LIGHT AND TEMPERATURE REQUIREMENTS FOR GERMINATION OF SEEDS OF AECHMEA NUDICAULIS (L.) GRIESEBACH AND STREPTOCALYX FLORIBUNDUS (MARTIUS EX SCHULTES F.) MEZ (BROMELIACEAE)

[Light and temperature requirements for germination of seeds of Aechmea nudicaulis (L.) Griesebach and Streptocalyx floribundus (Martius ex Schultes F.) Mez (Bromeliaceae)]. Seeds of A. nudicaulis and S. floribundus were collected in a restinga located at Presidente Kennedy (ES, Brazil). This sandy community is characterized by the presence of patches of vegetation interspersed by open areas with a sparse plant covering. In order to evaluate the capacity of A. nudicaulis and S. floribundus to recruit individuals from seeds in the distinct microclimates generated by this heterogeneous plant covering, experiments were conducted in laboratory to test the effects of light and temperature on the germination for both species. Seeds of A. nudicaulis were unable to germinate in darkness but they achieved high germinability (≥90%) from 15 to 40°C under red light. Seeds of S. floribundus germinated in darkness only at 40oC (22.5%), whereas the germinability was high (≥90%) from 15 to 30oC under 1 Departamento de Ecologia, Universidade de Brasília, CEP 70919-970, Brasília, DF, Brasil (pinheiro@unb.br) 2 Corresponding Author: Lab. Termobiologia L.G. Labouriau, Departamento de Botânica, Universidade de Brasília, CEP 70919-970, Brasília, DF, Brasil (fborghet@unb.br) 28 Pinheiro & Borghetti: Light and temperature requirements for germination of seeds light. For both species, the germination rate was highest at 25oC. For these bromeliads, alternating temperature (20/30oC) promoted high germinability (≥92.5%) only under red light. The seeds at 20/30°C in darkness were later transferred to light and a high germinability (>89%) was achieved for both species. No seed germinated at alternating temperatures of 20/50oC and they were dead after 720h of incubation. The seeds showed to be photoblastic and did not tolerate high temperatures. The results suggest that under natural conditions the germination of seeds of A. nudicaulis and S. floribundus is restricted to microclimates where light and moderate temperatures are present. The microsites generated by the plant covering seem to provide these requirements.


Introduction
The sandy coastal plain (Restinga) along the Brazilian coastal line has a great diversity of plant communities which species composition and spatial plant distribution result from topographic peculiarities and distinct environmental conditions (Araújo & Henriques 1984;Henriques et al. 1986).Some of those formations (Clusia and Ericaceae formation) are characterized by the scattered occurrence of patches of vegetation consisting of cacti, bromeliads, shrubs and trees, interspersed by open areas where a sparse vegetation constituted by bromeliads, palms and small shrubs is present (Henriques et al. 1986).High temperatures achieved by soils exposed to full sun (Franco et al. 1984;1996), water stress and oligotrophism (Reinert et al. 1996) are among the limiting factors for the establishment of species in this community.
It has been found that some plants make possible the establishment of other species due to the shade provided by their canopy (Franco & Nobel 1988, 1989, Fulbright et al. 1995;Barnes & Archer 1996).These "nurse plants" reduce the soil desiccation by solar radiation and provide water availability for the associated seedlings.Maximum soil surface temperatures are also reduced by shading, thus furnishing an appropriate microhabitat for seed germination and seedling establishment beneath the canopy (Franco & Nobel 1988;1989).This shading effect could be remarkably important in Restinga communities; Fialho (1990) reported a high mortality among seeds of Erythroxylum ovalifolium that fall on the sandy soil due to extreme temperatures.Indeed, Clusia sp.(Zaluar 1997) and Neoregelia cruenta (Hay & Lacerda 1980) were shown to reduce maximum soil temperatures and to increase soil nutrients and organic matter, respectively.
It is well established that an understanding of vegetation dynamics requires a combination of intensive field studies (phenology, mortality, seed bank) and experimental approaches (Schat & Scholten 1985).Accordingly, several reports have correlated spatial and temporal dynamics of species in the field to dormancy status and germination characteristics of the seeds (Fenner 1980;Meyer et al. 1989;Maunn 1994;Fulbright et al. 1995;Meyer et al. 1995;Schupp 1995).Unfortunately, few studies on seed germination have been conducted for species of Restinga (e.g.Dau & Labouriau 1974;Mercier & Guerreiro Filho 1990;Fialho & Furtado 1993).
The bromeliads Aechmea nudicaulis (L.) Griesebach and Streptocalyx floribundus (Martius ex Schultes F.) Mez are conspicuous components of Restinga communities.Individuals of A. nudicaulis usually occur bordering the patches of vegetation (Dau 1960;Pereira 1990).In the Restinga where the seeds were collected for the present investigation, both species were frequently found at the edges of the thickets, and a few individuals were also found in open areas (F.Borghetti, pers. obs.).In view of these observations, we ask the following: would these species tolerate the high soil temperatures of open areas, thus acting as early colonizers in patch dynamics?Considering the effects of light and temperature on seed germination we discuss these possibilities for A. nudicaulis and S. floribundus.

Material and methods
Aechmea nudicaulis (L.) Griesebach occurs as epiphytic, rupestrine or terrestial species, being described for several ecosystems including the sandy coastal plains of Brazil.Streptocalyx floribundus (Martius ex Schultes F.) Mez occurs as epiphytic or terrestrial species and can be found in coastal rain forests and sandy coastal plains of southeastern Brazil (Fontoura et al. 1991).
The seeds were collected in September 1996 in a Restinga located at "Barra de Itabapuana" (21°16' S and 40º48' W), municipality of Presidente Kennedy, southern coast of Espirito Santo State, southeastern Brazil.Warm and rainy summers and dry winters characterize the climate (Fig. 1).The distribution of the vegetation at that site is characterized by the occurrence of thickets interspersed by areas with a sparse plant cover.Temperature measurements of the soil surface were done daily during a week, using maximumminimum thermometers (MM5201 Incoterm) installed under shaded microhabitats (inside thickets) and in soil exposed to full sunlight.The thermometers were placed a few millimeters underground and covered with sand.The temperatures employed in the experiments described below were based on these field data.
In order to obtain representative seed samples of the local populations, several mature fruits were collected from at least 15 individuals of each species.In the laboratory, the seeds were removed from the fruits, dried at room temperature and stored at 22 ± 2°C in a dark room for about a month until use.Seed viability (n = 100) was estimated using a 0.5% tetrazolium salt solution (Copeland 1976).Seed germination studies were performed using incubators regulated at the following constant temperatures: 15,20,25,30,35,40 a Gaertner L-205 spectroscope (Borghetti 1992).This wavelength range is the most effective in promoting the germination of light-requiring seeds (Bewley & Black 1994).Four replicates of 50 seeds were used for each treatment.The seeds were placed in 9cm Petri dishes with one layer of qualitative filter paper.At 24h intervals the germinated seeds were counted and removed.
The observations were made under red light and deionized water was added as needed.For experiments conducted in darkness, the dishes were placed into black boxes in the beginning of the experiment and the observations were made under a green safelight (490-560nm) (Silveira & Labouriau 1995).The non-germinated seeds incubated at 20/30°C in darkness were transferred to red light after 240 h of incubation.Viability of the non-germinated seeds was evaluated at the end of the experiments using the tetrazolium solution.
Germinability (G = percentage of germinated seeds), the average germination time (t) and its variance (tS 2 ), and the average germination rate (V = inverse of the average germination time) and its variance (vS 2 ) were computed according to Labouriau (1972).The distributions of seed germination at different temperatures were tested for their normality through the Kolmogorov-Smirnov test (Sokal & Rohlf 1995).Comparisons between average germination rates were performed through the Mann-Whitney U-test (Sokal & Rohlf 1995).Confidence intervals for germinabilities were obtained from Tablas Cientificas (Documenta Geigy 1965).All statistics used an a level of 0.05.

Results
The tetrazolium test indicated a seed viability of 100% for both species.Irrespective of the temperature treatment, seeds of A. nudicaulis did not germinate in darkness, but achieved a high germinability from 15 to 40ºC under red light (Tab.1).A few seeds of S. floribundus germinated in darkness at 40°C, but under red light a high germinability was Table 1: Temperature dependence of the germinability, average germination rate and average germination time of seeds of Aechmea nudicaulis (n=200 seeds/treatment).a) The treatments were conducted under red light, except indicated in the table.b) Germinability (G) of pooled isothermal replicates (in percent); Gi and Gs are the lower and the upper confidence limits of the germinabilities, respectively (α = 0.05, Documenta Geigy, 1965).c) Calculated according to Labouriau (1972) (in hours).d) The seeds were incubated at 20/30 o C (16/8 h, respectively) under continuous red light (Light) and in darkness.Because no seed germinated in darkness, they were exposed to red light treatment after 240 hours of incubation (Dark / light).e) Kolmogorov-Smirnov test for the normality of the germination curves.* Differs from a normal distribution; (Sokal & Rohlf, 1995) observed from 15 to 30ºC (Tab.2).
For both species the temporal distribution of the germination differed from a gaussian distribution in most incubation temperatures (Kolmogorov-Smirnov test, Tab. 1 and 2).The large variances of the average germination times at 15 o C and above 35 o C indicate for both bromeliads that the germination can prolong for a week or more at extreme temperatures (Tab. 1 and 2).The Mann-Whitney U-test pointed out the interval from 20 to 30ºC as the optimum temperature range for germination of A. nudicaulis, and 25°C as the optimum temperature for germination of S. floribundus (Tab.3).
Despite the low germinability at 45°C (Tab.2), seeds of S. floribundus were still viable after 720 hours of incubation under red light and after 360 hours in darkness (tetrazolium test).When incubated at alternating temperatures of 20/30°C seeds of both species showed a high germinability only under red light (Tab. 1 and 2).The seeds in darkness were transferred to red light after 240 hours of incubation and a high germinability resulted for both species (Tab. 1 and 2; Fig. 2).For both species no germination was detected at 20/50°C and the seeds were not viable after 720 hours of incubation (tetrazolium test pointed out zero percent of viability).
Temperature measurements at the study site showed that the minimum average temperature at the soil surface under plant cover (21.5°C ± 0.8) was similar to that observed in sunny areas (21.0°C ± 0.9).On the other hand, the maximum average temperature verified in soil exposed to full sunlight (45.9°C ± 1.2) was higher than that found in the shaded microhabitats (28.3°C ± 4.5).

Discussion
The occurrence of patches of plants interspersed by open areas results in different microclimates in the Restinga.Dau (1960) found that the daily thermal range in sheltered Table 2. Temperature dependence of the germinability, average germination rate and average germination time of seeds of Streptocalyx floribundus (n=200 seeds/treatment).a) The treatments were conducted under light, except indicated in the table.b) Germinability (G) of pooled isothermal replicates (in percent); Gi and Gs are the lower and the upper confidence limits of the germinabilities, respectively (α = 0.05, Documenta Geigy, 1965).c) Calculated according to Labouriau (1972) (in hours).d) The seeds were incubated at 20/30 o C (16/8 h, respectively) under continuous red light (Light) and in darkness.Because no seed germinated in darkness, they were exposed to red light treatment after 240 hours of incubation (Dark / light).e) Kolmogorov-Smirnov test for the normality of the germination curves.* Differs from a normal distribution; NS, normal distribution (Sokal & Rohlf, 1995) et al. (1984) in a restinga located at Barra de Maricá (Rio de Janeiro State).While the minimum average temperature in the soil was similar in sites of different canopy densities (around 25ºC), the maximum average temperature could reach 50ºC or more in soil exposed to full sunlight (Franco et al. 1984).As expected, we found that the maximum soil temperatures were higher in sunny than in shaded microhabitats.An increase in incubation temperatures reduced the germinability for both species.At 35 o C and above the germinability of A. nudicaulis was significantly higher than S. floribundus, but seeds of S. floribundus incubated at 45 o C remained viable for more than Table 3.Multiple binary comparisons of isothermal average germination rates (V) for four samples of 50 seeds of Aechmea nudicaulis and Streptocalyx floribundus, using a two-tailed Mann-Whitney test (Sokal & Rohlf 1995).The symbol * corresponds to U-values larger than the critical value U 0.05 (4,4) = 13; lower U-values (non-significant differences at 0.05) are written in the nomogram.Seeds of A. nudicaulis germinated faster than seeds of S. floribundus in all temperatures tested, but even at the optimum temperature for germination (around 25 o C) both species required at least four days in an imbibed state to germinate.Also, the time variances indicate for these bromeliads that germination extends for more than a week at temperatures of 35 o C and above this.This suggests that under natural conditions a regular water availability is required for a successful germination, what restricts the seed recruitment for both species to sites where soil desiccation is not a recurrent phenomenon, as those occurring under plant cover (Franco & Nobel 1988;1989;Fulbright et al. 1995;Barnes & Archer 1996).

Aechmea nudicaulis
In contrast to bromeliads of shaded habitats as gallery and deciduous forests (Garcia-Franco & Rico-Gray 1991;Rosa & Ferreira 1998), but similar to bromeliads of restinga as Aechmea distinchantha and Neuregelia cruenta (Mercier & Guerreiro Filho 1990), seeds of A. nudicaulis and S. floribundus require light in order to germinate.Bromeliads that produce light-requiring seeds (Mercier & Guerreiro Filho 1990) were found to be shade-intolerant, as the case of N. cruenta and A. nudicaulis (Fischer & Araújo 1995).This suggests that the predominance of individuals of A. nudicaulis on the edges of the thickets could result from the shade-intolerance of the species.
The light and temperature requirements showed by seeds of A. nudicaulis and S. floribundus suggest that germination would be limited to specific microsites where proper temperatures and light intensity were found.The seed susceptibility to high temperatures make these species unlikely to recruit from seeds in soils exposed to full sunlight.However, the seeds would find suitable conditions for germination in sheltered microhabitats.In fact, it has been proposed that the shade produced by the palm Allagoptera arenaria in the bare sand facilitate the germination of cacti and bromeliads as A. nudicaulis and Neoregelia cruenta (Scarano 2000).Subsequently, these species would favor the entry of woody species (as Clusia spp.), in a facilitation process resulting in the development of the thicket (Zaluar & Scarano 2000).
The results showed here suggest that the bromeliads A. nudicaulis and S. floribundus would not act as pioneers in patch dynamics but could participate as important intermediates in thicket formation by generating soil and microclimate conditions for the establishment of less tolerant species, as already addressed for N. cruenta (Hay & Lacerda 1980).Indeed, studies regarding plant growing in the field are necessary for a better understanding of the participation of these species in patch dynamics and to verify whether the bromeliads frequently found in the bare soil result from seed germination or clonal growth.
Figure 1.Average temperature ( o C) and precipitation (mm) recorded in 1996 for the municipality of Vitória (ES).Source: Instituto Nacional de Meteorologia.
Germinability of seeds of Aechmea nudicaulis and Streptocalyx floribundus at 20/30º C. The seeds were incubated under red light (open symbols) and in darkness (black symbols) (n=200 seeds/treatment).After 240 hours of incubation in darkness the non-germinated seeds were exposed to red light.Continuous line -A.nudicaulis; Dashed line -S.floribundus.