm2/cm2 14-32 4-9(2) ***
Mesophyll volume in leaf, % 7-25 7-22(2) ns
Mesophyll conductance, cm/sec 0.7-1.7 0.2-0.5(2) ***
Carbon 430-470 430-470(3) ns
Nitrogen 25-35 20-30(3) *
Carbon/nitrogen 11-18 13-22(3) ns
Mineral 60-105 50-85(3) ***
Nitrates 3.6-7.6 2.5-5.7(3) **
Organic acids 55-100 70-110(3) ns
Soluble carbohydrates 60-150 35-75(3) **
Quantitative leaf anatomy. Structural parameters of the photosynthetic apparatus were determined according to Mokronosov's method (Pyankov et al., 1998) and a projective method (Ivanova, Pyankov, 2002). For analysis, we took five to ten leaves from each of ten to fifteen plants. Leaf area, leaf specific mass (leaf mass per area, LMA) and leaf thickness were determined by the system of images analysis "Siams Mesoplant" in module "Macro". Leaf thickness was determined by examining leaf cross-sections, placed in Tris-HCl-sorbitol buffer (pH 7.4) in ten replications. The number of cells per unit leaf area was determined in samples fixed with 3.5% glutaraldehyde dissolved in phosphate buffer (pH 7.0). Fixed leaf fragments were macerated with 20% KOH under heating. The number of cells in macerate was counted in 20 replicates for 90 squares of the Goryaev haemocytomer. The volume of mesophyll cells was determined by the projective method (Ivanova, Pyankov, 2002).
The significance of differences between the parameters studied in the groups of species was estimated using Mann-Whitney test.
Whole plant mass and its allocation
The absolute values of total biomass in Mongolian plant species ranged from 0.5 g dry mass in Mollugo cerviana to about 100 g in Ephedra equsetina. 75% of the species has this parameter less than 4 g (table 1). We found significant positive correlations between total weight of whole plant and the weight of individual organs (stem, leaves, underground and generative organs) - correlation coefficient r ranged from 0.83 to 0.97 at <0.05.
The UMR varied from 5.5% in annual species Tribulus terrestris to 86.2% in perennial plant Poligonum angustifolium. Unlike this parameter, the minimal LMR and SMR were observed in Poligonum angustifolium (8.3, 4.5%, respectively) and maximal values were in Mollugo serviana (59% - LMR) and Tribulus terrestris (47% - SMR). The stem weight and leaf weight did no exceed 1 g in 75% of the species (table 1). The weight of generative organs contributed the least in total weight of the whole plant. 50% of the species examined had GMR from 6 to 16%.
Comparative analysis results of the biomass allocation of Mongolian plant and Middle Ural plants species showed that adaptation of Mongolian plants to arid environment expressed in the increasing of UMR and decreasing of SMR and LAR. The forming of large underground organs allows Gobi plants to improve water supply and to store nutritional substances for surviving plants during unfavorable periods. Reduction of leaf area per unit dry weight of plant connected with the necessity of decreasing water loss in aridity stress condition. Pyankov and Ivanov (2000) showed with boreal plant species analysis that plants with different types of ecological strategies (S-stress-tolerators, C-competitors, and R-ruderals) have differences in biomass allocation. The increasing of UMR and decreasing of SMR and LAR were found for S-strategists. This fact allows concluding that existence in severe arid stress conditions results in forming stress-tolerators properties in biomass allocation of Mongolian plant species.
Parameters of biomass allocation were used for estimation of the functional activity of plant species and evaluation of their state in succession rows. We applied classification of species-indicators of pasture digression in Mongolia according Gunin and Vostokova (1989) and divided species in relation to their digressive activity on three groups: plants with low digressive activity, mainly native dominant species (N); species with high digressive activity (D), and ruderal plants, which abundantly occupied overgrazed pastures
(R) (fig. l).
В
&
i
cn 5
• %
□
qP
J? 4
-5
-4
-3
-2
o -4—
Rootmass ratto (OjK). LAR MJEi.geierati); urgais mass ratto (ИЗ)
Discriminant function 1
Fig. 1. Box plots of the biomass allocation parameters in Mongolian plants with different functional activity: native dominant species with low digressive activity (N, n=14), species with high digressive activity (D, n=4), ruderal plants (R, n=4). The species in the groups are presented according Gunin, Vostokova (1989). The square symbol in each box indicate the median value, the bottom and top parts of the box the 25th and 75th percentiles, and the bottom and top of the bars minimal and maximal values.
Comparative analysis of biomass allocation for Mongolian plant species studied showed that species from groups with distinct digressive activity significantly differed from each other with respect to the UMR, SMR, and GMR (table 2). The UMR significantly decreased in order N^D^R. Unlike this, the GMR and
SMR increased in order N^D^R. Therefore these indices may be used as criteria for identification of species with different digressive activity. We applied the procedure of discriminant analysis based on plant biomass parameters for identification digressive activity of Mongolian plants, which activity was unknown. We used the values of discriminant functions to determinate a belonging of species to the certain class:
Root1 = -2.6 + 0.13*LAR + 0.05*SW + 0.10*UMR - 0.07*GMR,
Root2 = 4.4 - 6.70*LAR + 0.19*SW- 0.03*UMR - 0.14*GMR,
where LAR - leaf area ratio (the ratio of common area of leaves to total plant weight), SW - stems weight, UMR - underground organs mass ratio, GMR - generative organs mass ratio.
The results of discriminant analysis are represented at figure 1. The most significant for distinguishing groups by function 1 was UMR (r=0.95). Function 2 associated with the stem weight (r=0.62) and LAR (r=-0.45). The levels of digressive activity of species, which functional activity are unknown, were determined on the basis of the coordinates for each species in the space of discriminant functions (fig. 1). The results obtained indicate that species with different digressive activity may be characterized by different contributions of their organs (particularly underground organs, stems, and generative organs) to the total plant weight. In connection with this the biomass allocation parameters may be apply for identification the levels of digressive activity in other species with unknown functional activity.
Structural and functional adaptation of photosynthetic apparatus
Table 1 represents the main parameters of photosynthetic apparatus in Mongolian plants studied. Mongolian plant species examined have the leaf area which varied from 0.04 cm2 to 8.90 cm2. In 50% of the studied species the leaf area did not exceed 1.40 cm2. The leaf thickness of Gobi plants varied from 210 to 2250 мш. Maximal values of this parameter had succulents such Bassia dasyphylla, Haloxylon ammodendron, Reaumuria soongorica. Like this parameter the specific leaf weight (SLW) varied also in 10fold range - from 230 to 2560 mg/dm2. Mesophyll cell volume varied from 1 300 to 160 000 мш3. The minimal cell volume was characteristic for grasses, and the largest cells were observed in succulents. The Gobi plant species had also different values of chloroplast number in mesophyll cell (minimal value -10 and maximal value 285) and per unit leaf area (from 6.9 million up to 120 million per 1 cm2 of the leaf). Leaf mesophyll conductance for CO2 differed also in the large limits among species studied, from the lowest 0.3 cm/sec up to the highest 4.1 cm/sec. Thus, Mongolian plant species characterized by the great variety of morphological, structural and functional parameters of the photosynthetic organs. This is probably connected with a large number of ecological possibilities and adaptation mechanisms among Gobi species to aridity stress. A common anatomical research of Gobi plants showed that there is a large diversity of the xerophytisation forms in desert ecosystems, which corresponds to a variety of ecological niches (Gamalei, 1988). Our study results show also a great diversity of the ways in structural and functional adaptation of photosynthetic apparatus in Gobi plants, but they have common futures in comparison to plants from humid region (table 1).
Adaptation of Mongolian plants to arid conditions involves different levels of organization, including leaf morphology, photosynthetic tissues, cells and chloroplasts parameters. The morphological adaptation of Mongolian plants' leaves to aridity stress is expressed in decreasing leaf area and in enhancing leaf thickness and density (table 1). The higher weight of unit leaf area was caused not only by developed mechanical, conducting and covering tissues apparently, as it was already shown (Gamalei, 1988; Voronin et al., 2003), but also by the structural adaptation of photosynthetic tissues to aridity stress consisted in change of quantitative parameters. Species of arid and humid regions did not differ in cell volumes, number chloroplasts per cell, and mesophyll volume of leaf. However Gobi plants had larger values of photosynthetic cell and chloroplast number per unit leaf area. The increasing of the mesophyll cells and chloroplasts number is very important for formation of the inner leaf surfaces that are common membrane for CO2 fixation and water transpiration, i.e., responsible for plant carbon/water balance. Our study showed that these parameters were two - three times higher for Mongolian plants than for boreal plants. Accordingly, mesophyll conductance gm for Mongolian plants was approximately three times higher than for boreal plants (table 1).
At the same time there was a large diversity of quantitative parameters of photosynthetic apparatus among Mongolian plants. We found significant differences between groups with different digressive activity, which was identified according (Gunin, Vostokova, 1989).
The native dominant species in comparison with D- and R-plants had larger cell sizes (about 3-5 times) and smaller volume of chloroplasts in cell (about 3 times; table 2). Ruderal plants had many chloroplasts per
cell and the most cells and chloroplasts per unit leaf area. Consequently, ruderals had the largest inner surface area of mesophyll cells and chloroplasts. These parameters - Ames/A and Achl/A - were about 2-3 times grater than in D- and N-species. The group of species with high digressive activity (D-species) was characterized by the low cell sizes with low number of chloroplasts in cell and high mesophyll cell number. D-species had the lowest level of variety of mesophyll structure parameters. Mesophyll conductance gm increased in the row N^D^R.
Table 2. Comparison of plants with different digressive activity and ecological strategy on morphological, structural and biochemical and physiological parameters. *Data are adduced from references: 1 - Grime, 1974; 2 - Gunin, Vostokova, 1989; 3 - Pyankov et al., 1998; 4 - Pyankov, Ivanov, 2000; 5 - Pyankov et al., 2001a; 6 - Pyankov et al., 2000b.
Parameter Digressive activity, Mongolia Ecological strategy, Boreal zone, Middle Ural
Habitat N(2) High stress, No grazing S(1) High stress, Low disturbance
D(2) High stress, High grazing C(1) Low stress, Low disturbance
R(2) High stress, overgrazing R(1) Low stress, High disturbance
Life form N(2) Shrubs, perennial herbs S(1) Shrubs, perennial herbs
D(2) Shrubs, perennial herbs C(1) Trees, shrubs, perennial herbs
R(2) Majority annual herbs R(1) Majority annual herbs
Total plant weight, mg/g N 0.5-5.0 S (4) 0.5-4.0
D 4-45 C (4) 8-73
R 5-6 R (4) 1-2
Underground organ mass ratio, % N 45-65 S (4) 25-50
D 25-50 C (4) 25-50
R 10-15 R (4) 4-7
Stem mass ratio, % N 12-25 S (4) 15-25
D 23-33 C (4) 24-42
R 30-50 R (4) 40-45
Mesophyll cell volume, 103 мm3 N 4-21 S (3) 8-36
D 2-8 C (3) 2-4
R 5-50 R (3) 10-20
Chloroplast number per cell N 20-50 S (3) 40-60
D 17-30 C (3) 16-27
R 35-110 R (3) 30-73
Ames/A N 20-27 S (3) 6-14
D 21-28 C (3) 10-14
R 28-65 R (3) 14-22
Mesophyll conductance, cm/sec N 0.9-1.7 S (3) 0.3-0.5
D 0.5-1.2 C (3) 0.5-0.8
R 1.0-2.5 R (3) 0.7-1.2
C/N N 7-20 S (5) 17-34
D 15-20 C (5) 13-23
R 11-14 R (5) 10-14
Mineral substances, mg/g Soluble carbohydrates, mg/g Construction cost, g glucose/g N 70-100 S (5) 30-60
D 60-80 C (5) 55-80
R 80-200 R (5) 85-120
N 90-150 S (5) 50-70
D 70-100 C (5) 25-75
R 50-120 R (5) 20-30
N 1.45-1.60 S (6) 1.45-1.55
D 1.40-1.45 C (6) 1.28-1.45
R 1.30-1.50 R (6) 1.11-1.37
Thus, we found out the significant divergence between Mongolian plans with different digressive activity. Figure 2 shows the results of discriminant analysis for Mongolian plants based on the parameters of photosynthetic tissues. Plants with different digressive activity were separated by discriminant functions: Root1 = 20,7 + 0,44* VcelP - 0.28*ChlP - 0.07* Vchl - 0.40*gm,
Root2 = 1.45 + 0.11* Vchl - 0.15* %ChlP - 0.21*gm,
where VcelP - palisade cell volume, ChlP - chloroplast number in palisade cell, Vchl - chloroplast volume, gm- mesophyll conductance, %ChlP - volume of chloroplasts in cell.
The most significant variables according to discriminant analysis were the same as according to one-dimensional statistical criterions between N-, D- and R-groups. The discriminant analysis mechanism allowed determining the degree of digressive activity of species, which digressive activity was unknown yet (fig. 2).
Z1
О
о
й Ш
Tt-
О
С1
О
-12
■ш
-4
<Б
О
о D
■ N
о R
V $ ci
" i
аГ
О
TU
Чп
« )U Si „ $
t i.-i ■ ш а ■ ■
■ Î J':, ■ S ■ - #2®® я
К ■
-2
Volume of chlorophsts in cell (-0.35), mesophyll conductance (-0.3S)
ROOT1
Fig. 2. Discriminant analysis of species with different functional activity on the basis of biomass allocation parameters. Species numbers are presented in table 3. N - native dominant species with low digressive activity, D - species with high digressive activity, R - ruderal plants. Correlation coefficients between discriminant functions and variables are shown in brackets.
Chemical composition and construction cost of leaves
Chemical composition (nitrogen, carbon, minerals, NO3, organic acids) and construction cost were determined in the leaves of the same 54 species which functional activity identified on the basis of biomass parameters.
The content of soluble carbohydrates, nonstructural polysaccharides, organic acids and minerals differed considerably among studied Gobi plant species (table 1). The concentration of minerals and nitrates in leaves was minimal for perennial plant Sophora alopecuroides (39, 1.8 mg/g, respectively) and maximal for annual species Tribulus terrestris (547, 22 mg/g, respectively) and Chenopodium acuminatum (271, 50 mg/g, respectively). In 75% of the species content of mineral substances and nitrates did no exceed 85 and 7.6 mg/g respectively. Leaf carbon concentration varied less considerably then content of other substances. Values of this parameter ranged from 267 mg/g in Zygophyllum potaninii to 496 mg/g in Veronica incana. In 50% of the species studied, leaf carbon concentration varied from 430 to 470 mg/g.
The results of comparison of leaf chemical composition for Gobi plant species and Middle Ural plants are shown in table 1. The leaves of Mongolian plants had higher concentration of nitrogen, nitrates, mineral substances and soluble carbohydrates then ones in Middle Ural plants. Earlier it was shown (Grime et al., 1988; Garnier et al., 1999) that increased nitrogen concentration in leaves is depended on the high content of functional proteins and resulted in high rate of CO2 fixation. In our opinion enhanced nitrogen concentration in leaves of Mongolian plant species is not connected with high content of functional protein, but it was a consequence of increased nitrate concentration (table 1). Accumulation of nitrate, minerals and soluble carbohydrates in leaf cells may provide high suck power for improving water supply.
The leaf construction cost was higher for Mongolian plant species than for Middle Ural plants (table 1). Pyankov et. al. (2001b) demonstrated that maximal values of this parameter were in boreal plants with stresstolerant strategy and is explainable by synthesis of compounds (for example, cuticle waxes) that protect leaves from water loss during drought, freezing, and exposure to ultraviolet radiation. It can be assumed that Mongolian plants manifest high leaf construction cost because their arid stress environment requires the synthesis of specific protective compounds.
Comparative analysis of chemical composition of studied Mongolian plants with different digressive activity showed that species belonging to ruderal plants significantly differed from other species groups (table 1). They had less content of carbon and soluble carbohydrates and higher concentration of mineral substances and nitrates.
This indicates that the digressive activity of species is associated not only with biomass indices and mesophyll structure but also with specific chemical composition of the leaves.
Physiological mechanisms of adaptation to aridity stress condition
The main specific structurally-functional features of adaptation of Gobi plant species to arid stress conditions were found. These adaptation mechanisms are revealed at the different levels of plant organization - total plant biomass, photosynthetic organs morphology, photosynthetic tissues structure and leaf biochemical composition. This complex of plant traits expresses the efficiency of environmental recourses utilization, i.e. ecological strategy of plant species.
A concept of plant ecological strategies was proposed and empirically justified by Ramenskii (1935) and quantitatively elaborated by Grime independently (1979). According to Ramenskii-Grime's approach (Ramenskii, 1935; Grime, 1979) three types of strategies exist among plants - the competitors, ruderals and stress-tolerators (figure 3a). J.P. Grime (1979) determined the types of plant ecological strategies analyzing relative growth rates and morphological indexes of species. Pyankov et al. (1998, 2000, 2001a, 2001b) showed that complex analysis of structural and functional parameters of plants on several organization levels allows to identify the types of plant strategies in natural habitats. These studies found out the main traits of the competitors, ruderals and stress-tolerant species of boreal plants. The manifestation of stress-tolerant features were associated with the high UMR, low SMR and LAR (Pyankov, Ivanov, 2000), small leaves with high SLM (Pyankov et al., 1998) and high construction cost (Pyankov et al., 2001b). Our investigations showed that in comparison to boreal plants Gobi plants had strongly expressed stress-tolerant features, which were associated with increased UMR, decreased SMR and LAR, small leaf area with high SLM and enhanced construction cost (table 1).
On the contrary to boreal species peculiarity of photosynthetic tissues structure in Mongolian plants was large number of photosynthetic cells and chloroplasts per leaf area - 2-3 times greater than in boreal plants (table 1). As the mesophyll cell and chloroplast sizes in Gobi and Ural plants were the same, so derivatives of numbers and sizes of cells and chloroplasts - total surface area of cells and chloroplasts (Ames/A and Achl/A) - were also 2-3 times greater in Gobi plants than in boreal species. At the same time stress-tolerant species of boreal zones characterized by the lowest number of cells and chloroplasts and the lowest values of Ames/A and Achl/A (Pyankov et al., 1998). These distinctions are connected with a structural adaptation of photosynthetic tissues to different kinds of stress. Stress-tolerators of boreal zone are mainly shade plants and their photosynthetic organs are adapted to low light intensity under a forest canopy. Therefore they have thin leaf blade and homogenous mesophyll with large cells. Gobi plants are adapted to another kind of stress, so-called arid stress, which is associated with the water deficit, high temperature and irradiance. In agreement with this, Mongolian plants mainly are xerophytes (Gamalei, 1988). Xeromorphy is often expressed in decreasing of leaf surface, enhancing of leaf thickness and density (Vasilevskaja, 1954; Vosnesenskii, 1984; Gamalei, 1988). The leaf surface reduction frequently led to the increasing of cells and chloroplasts number per leaf area unit (Gamalei, 1988). So xerophytes of Karakums desert had 1800-2000 thousand mesophyll cells per 1 cm2 of leaf surface though mesophytes of temperate zone had mainly 400600 thousands of cells per 1 cm2 (Mokronosov, 1978). Besides that a large photosynthetic cell and chloroplast number per leaf area unit is not only consequence of the leaf area decreasing, but also the structural reason for enhancing leaf inner assimilate surface - Ames/A and Achl/A. As result Gobi plants have 2-time greater mesophyll conductance for CO2 than boreal plants (table 1). Thus is very important for support of common leaf CO2 diffusion rate in water deficit conditions, as level of stomata conductance is very low. In this connection we concluded that the great density of cells and chloroplasts, high Ames/A and Achl/A, and large mesophyll conductance in Gobi plants can be considered as the futures of mainly stress-tolerant
species, which are adopted to arid stress.
Functional diversity of Gobi plant species
Our results showed that Gobi plants alongside with the found similarity of features characterized by a structural and biochemical variety connected with a species functional diversity. As all Gobi plants are under the same strong ecological stress and the main changing factor acting on the vegetation is an anthropogenic influence (grazing, trampling), so the functional diversity of Gobi plants depends on en expression of pasture digression. The pasture digression (degradation) is a gradual changing of vegetation by an influence of excessive pasturing. Plants of Gobi's desert steppe are characterized by a different digressive activity - i.e. a reaction of species on different levels of the grazing. Plants with a different digressive activity get an advantage on different stages of plant community succession. The succession under grazing pressure usually occurs on the following way: 1. Stage of primary (native) dominants; 2. Replacement of primary (native) dominants by co-dominants (secondary, pasture dominants) and appearance of weeds (ruderals); 3. Stage of ruderal pasture weeds (Gunin, Vostokova, 1989).
We analyzed the groups of Mongolian species with the different digressive activity and found out definite specific features at different levels (table 2). Tendencies in the change of biomass indices, mesophyll structure, and leaf chemical composition among Mongolian plants with the different digressive activity were similar to the tendencies for plants with different types of ecological strategies demonstrated on boreal species (table 2). The conformity between two classifications is also in some similarity in the characteristics of plant habitats and life forms in plants with the different digressive activity/ecological strategies. A-species like S-strategists submitted by perennial herbs and shrubs prevail at the first stage of pasture digression with a week disturbance degree. Taxonomy of this group is a rather diverse, but it also includes large number of grasses and species of genus Allium. Big perennials and also shrubs and subshrubs represent larger part of D-species. D-species get an advantage at the second stage of degradation and are as a rule co-dominant. The most annuals and taproot perennials are R-species like R-strategists and their proportion enhance at the overgrazing stage.
Therefore results of our comparison suggest that the classification of plants on the type of digressive activity (N, D, R) is consistent with the system of ecological strategies (S, C, R). A-species (low digressive activity) as S-strategists are the most tolerant to ecological stresses (aridity, salinity) in natural habitats of Mongolia, non-enduring anthropogenous disturbance. This species had lower values of total biomass and Ames/A and higher UMR and construction cost in comparison with other two groups - D and R (table 2). D-species (high digressive activity) like C-strategists have a competitive ability and are characterized by the greatest total biomass, leaf area, small mesophyll cells and large their number per leaf area unit. R-species (ruderal plants, high activity on the overgrazed pastures) like R-strategists according to Grime are characterized by low total biomass, the lowest UMR and construction cost, the highest values of Ames/A and mesophyll conductance. This indicates that the groups with the different digressive activity physiologically close to plant groups with the different ecological strategies of Ramenskii-Grime (Ramenskii, 1935; Grime, 1979).
In our opinion, the major differences between the plant species of the different digressive activity/ecological strategies consist in the efficiency of environmental resources utilisation, resistance to environmental stress, and relative growth rate. Possibly metabolism of ruderal plants is directed towards the formation of active enzyme complexes (protein compounds) and on the synthesis of terminal carbohydrate products. Ruderal species usually have a short life cycle; their metabolism is oriented to the synthesis of the enzymes determining their high functional activity. Lambers and Poorter (1992) demonstrated that low carbon content and high concentration of mineral substances are characterized for fast-grow species. The high functional activity of R-plants allows them to occupy quickly free places in overgrazed pastures. Ruderals, mostly invasive species in overgrazing pastures in Mongolia, such as Artemisia scoparia and Peganum nigellastrum characterized by the rapid development and formation of generative organs, as well as high seed production. In disturbed ecosystems these species the most effectively use photosynthetic products for formation of seeds and the less spend for underground organs (table 2).
On the contrary, week-digressive stress-tolerant species, a majority of photosynthetic assimilated carbon remains in leaves, and the fraction of this carbon is probably utilized for synthesis of secondary compounds. These plants also characterized by large energy expenses for metabolism of protective compounds that enhance a plant tolerance to environmental stress and seasonal weather changes. Stress-tolerant plants (native dominant species of Mongolian ecosystems) are perennials requiring additional amounts of
assimilated carbon to build structures and compounds that increase a plant resistance to environmental factors.
Fig. 3. Box plots distribution of the photosynthetic tissues structure parameters in Mongolian plants with different functional activity: native dominant species with low digressive activity (N, n=12), species with high digressive activity (D, n=4), ruderal plants (R, n=4). The species in the groups are presented in table 5 according Gunin, Vostokova (1989). The means on the graph see fig. 1.
However we cannot associate groups on the digressive activity with the primary ecological strategy of Grime (1979) because plant species studied grow in conditions of strong arid stress and all of them have well expressed properties of stress-tolerators in biomass allocation, mesophyll structure, and leaf chemical composition (table 2). Because of this we disposed Gobi plants in Grime's triangle in stress-tolerators area (fig. 3). This disposition has a good conformation with the data on morphological, structural and biochemical parameters of plants with different ecological strategies in general table 7. We can consider the projection of N- species by Gunin, Vostokova (1989) on S-strategists according to Grime (1979), the projection of D-species on CS-strategists, and the projection of ^-species on RS-strategists of Grime's system. Madon, Medail (1997) also did not find out "clear" R-strategists by Grime in Mediterranean limestone grasslands which climate characterized by an extreme summer drought and noted the well-expressed stress-tolerators properties in annuals of xeric habitats (Madon, Medail, 1997).
Fernandez-Gimenez, Allen-Diaz (2000) also did not observe consistent vegetation changes attributable to the grazing gradient in desert-steppe. There were no ruderal species by a grazing increasing and the grazing had a little influence on shrubs and grasses abundance. On our opinion, it does not contradict to basic natural laws of grazing influence on vegetation but opposite is consistent with the ecological strategies concept. The enhancing of arid stress that corresponds to desertification processes does not let to develop ability of species with the well-expressed R-strategy. Only species with the S-strategy prevailing are able to survive in deserts and desert-steppe conditions as these authors observed Caragana shrubs, salt-shrub species Ananbasis brevifolia, Reaumuria soongorica, Salsola passerina, grasses Stipa gobica/glareosa, blueflags Iris bungei, I. tenuifolia. All of these species belong to S- and CS-strategies according Grime and an increasing of pasture grazing in deserts conditions in our opinion can influence on vegetation change due to the enhancing of CS-strategists (i.e. D-species) proportion. D-species get an advantage by the high level of pasture digression, but not at overgrazing stage (Gunin, Vostokova, 1989). A reduction of grazing pressure leads to the decreasing of D-species (CS-strategists) portion and increasing of N-species portion (S-strategist). Grime (1979) also showed that a reduction of pasture grazing let to the enhancing number of stress-tolerant species.
An absence of grazing during a long period (more than 11 years) has a non-favorable impact on soil properties and plant species diversity (Nechaeva, 1979). In this case occur a hardening of upper layers of soil and spread out of cespitose grasses which prevent expansions of shrubs and semishrubs and favor the reducing of these biomorphs' yield (Nechaeva, 1979). Thus an absence of the grazing further to the increasing of ecological stress and result in the enhancing of stress-tolerant species number in community.
The offered concept of functional activity determination for Gobi plants on their digressive activity is coordinated with the opinion about non-linearity of plant response to grazing intensity (Fensham et al. 1999; Vesk, Westoby, 2001). The plant response along disturbance gradient is consistent with unimodal peaks of abundance along physical environmental gradients (Fensham et al. 1999).
Thus we concluded that the classification of Gobi plants on the ecological strategy types is consistent
with the plant adaptation to both the ecological stress and disturbance. In connection with this the ecological strategy reflects a functional and digressive activity of species. The identification of ecological strategy type leads us to determine a functional and digressive activity of species. In our opinion ecological strategies for Gobi plants correspond to their functional types. An ecological strategy may be identified with the complex of different approaches at levels of whole plant (biomass allocation indices), leaf anatomy (development of inner assimilation surface), and biochemical composition of plant tissue (C, N content, and construction cost). The objectivity and quantitative character of these parameters allow detecting the functional activity and state of each studied species in succession rows revealing the degrees of its ruderal ability and tolerance to ecological stress. A percentage of species with some functional activity in plant communities will be able to show the level of digressive disturbance of the ecosystem.
The results obtained let to prognosticate also the reaction of plant species and predicate vegetation changes under increase of anthropogeneous pressure or desertification processes in Gobi ecosystems.
We are grateful to collaborators of Joint Soviet-Mongolian Integrated Biological Expedition for their assistance in the providing of investigations on Mongolian territory.
This study was supported by the foundation of EU "INCO-Copernicus-2" (ICA2-1999-10110) and by the State Programs "Russian University" (07.01.045).
REFERENCES
1. Box E.O. Plant functional types and climate at the Global Scale. Journal of Vegetation Science. 1996. 7. pp. 309-320.
2. Bullock J.M., Franklin J., Stevenson M., Silvertown J., Coulson S. J., Gregory S.J., Tofts R. A plant trait analysis of responses to grazing in a long-term experiment. Journal of Applied Ecology. 2001. 38: 253267.
3. Diaz S., Noy-Meir I., Cabido M. Can grazing response of herbaceous plants be predicted from simple vegetative traits? Journal of Applied Ecology. 2001. 38. pp. 497-508.
4. Fensham R.J., Holman J.E., Cox M.J. Plant species responses along a grazing disturbance gradient in Australian grassland. Journal of Vegetation Science. 1999. 10. pp. 77-86.
5. Fernandez-Gimenez M., Allen-Diaz B. Vegetation change along gradients from water sources in three grazed Mongolian ecosystems. Plant Ecology. 2000. 00. pp. 1-18.
6. Gamalei Ju.V. Structural and functional evolution of deserts plants. In book: Deserts of Zaaltajskaja Gobi: Characteristic of Dominant Plants. Leningrad, Nauka, 1988. P. 104-106.
7. Garnier E. Growth analysis of congerenic annual and perrenial grass species. Journal of Ecology. 1992. 80. P. 665-675.
8. Garnier E., Salager J.-L., Laurent G., Sonie L. Relationships between photosynthesis, nitrogen and leaf structure in 14 grass species and their dependence on the basis of expression. New Phytologist. 1999. 143. P. 119-129.
9. Grime J.P. Plant Strategies and Vegetation Processes. Chichester: Wiley and Sons, 1979. 222 р.
10. Grime J.P., Hodson J.G., Hunt R. Comparative Plant Ecology: A Functional Approach to Common British Species. London: Unwin Hyman, 1988. 742 р.
11. Gunin P.D., Vostokova E.A. Methodical recommendations on estimation and vegetation mapping of modern state of ecosystems of Mongolia. Ulan-Bator, Joint Soviet-Mongolian Integrated Biological Expedition. 1989. 108 p.
12. Ivanova L.A., Pyankov V.I. Structural adaptation of the leaf mesophyll to shading. Russian Journal of Plant Physiology. 2002. 49. P. 419-431.
13. Lambers H., Poorter H. Inherent Variation in Growth Rate between Higher Plants: A Search for Physiological Causes and Ecological Consequences. Adv. Ecol. Res. 1992. 23. P. 188-261.
14. Lavorel S., McIntyre S., Landsberg J., Forbes TDA. Plant functional classifications: from general groups to specific groups based on response to disturbance. Trends in Ecology & Evolution. 1997. 12. P. 474-478.
15. Madon O., Medail F. The ecological significance of annuals on a Meditteranean grassland (Mt Ventoux, France). Plant Ecology. 1997. 129.P. 189-199.
16. McIntyre S., Lavorel S. How environmental and disturbance factors influence species composition in temperate Australian grasslands. Journal of Vegetation Science. 1994. 5.P. 373-384.
17. Mokronosov A.T. Mesostructure and functional activity of the photosynthetic apparatus. In book: Mesostructure and functional activity of the photosynthetic apparatus. Mokronosov A.T. Ed., Sverdlovsk:
Ural Gos. Univ. 1978. P. 5-15.
18. Nechaeva N.T. Vegetation Productivity Dinamics. Productivity of vegetation of Central Karakums in connection with different use mode. M., Nauka. 1979. P. 54-90.
19. Poorter H. Construction Costs and Payback Time of Biomass: A Whole Plant Perspective. In: A Whole Plant Perspective on Carbon-Nitrogen Interaction. Roy J. and Garnier E. (eds.), The Hague: SPB. 1994.pp.111-127.
20. Pyankov V.I., Mokronosov A.T. Basic tendencies in changes of the Earth's vegetation in relation to global warming of the climate. Russian J. Plant Physiology. 1993. 40 (4). P. 451-466.
21. Pyankov V.I., Ivanova L.A., Lambers H. Quantitative anatomy of photosynthetic tissues of plant species of different functional types in a boreal vegetation. In: Inherent variation in plant growth. Physiological mechanisms and ecological consequences. Lambers H., Poorter H., Van Vooren M.M.I. (eds.), Leiden: Backhuys Publishers. 1998. pp. 71-87.
22. Pyankov V.I., Ivanov L.A. Biomass structure in boreal plants with different types of ecological strategies. Russian J. Ecology. 2000. 31 (1). P. 1-7.
23. Pyankov V.I., Ivanov L.A., Lambers H. Chemical composition of the leaves with different ecological strategies from the boreal zone. Russian J. Ecology. 2001a. 32. P. 221-229.
24. Pyankov V.I., Ivanov L.A., Lambers H. Plant construction cost in the boreal species differing in their ecological strategies. Russian J. Plant Physiology. 2001b. 48: 67-73.
25. Ramenskii L.G. On Principles, Main Concepts, and Terms in Industrial Land Typology, Geobotany, and Ecology. Soviet Botany. 1935. no.4. pp. 25-42
26. Slemnev N.N. Photosynthetic Characteristics of mongolian plants: evolutionary, ecological, and phytocenological aspects. Russian J. Plant Physiology. 1996. 43. P. 366-382.
27. Tilman D., Knops J. The influence of functiona1 diversity and composition on ecosystem processes. Science. 1997. Issue 5330. 277. P. 1300-1302.
28. Vasilevskaja V.K. Leaf formation of drought resistance plants. Ashhabad: AN TSSR. 1954. 126 p.
29. VerstraeteM.M., Schwartz S.A. Desertification and global change. Vegetatio. 1991. 91. P. 3-13.
30. Vesk P.A., Westoby M. Predicting plant species' responses to grazing. Journal of Applied Ecology. 2001. 38. P. 897-909.
31. Voronin P.Yu., Ivanova L.A., Ronzhina D.A., Ivanov L.A., Anenkhonov O.A., Black C.C., Gunin P.D., Pyankov V.I. Structural and Functional Changes in the Leaves of Plants from Steppe Communities as Affected by Aridization of the Eurasian Climate. Russian J. Plant Physiology. 2003. 50 (5). P. 604-601.
32. Vosnesenskii V.L. Morphologically-physiological adaptation of plants to arid conditions. Biosphere Resources of deserts of Middle Asia and Kazachstan. M.: Nauka. 1984. 154 p.
33. Westoby M. A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and soil. 1998. 199. P. 213-227.
