The International Journal of Plant Reproductive Biology 3(2) pp. 133–140, 2011

Correlation Between Sexual Reproduction in Phragmites australis and Die-back Syndrome

1Lara Reale , Daniela Gigante, Flavia Landucci, Roberto Venanzoni, Francesco FerrantiDepartment of Applied Biology, Borgo XX Giugno 74, 06121 Perugia, Italy1e-mail: citolabo@unipg.it

Received: 02.05.2011; Accepted: 12.06.2011

ABSTRACT

The common reed Phragmites australis (Cav.) Trin. ex Steud. (syn. P. communis Trin.) is a member of Poaceae with a widespread distribution in both the hemispheres, in different habitat types from river/lake shores, wetlands, coastlands and estuarine habitats, to ruderal, disturbed and even urban areas, hence it is considered a sub-cosmopolite species (Bjork 1967, Haslam 1973, Ostendorp 1993, Brix 1999a) In the last decades, reeds are dying back at a fast rate in sizeable areas of Europe, with significant impacts on important wetland functions (biodiversity, stability of river and lake margins, water quality) and local economy (Den Hartog et al. 1989, Van Der Putten 1997, Brix 1999b). Similar symptoms have been detected even in central Italy (Gigante et al. 2011). Besides ecological, morphological or anatomical parameters studies concerning some reproductive aspects might also be interesting in order to detect the health condition of reed-dominated ecosystems. In all the plant species the flower biology can be a good indicator of the health condition as the seed production and the seed viability. In plants exhibiting both vegetative and sexual reproduction, the energy allocation can be shifted from one to the other strategy in response to environmental stress; it is well known that sexual reproduction decreases the vulnerability of a population to disturbances and biotic stresses by increasing the genetic variability (Lei 2010).

On this ground, we took into account the reproductive features of a declining reed stand in central Italy, where the die-back syndrome was recently detected (Gigante et al. 2011), with the aims to extend knowledge on sexual reproduction in P. australis and to highlight links between sexual reproduction and die-back symptoms. In this frame, cyto-histological analyses of inflorescences at different developmental stages were carried out in permanent plots where morphological investigation and histochemical analyses were also carried out to verify pollen and seed viability. An interesting link between some decline symptoms and the rate of spikelets containing at least one viable seed was observed. In detail, it appears that higher rates of viable seeds are recorded in the declining reed stands.

Keywords: Phragmites australis, die back, flower development, pollen and seed viability

.

INTRODUCTION

The common reed Phragmites australis (Cav.) Trin. ex Steud. (syn. P. communis Trin.) is a member of Poaceae with a widespread distribution in both the hemispheres, in different habitat types from river/lake shores, wetlands, coastlands and estuarine habitats, to ruderal, disturbed and even urban areas, hence it is considered a subcosmopolite species (Bjork 1967, Haslam 1973, Ostendorp 1993, Brix 1999a). The reed has annual cane-like shoots, reaching heights of between 80 and 350 (1000) cm, usually simple, and an extensive perennial rhizome system (Tutin et al. 1980).

The leaves are arranged alternately and attached to the stem by a smooth sheath, bearing prominent wing-like extensions at the leaf base, with a fringe of fine hairs next to the stem. Flowers, usually purple in colour, are borne on branched inflorescences 20-60 cm in length that is composed of spikelets. Each spikelet, 10-15 mm in length, consists of a short axis, the rachilla, bearing two overlapping bracts, commonly called glumes; the lowermost glumes bear no flowers in their axis and are called empty glumes. Above the empty glumes there are from 3 to 5 flowers subtended by two bracts: the lower or abaxial called “lemma” and the upper or adaxial called “palea”. The small branches between the flowers bear long white silky hairs. The reproductive part of

flowers consists of three stamens with thread like filaments and rather large anthers and a single unilocular pistil with a short style and two feathery stigmas (Tonzig 1956). The fruit of P. australis is a caryopsis with an albuminous seed with starchy endosperm and peripheral embryo. P. australis has the potential to invade new areas by either sexual reproduction through germination of viable seeds or vegetative reproduction via rhizomes.

In Europe, the flowering occurs from late August to early September (Haslam 1972, Van der Toorn 1972), although there is some variation due to genetic and environmental factors (Bjork 1967). The weather during the month of pollination and seed maturation is also important for the seed production.

Until recently, it was thought that P. australis produced few viable seeds (Gervais et al. 1993, Haslam 1972, Tucker 1990) or that seed viability was highly variable (Kettenring & Whigham 2009, McKee & Richards 1996) and that reed typically spread via rhizomes (Bart et al. 2006). Through this method of asexual reproduction, a patch grows by extending rhizomes and sending up new ramets to produce large patches of genetically identical plants. Monoclonal patches have been found within its introduced range, suggesting that reproduction by rhizomes is a significant contributor of P. australis spread (Pellegrin & Hauber 1999). However, a recent study of genetic diversity in reed patches in the Rhode River subestuary indicated that nearly all the sampled patches consisted of multiple genotypes and many patches were genetically distinct from others (i.e., few shared genotypes; McCormick et al. 2010), suggesting a more significant role of seeds and sexual reproduction in P. australis spread than previously suggested. Consistent with these results, Kettenring & Whigham (2009) found that in six sub-estuaries of the Chesapeake Bay, many but not all, reed patches produced abundant viable seeds. These two studies (Kettenring & Whigham 2009, McCormick et al. 2010) underline the importance of understanding the role of seeds in P. australis invasion to complement recent research on reed clonal spread (Amsberry et al. 2000, Bart & Hartman 2003, League et al. 2006).

The seed setting potential of reed seems to be influenced by different factors; several Authors found that the seed setting potential of northern populations of P. australis is lower than at lower latitudes (Bjork 1967, Nikolajevsky 1971). Whilst this might be true in general terms, McKee & Richards (1996) showed that there are many other influences on seedsetting ability in P. australis in Great Britain. Measured seed production was not exclusively high in the South of the country and low in the North, similar to the situation in Quebec (Gervais et al.1993), and can differ dramatically between years, as shown for Swedish populations by Gustafsson & Simak (1963). Weather and resource factors as well as genetic factors must all be

considered, as suggested by Bjork (1967) and Gustafsson & Simak (1963).

In the last decades, reeds are dying back at a fast rate in sizeable areas of Europe, with significant impacts on important wetland functions (biodiversity, stability of river and lake margins, water quality) and local economy (Den Hartog et al. 1989, Van Der Putten 1997, Brix 1999b).

The stability of reed-dominated ecosystems depends on a balance between the processes of progression and retreat. In large areas of Europe, however, this balance has shifted towards a process of die-back and retreat (Den Hartog et al. 1989). Since the balance between open water and reed vegetation is a key parameter for the ecological quality of many European wetlands, the reed dieback has large consequences for important wetland functions.

Ecological, morphological or anatomical parameters were used in order to detect the health condition of reed-dominated ecosystems, but studies concerning some reproductive aspects might also be interesting. In all the plant species the flower biology can be a good indicator of the health condition as the seed production and the seed viability. In plants exhibiting both vegetative and sexual reproduction, the energy allocation can be shifted from one to the other strategy in response to environmental stress; it is well known that sexual reproduction decreases the vulnerability of a population to disturbances and biotic stresses by increasing the genetic variability (Lei 2010). For instance, in Artemisia tridentata shrubs shift their energy allocation to sexual reproduction in response to environmental stress while other plant species, such as Fragaria virginiana, exhibit vegetative propagation in open habitats and sexual reproduction as space becomes more limited (Lei 2010). In P. australis, vegetative analysis of contaminated and uncontaminated wetlands allowed to identify significant differences in the growth patterns which showed an evident reduction in shoot height and seed development (Batty & Younger 2004).

On this ground, we took into account the reproductive features of a declining reed stand in central Italy, where the die-back syndrome was recently detected (Gigante et al. 2011), with the aim both to extend knowledge on sexual reproduction in P. australis and to highlight links between sexual reproduction and die-back symptoms.

MATERIALS & METHODS

Site details — The Lake Trasimeno is the largest lake in 2peninsular Italy, with an average size of ca. 121.5 km and an

average depth of 4.2 m (Dragoni & Evangelisti 1999, Burzigotti et al. 2003). The period of water replacement is 24.4 years, the rate of evaporation is 155x106 m3/year (Viotti et al. 2002, De Bartolomeo et al. 2004). The lake has no natural tributaries and the water supply is largely

134 The International Journal of Plant Reproductive Biology July, 3(2)

Site Plot code Geographical coordinates Type of submersion Presence of litter

OAS3 N43° 05.694’ E12° 11.044’ permanent abundantOasi La Valle

OAS5 N43° 05.681’ E12° 10.946’ permanent abundant

Porto di Panicarola POR3 N43° 04.848’ E12° 06.523’ temporary absent

Rio Pescia POM1 N43° 06.752’ E12° 02.943’ temporary absent

Passignano PAS1 N43° 11.412’ E12° 06.491’ temporary absent

Table 1 — Names, codes, geographical coordinates and main environmental features of sites and plots.

dependent on rainfall, even after the drainage basin was artificially expanded, between 1959 and 1961 (Gambini 1980).

According to Gigante et al. (2011), the South-Eastern sector of the Lake, represented by a large bay with an extremely shallow water column, is affected by die-back and, during the last half-century, suffered a remarkable retreat which led to the loss of a large amount of its surface, especially in the area La Valle. Starting from this background, the experimental design of the present study

2took into account 5 permanent plots (1x1 m ) which, as indicated in Gigante et al. (cit.), belonged to 2 distinct conditions on the ground of a number of macro-morphological plant traits: OAS3, OAS 5 were seemingly affected by die-back (dying-back group: DB), while PAS1, POM1, POR3 didn’t show any macroscopic sign of decline (optimal group: OP).

Names, codes, geographical coordinates and main environmental features of sites and plots are reported in Table 1.

2011 Sexual reproduction in P. australis and die-back — REALE et al 135

Macromorphologic traits — In order to confirm the dying-back condition, some macromorphological analysis were repeated at the end of August 2009 in the 2 groups of plots by adopting the same techniques described in Gigante et al. (2011). Among the most significant symptoms related to die-back, the stem diameter was selected; this trait was measured by using a calibre, on 15 individual culms randomly selected in each of the 5 plots. We checked also the occurrence of clumped habit, a typical symptom of reed die-back, by counting in each of the 5 plots the number of stems involved in the clumps; this number has been reported as percentage with reference to the stem density, which we also

2measured as the total number of stems per plot (n/m ). Finally, we took into account the occurrence of flowering stems; also in this case, the counted number of flowering stems per plot has been reported as percentage with reference to the stem density. To ascertain the actual occurrence of a statistically significant difference between the two groups of plots (DB and OP), the stem diameter was tested by applying ANOVA with a post hoc analysis based on Fisher’s Protected LSD test, while the presence of clumped habit and flowering

were checked by the Student T-test; for the statistics, we used the software Analyst Soft Stat Plus: mac v2009.

Light microscopy — Cyto-histological observations were carried out on flowers collected four weeks before and during anthesis.

To obtain semi-thin sections, the flowers were fixed in 3% (w/v) glutaraldehyde in 0.075 M phosphate buffer, pH 7.2, for 5h. The samples were then washed four times for 15 min each in 0.075 M phosphate buffer, pH 7.2 and post-fixed in 1% (w/v) OsO4. At this stage, samples were dehydratedin increasing concentrations of ethanol and then includedin resin (Epon, 2-dodecenylsuccinic anhydride, and methylnadic anhydride mixture). The semi-thin sections (1–2 µm) were cut with an ultramicrotome (OmU2, Reichert, Heidelberg) equipped with a glass blade, stained with toluidine blue and observed under a light microscope (DMLB, Leica, Wetzlar, Germany)

Pollen viability — To check pollen viability, the flowers were collected during anthesis and their pollen was labelled with fluorescein diacetate (FDA) according to Heslop-

Harrison and Heslop-Harrison (1970) and observed using epifluorescence microscopy.

Seed setting and Seed viability — Using a binocular microscope (x20), for each plot, 100 randomly selected spikelets were collected and for each spikelet, all set seeds were removed, dried and placed in clean Petri dishes lined with fresh filter paper. 100 randomly selected seeds were classified as viable or not viable by labelling with TTC (Waes & Bebergh 1986) and the percentage of spikelets containing at last one viable seed for each plot was reported.

The occurrence of a statistically significant difference between the two groups of plots (DB and OP) with regard to pollen and seed viability was checked by applying the Student T-test (Analyst Soft Stat Plus: mac v2009).

OBSERVATIONS

Before considering the reproductive features, as a first

step we wish to check the condition of decline in the

considered sites. With this aim, the analysed diagnostic traits

are represented as box-and-whisker charts where median,

Fig. 1 — Values of the stem diameter (mm) measured in the 2 groups of plots (n = 15 per plot); median, average, maximum and minimum values, 1st and 3rd quartiles, mild and extreme outliers are reported. DB: dying-back group (OAS3, OAS5); OP: optimal group (PAS1, POM1, POR3). The significance level is indicated (ANOVA, Fisher’s Protected LSD test).

Fig. 2 — Values of the number of culms affected by clumped habit measured in the 2 groups of plots (expressed with reference to the stem density); median, average, maximum and minimum values, 1st and 3rd quartiles, mild and extreme outliers are reported. DB: dying-back group (OAS3, OAS5); OP: optimal group (PAS1, POM1, POR3). The significance level is indicated (Student T-test).

Fig. 3 — Values of the flowering rate measured in the 2 groups of plots; median, average, maximum and minimum values, 1st and 3rd quartiles, mild and extreme outliers are reported. DB: dying-back group (OAS3, OAS5); OP: optimal group (PAS1, POM1, POR3). There are no statistically significant differences (Student T-test).

136 The International Journal of Plant Reproductive Biology July, 3(2)

2011 Sexual reproduction in P. australis and die-back — REALE et al 137

Fig. 4A & B — Light microscopic photographs of flowers 4 weeks before anthesis. A. Longitudinal section of young spikelet consisted of 3-5 flowers with the flower meristem still active, bar 500 µm. B. Detail of young flower in which integument (t) were not completely fused and the archesporial cell (m) that will differentiate in the megaspore mother cell was evident. In the anther the exothecium and the fribous endotecium were differentiated and archesporial cells (mi) that will differentiate in the microspore mother cell were distinguishable, Bar 250 µm.

Fig. 5 – Percentage of pollen viability (mean ± standard error) in the analyzed plots. The y-axes show the percentage (%) of viable pollen, while the x-axes show the plots. The percentage of pollen viability was checked, at the anthesis, with fluorescein diacetate (FDA); fluorescein-labeled pollen (reflecting viability) appeared light green, while not viable pollen appeared bull green.

Fig. 6 – Percentage of spikelets containing at last one viable seed (mean ± standard error) and statistical significance of the 2 groups of plots according to the T-test (P < 0.01). Seed viability was checked by TTC test, the embryo of the viable seeds (left in the picture) appeared pink, while the embryo of the not viable seeds (right in the picture) appeared white.

138 The International Journal of Plant Reproductive Biology July, 3(2)

Fig. 7 — Rate of spikelets containing at least one viable seed (expressed as percentage of the total analysed seeds per plot) and average stem diameter per plot (mm); the regression line is reported.

average, maximum and minimum values, 1st and 3rd

quartiles, mild and extreme outliers are reported for each trait

and group of plots (Figs. 1 & 2). The statistical significance

for each trait, with regards to the 2 groups respectively

referred to dying-back and healthy conditions, is indicated.

Considerably smaller values for the stem diameters (Fig. 1)

have been measured in the plots affected by reed die-back

(OAS3, OAS5). It is remarkable that the clumped habit was

detected only in the declining plots (Fig. 2), while it results

totally absent in the healthy stands (PAS1, POM1, POR3).

On this ground, we could confirm the presence of die-back in

the same areas indicated by Gigante et al. (2011).

As concerns the flowering stems, even if lower values

were recorded in the dying-back plots, there is no significant

difference among the 2 groups of stands (Fig. 3).

As concerns the reproductive features, as general

outcome, four weeks before anthesis, spikelet consisted of 3-

5 flowers but the flower meristem was still active and new

flowers differentiating (Fig. 4A). The pistil was composed of

a superior, tricarpellate and unilocular ovary and stigmas

were plumose; three stamens with thread like filaments were

also observed. The basal flower was the most differentiated

and it was possible to identify the archesporial cells in the

ovule and in the anthers (Fig. 4B). At this stage in the ovule,

which in Phragmites was anatropous, bitegmic and

tenuinucellate, the integument were not completely fused

and the archesporial cell that will differentiate in the

megaspore mother cell was evident (Fig. 4B). In the anther

the exothecium and the fibrous endothecium were

differentiated while the tapetum and the transition layers

were not distinguishable (Fig. 4B).

At the time of anthesis, the pollen grains were uniporate-

operculate and three-celled at the dispersal stage and the

embryo sac was Polygonum type. The pollen viability was

checked, at the anthesis, with fluorescein diacetate (FDA).

The percentage of fluorescein-labeled pollen (reflecting

viability) was lower than that reported in bibliography (Ishii

& Kadono 2002); in OAS3, OAS5, POM1 and POR3 the

values did not differ significantly (Fig. 5). In PAS1 the

highest percentage of pollen viability was observed but this

value only differed significantly from OAS3 and POM1.

Eight weeks after anthesis the differentiation of

caryopsis was completed and the seed setting was

investigated. Spikelets were collected and the percentage of

spikelets producing at least one viable seed was determined.

The highest values were observed in the group of declining

plots (OAS3, OAS5) (Fig. 6).

In order to point out the relation existing between die-

back and seed viability, we took into account the

morphologic traits for each plot and found an interesting link

between the average stem diameters and the rate of spikelets

containing at least one viable seed. In detail, it appeared that

higher rates of viable seeds are recorded in the declining reed

stands (Fig. 7).

DISCUSSION

P. australis is one of the most frequently studied plants as it is an almost cosmopolite taxon, playing a significant role in many wetland ecosystems. In the last decades reeds are dying back in many areas of Europe with remarkable impacts for biodiversity and local economy; therefore many studies were focused on understanding the causes of the die-back phenomenon and identifying its symptoms. Our study pointed out that beside morphological, ecological and anatomical parameters, generally considered by most of the scientists, also reproductive traits might be very interesting when regarded as useful traits to detect the reed die-back. The seed setting was very different in reed stands affected by die-back compared to plots lacking in macroscopic sign of decline; the rate of spikelets containing at last one viable seeds was significantly higher in the dying-back group of plots than in the optimal group. The condition of decline in the considered sites was detected by comparing traits like the stem diameter and the clumped habit.

The seed viability can be influenced by different factors as maternal genotypes (Karkkainen et al. 1999), resource limitation (Suzuki 2000), inbreeding depression (Guerra Araujo 2007) and, during the pre-fertilization, by the limitation of pollen grains (Tandon et al. 2001), self-incompatibility and gametic sterility (Pandit & Babu 2000).

s

The cyto-histological analysis of flower differentiation, the pollen viability and the detection of the flowering culms that we carried out in our sites showed no differences among plots referred to dying-back and healthy conditions. The flower differentiation and the pollen viability don’t seem, therefore, influenced by the die-back conditions but only by the seed differentiation. Considering that the rate of flowering stems doesn’t statistically differ in the two groups of plots, the seed viability can be considered a good indicator trait of the presence of die-back

The relation between the seed viability and the presence of die-back confirmed the observations carried out in other perennial plant species, exhibiting both vegetative and sexual reproduction, which can prefer one or the other way depending on growth condition. Artemisia tridentata grows vegetatively when environmental conditions are optimal, shifting to sexual reproduction when conditions are unfavourable (Evans et al. 1991). Generally the sexual reproduction is preferred in unfavourable growth condition, also because it promotes dispersal ability of seeds into new areas, as well as decreases the vulnerability of a population to disturbances, biotic stresses, and environmental upheaval by increasing genetic variability. Although reed, similarly to other rhizomatous plants, can be propagated both by sexual and vegetative ways, development from seeds is much rarer than vegetative colonization (Haslam 1973, Mauchamp et al. 2001, Nikolajevskij 1971), in particular in an established stand (Koppitz 1999). In fact, germination is a highly vulnerable period and the optimum habitat conditions for this process differ from the adult plant typical environment (Haslam 1973). Our observations demonstrated however that in unfavorable condition, as die-back, the production of seed increases and these results are consistent with the statement that the possible lack of genetical diversity of Phragmites over large areas, resulting from its clonal habit, might increase its susceptibility to die-back (Clevering 1995, Kohl et al. 1995).

LITERATURE CITED

Bjork S 1967 Ecological investigations of Phragmites communis: studies in theoretic and applied limnology, Folia Limnol Scand 14 1–248.

Brix H 1999a Genetic diversity, ecophysiology and growth dynamics of reed (Phragmites australis), Aquatic Botany 64 179–184.

Brix H 1999b The European research programme on reed die-back and progression (EUREED), Limnologica 29 5–10.

Burzigotti R, Dragoni W, Evangelisti C & Gervasi L 2003 The Role of Lake Trasimeno (central Italy) in the History of Hydrology and Water Management, IWHA 3rd International Conference. Alexandria, Egypt.

Clevering OA 1995 Ideological significance of genetic diversity of Phragmites australis in relation to nutrient status and occurrence of die-back. Dutch Group Report. In: van Putten

.

s

WH, ed. Reed News Reports of the EC Project EUREED – EV5V-CT92-0083, vol 3, Heteren, The Netherlands: Netherlands Institute of Ecology 33-36.

Den Hartog C, Kvet J, & Sukopp H 1989 Reed: A common species in decline, Aquat. Bot. 35 (1) 1-4.

De Bartolomeo A, Poletti L, Sanchini G, Sebastiani B, & Morozzi G 2004 Relationship among parameters of lake polluted sediments evaluated by multivariate statistical analysis, Chemosphere 55 1323-1329.

Dragoni W & Evangelisti C 1999 Il problema del lago Trasimeno, Spazio Ambiente 3 14-17. Perugia.

Evans R D, Black RA, & LINK SO 1991 Reproductive growth during drought in Artemisia tridentate Nutt., Functional Ecology 5 676-683.

Gambini E 1995 Le oscillazioni di livello del lago Trasimeno, Quaderni del Museo della Pesca del Lago Trasimeno, 2 Grafiche Piemme Perugia 139 pp.

Gigante D, Venanzoni R & Zuccarello V 2011 Reed die-back in southern Europe? A case study from Central Italy, C R Biol, 334 (4) 327–336 Elsevier

Guerra Araujo AC, Falcao R, de Campos C & Vera T 2007 Seed abortion in the sexual counterpart of Brachiaria brizantha apomicts (Poaceae), Sex. Plant Reprod. 20 109–121.

Haslam S M 1973 Some aspects of the life history and autecology of Phragmites communis, Polish Archives of Hydrobiology 2079–2100.

Heslop-Harrison J & Heslop-Harrison Y 1970. Evaluation of pollen viability by enzymatically induced fluorescence: intracellular hydrolysis of fluorescein diacetate, Stain Technol. 30 115–120.

Karkkainen K, Savolainen O & Koski V 1999 Why do plants abort so many developing seeds bad offspring or bad maternal genotypes? Evol. Ecol. 13 305–317.

Kohl JG, Henze R, Kuhl H, Koppitz H, Woitke P, &Henze 1995 German Group Report. In: van Putten WH, ed. Reed News Reports of the EC Project EUREED – EV5V-CT92-0083, vol 3, Heteren, The Netherlands: Netherlands Institute of Ecology 25-32.

Koppitz H 1999 Analysis of genetic diversity among selected populations of Phragmites australis world-wide, Aquat. Bot. 64 209–221.

Lei SA 2010 Benefits and costs of vegetative and sexual reproduction in perennial plants: A review of the literature, Journal of the Arizona-Nevada Academy of Science 42 9-14.

Mauchamp A, Blanch S & Grillas P 2001 Effects of submergence on the growth of Phragmites australis seedlings, Aquat. Bot. 69 147–164.

Nikolajevskij VG 1971 Research into the biology of the common reed (Phragmites communis Trin.) in the USSR, Folia Geobot. Phytotaxon. 6 221–230.

Ostendorp W 1993 Reed bed characteristics and significance of reeds in landscape ecology. In: Ostendorp W, Krumscheid-Plankert P, eds. Lakeshore deterioration and restoration Works in Central Europe, 149–161. Fischer-Verlag, Stuttgart, FRG.

Pandit MK & Babu CR 2000 Synaptic mutation associated with gametic sterility and population divergence in Coptis teeta (Ranunculaceae), Bot. J. Linn. Soc. 133 525-533.

Suzuki N 2000 Pollinator limitation and resource limitation of seed production in the Scoth broom, Cytisus scoparius (Leguminosae), Plant Species Biol. 15 187-193.

2011 Sexual reproduction in P. australis and die-back — REALE et al 139

Tandon R, Shivanna KR & Mohan HY 2001 Pollination biology

and breeding system of Acacia Senegal, Bot. J. Linn. Soc. 135

251–262.

Tonzig S 1956 Elementi di Botanica. Casa Editrice Ambrosiana,

Milano.

Tutin T G, Heywood V H, Burges N A, Moore D M, Valentine D H,

Walters S M & Webb D A (eds.) 1980 Flora Europaea 5

Cambridge University Press.

Van der Putten W H 1997 Die-back of Phragmites australis in

European wetlands: an overview of the European Research

Programme on reed die-back and progression (1993-1994),

Aquat. Bot. 59 263–275.

Van Waes JM &Bebergh PC 1986 Adaptation of the tetrazolium method for testing the seed viability, and scanning electron microscopy study of some Western European orchids, Physiol. Plant. 66 435-442.

Viotti P, Galeotti L, Sbaffoni S, Sappa G, Leccese M & Stracqualursi N 2002 Analisi sperimentale dei flussi di fosforo dai sedimenti di un lago: il caso del Lago Trasimeno (Italia), VI Simpósio Ítalo Brasileiro de Engenharia Sanitária e Ambiental.

140 The International Journal of Plant Reproductive Biology July, 3(2)