Diversity of conidial fungi and some abiotic variables of the water after the reopening of the Pirarungaua stream in the Jardim Botânico , São Paulo , São Paulo State , Brazil

(Diversity of conidial fungi and some abiotic variables of the water after the reopening of the Pirarungaua stream in the Jardim Botânico, São Paulo, São Paulo State, Brazil). To evaluate the fungal diversity in the Pirarungaua stream reopened after remaining channeled about 40 years in the Botanical Garden of São Paulo, Brazil, two experiments in the cold and dry season and two in the hot and rainy season were performed. Leaves of Tibouchina pulchra Cogn and mixed leaf litter were confined in nylon litter bags, submerged and collected monthly, together with free leaf litter samples. Some abiotic factors were measured in the water. The leaves were fragmented, incubated in sterile water at 22 oC and microscopically analyzed. We identified 33 conidial fungal taxa, among which, 23 are new to this area. Comparing seasons, the similarity of the mycota of confined T. pulchra leaves corresponded to 39% and to 56% regarding the confined mixed leaf litter. Similarities of 48-50% occurred between the mycota of confined and free leaf litter samples. Confined leaves showed a tendency to support higher numbers of species, whereas some rare fungi like Ulocoryphus mastigophorus Michaelides L. Hunter & W.B Kendr were more frequent in free leaf litter. Based on the results, it is recommended to monitor the limnological conditions of the stream and to promote awareness among the local population and visitors about the importance of a water body, even artificial, for the establishment of microclimatic features, and for the biodiversity of fungi.


Introduction
According to former research reports of the Botanical Garden in São Paulo, Brazil (Hoehne 1943), the Pirarungaua stream joins with the Córrego dos Simões stream, forming the River Ipiranga, whose importance lies in the historical fact that the independence of the country was declared on its margins in the nineteenth century.
Since the 1940 decade, of the 1380 m in length of the stream, only 500 m were flowing in its natural bed, from the springhead through the exuberant plateau Atlantic rainforest, being the remainder completely channeled (Hoehne 1943).
However, in August 2007 the underground gallery suffered a severe breakdown, which resulted in a hole of 5.00 m wide by 10.00 m long and 2.50 m deep.The underground channel was demolished and in its place, a ditch of 230 m long by 14 m wide was built, with smooth, inclined slopes composing a new profile for the new channel designed to remain in the open and integrated with the landscape.
Once the stream was opened, communities of aquatic plants, algae, mollusks and fishes were settling and thus probably fungi that fulfilling one of its main functions, may have become important decomposers of new sources of organic matter.As practically all major limnological characteristics of the stream were drastically modified and gradually the riparian vegetation has been established, it can be considered that the Pirarungaua is a waterbody in revitalization, constituting a "natural laboratory" (Tundisi & Tundisi 2008) to accompany changes in those communities.
The aquatic mycota of the State Park Fontes do Ipiranga -PEFI, where the Botanical Garden is located, is fairly well known and studies about fungal succession on leaves of Ficus microcarpa L. and Tibouchina pulchra Cogn.(Schoenlein-Crusius & Milanez 1989, Moreira 2006), surveys of aquatic facultatives and ingoldian fungi in mixed litter collected from different water bodies with several levels of eutrophication, in dry, cold and rainy and hot seasons (Schoenlein-Crusius et al. 2009).Several conidial fungi, specifically ingoldian fungi have been observed in lotic and also in lenthic waters in urban parks, with richness and diversity higher than expected for an anthropic environment (Schoenlein-Crusius et al. 2014).Revisions about the main surveys of ingoldian fungi in Brazil and South America may be consulted in Schoenlein-Crusius & Grandi (2003) and Schoenlein-Crusius & Malosso (2007).More recently outstanding conidial species found in very dry regions in the Northeast region of Brazil have been cited and excellently taxonomically described by Barbosa & Gusmão (2011), Almeida et al. (2011;2012), among others.
Zoosporic organisms were also studied in several taxonomic surveys in the Botanical Garden of São Paulo (Milanez et al. 2007, Pires-Zottarelli & Rocha 2007); most recently including free mixed litter collected in the Pirarungaua stream (Jesus et al. 2013).
The present study proposes to add new knowledge about the diversity and richness of conidial fungi, which approaches aquatic facultatives and ingoldian fungi, taking the opportunity to study, during a revitalization process, a stream that is increasingly integrating nature.Some abiotic factors are presented as first measurements in the reopened lotic system.Measurement of some abiotic factors in the water -During the monthly sampling, using Multiprobe equipment U 10 from Horiba, temperature, pH, dissolved oxygen and conductivity were measured at 20 cm of water depth at each collection point.

Characterization
Field experiments -From May/2010 to April/2014, four experiments were performed: the first and the third initiated in the dry and cold season, and the second and forth experiment in the rainy and hot season.To accumulate Tibouchina pulchra Cogn.fallen leaves, during 30 days, avoiding their contact with the ground, two nylon nets (1mm diam.Mesh), with the size of fairly 3 × 5 m, were tied on four trees in the Botanical Garden of São Paulo, at 1 m above the soil surface.The entire area of the extended nets (15 m 2 ), from one edge to the other retained leaves of the canopy of at least 10 trees of T. pulchra in June (dry and cold season) and in November (rainy and hot season) of 2010.
In May (dry and cold season) and in October (rainy and hot season) of 2011, about 10 kg of fallen mixed leaf litter from one site of the Pirarungaua stream was collected.For each experiment the amount of leaves was divided in samples of 10 g each, which were confined in 120 nylon net litter bags (10 × 10 cm with 1 mm diam.mesh) and submerged in the six collection points along the stream.
The leaves of T. pulchra remained submerged from July to November of 2010 (1 st .Experiment, 5 months, dry and cold season) and from December of 2010 to May of 2012 (2 nd .Experiment, 6 months, rainy and hot season).The mixed leaf litter samples remained submerged in the stream from June to November of 2011 (3 rd .Experiment, 6 months, dry and cold season) and from November of 2011 to April of 2012 (4 th .Experiment, 6 months, rainy and hot season).
Monthly, one nylon net litter bag was taken from each collection point.Concomitantly during the 3 rd and 4 th .Experiments, two samples of 10-20 leaves of submerged free mixed leaf litter were taken at 20 to 40 cm deep at each point, totalizing 12 monthly samples from the stream.
Fungal isolations -At the laboratory, the leaves were taken from the nylon net litter bags and from the collection flasks, disposed on a strainer of 120 µm mesh and gently washed with tap water to get rid of adhered debris.
The confined leaves and also the free mixed leaf litter samples were fragmented in 1 cm diam.Aliquots, which were incubated in Petri dishes containing sterile distilled water (Ingold 1975).The number of Petri dishes obtained at this phase depended on the decomposition rate of the leaves, because in each dish the leaf fragments should remain submerged, not overlaying, but free, in a mostly similar way to what would occur in the field.From the 5 th day on of incubation of the Petri dishes at a temperature about 20 o C, slides of the leaf fragments were prepared, using only drops of sterile distilled water and covered with coverslip, to be observed under an optical microscope, in order to look for taxonomical relevant features, which allow the identification of the species according to current literature (Nilsson 1964, Ingold 1975, Marvanová 1997, Santos-Flores & Betancourt-Lopez 1999, Gulis et al. 2005).
Statistical treatment of the data -In order to verify the existence of significant differences between the abiotic factors such as temperature, pH, dissolved oxygen and conductivity of the stream water, a Kruskal-Wallis test was employed.Calculations were made by the PAST Program, version 2.17 (Paleontological statistics software package for education and data analysis) of Hammer et al. (2001).
To compare the mycota observed in the different months and collection points, the Similarity Index of Sörensen (S) was used, expressed by the formula S = 2xc/a + b, where a is the total number of fungal taxa in a given situation, b is the number of fungal taxa in a situation to be compared and c is the number of fungal taxa that are common to the two situations which are being compared.The similarity is expressed in percentual values, multiplying the S value by 100 (Müller-Dombois & Ellenberg 1974).

Results and Discussion
Characterization of climatic conditions during the experiments -According to table 1, the climatic data confirm that the 1 st and 3 rd experiments were performed in periods with lower rainfall, lower number of rainy days and cooler (lower monthly average temperatures) and the 2 nd and 4 th experimentsin more rainy periods (higher values of volumetric rain fall and number of rainy days) and hot (higher rain fall and average monthly temperatures).The relative air humidity remained between 74.5 and 86.9% during the entire study.Some abiotic factors in stream water during the collections -Table 2 shows the individual values, means and table 3 the results of the Kruskal-Wallis test regarding the temperature of stream water measured during the experiments.The Kruskal-Wallis test showed no significant differences between the sampling points, but between collections.As the experiments were performed subsequently, approaching cold and dry seasons and also hot and rainy seasons, the water temperature followed climatic variations, reaching minimum average values in July/10 (19 ºC, 1 st experiment), May/11 (17,8 ºC, 2 nd experiment), June/11 (16,1 ºC, 3 rd experiment) and April/12 (21.2 ºC, 4 th experiment), whereas the maximum mean values were registered in October/10 (1 st experiment), December/10 (24.7 ºC, 2 nd experiment), November/11 (22.9 ºC, 3 rd experiment) and in February and March of 2012 (24.9 ºC, 4 th experiment).The mean values found in this study are similar to those cited by Bicudo et al. (2002) in the Lago das Ninféias, which is the reservoir that supplies the Pirarungaua stream.Also within the range of 26.6 ºC (February 2005) and minimum 12.6 °C (July 2004) cited in previous study conducted in several collection sites in the Parque Estadual das Fontes do Ipiranga (Schoenlein-Crusius et al. 2009).
Table 4 shows the individual values, means, and table 5, the result of the Kruskal-Wallis test related to the pH of the water stream measured during the experiments.
As with temperature, the Kruskal-Wallis test revealed significant differences only between collections.
A tendency of lower values to occur in more rainy and hot seasons, and the contrary in the opposite climatic seasons may be verified.However, in view that not even the highest pH values exceeded 7.0, probably the stream presents the tendency to have more acid waters.This result is consistent with what Bicudo et al. (2002) found in the Lago das Ninféias, where the pH of the water ranged from 5.5 to 6.5 between 1997 and 1998.It is also consistent with the amplitude of the pH of the waters of PEFI, cited by Schoenlein-Crusius et al. (2009).
Table 6 shows the individual values and averages and table 7, the result of the Kruskal-Wallis test regarding the content of dissolved oxygen (mgO 2 L -1 ) of stream water measured during the experiments.In September/10 occurred the maximum mean value -13.5 mgO 2 L -1 and in November/10 the lowest, corresponding to 7.2 mgO 2 L -1 .With exceptions of point 1, which mean value of oxygen corresponded to 5.72 mgO 2 L -1 , the remaining points presented means above 7 mgO 2 L -1 during the 1 st experiment.In the 2 nd experiment, initiated in the rainy and hot season, when according to the literature, the oxygen levels are usually lower (Esteves 1998), the mean values were significantly higher in the 6 th collection made in May of 2011, in relation to other collections.In the 3 rd experiment, started in the dry and cold season, the average values of the levels of dissolved oxygen were higher in July/11 (8.83 mgO 2 L -1 ) and significantly lower in October/11 (6.19 mgO 2 L -1 ; 5 th collection).
In the last experiment, started in the rainy and hot season, the average dissolved oxygen concentration was significantly higher in January/12 (7.54 mgO 2 L -1 -3 rd collection) and significantly lower in April/12 (6.41 mgO 2 L -1 -6 th collection), when a sharp decrease of the rainfall was recorded (table 1).These results are consistent with the aforementioned climatic variations in the literature (Esteves 1998, Tundisi & Tundisi 2008).The variation of the values found during the entire study is larger than the range between 0.4 and 5.6 mgO 2 L -1 cited by Bicudo et al. (2002)  Crusius et al. (2009) for other water bodies located in PEFI.Although the statistical test did not pointed mathematically significant differences, it is notable that the average levels of dissolved oxygen tended to rise (with some exceptions) from the point 1 to 6, probably due to the turbulence of the waters, favored by the artificial falls built along the channel, thus fulfilling its goal to keep the water moving and sufficiently aerated.
Table 8 shows the individual values, means, and table 9, the results of the Kruskal-Wallis test concerning conductivity (µScm -1 ) of the water stream measured during the experiments.According to the Kruskal-Wallis test, there are significant differences only between collections.
The values of the water conductivity of the Pirarungaua stream are in the range of the variation of this parameter cited by Bicudo et al. (2002), Moreira (2006) and Schoenlein-Crusius et al. (2009) for other collection sites in the PEFI.As no abiotic variables measured before the reopening of the stream were found, it seems interesting to present some of them in the present paper for historical purposes.
Diversity of aquatic Hyphomycetes -During the entire study period, 33 taxa of fungi associated with submerged decaying substrates were obtained.Among these, eight (Beltrania rhombica, Beltraniopsis sp., Beltraniella sp., Dictyochaeta sp., Pestalotiopsis-like., Fusarium oxysporum, Pestalotiopsis sp. and Ulocoryphus mastigophorus) are more common in terrestrial environments, been considered 18 times, being the most frequent taxa, followed by Margaritispora aquatica (8 occurrences) and Camposporium pellucidum (6 occurrences).Table 11 presents the fungal taxa obtained during the 2 nd experiment, initiated in the rainy and hot season, with T. pulchra leaves too.In relation to the1 st experiment, the number of taxa was slightly higher, corresponding to 17.The number of occurrences ranged from 13 in 5 th collection, to only one recorded occurrence in the last collection.Point 2 in this experimental presented the highest number of taxa (8), while point 1 had the highest number of occurrences ( 13).The number of taxa remained the same during all collections, except for the last, when only one taxa was isolated.However, the similarity index of Sörensen between the collections remained around 50%, demonstrating that, although the number of taxa had been equal, the taxonomical composition of the mycota changed according to the collections.
Camposporium pellucidum occurred 12 times, followed by Triscelophorus monosporus, registered 10 times and Endophragmiella sp., present 9 times during the experiment.Both in the first and in the second experiment, it was verified that some taxa occurred exclusively at certain collection point, and also some taxa occurred only incertain collections.According to Sörensen index (table 16), there is 39% similarity between the mycota of the experiments with T. pulchra leaves, that represents one of the lowest values found in the entire study, indicating that the composition of mycota differed considerably in climatically distinct seasons.
Regarding the fungal succession, the presence of higher number of taxa and occurrences of aquatic Hyphomycetes in the 2 nd and 3 rd collections of the 1 st experiment (table 11), followed by the single presence of species was also reported by Moreira (2006) for leaves of T. pulchra submerged in the Lago das Ninféias.In the 2 nd experiment (table 11), initiated in the rainy and hot season, with a greater number of fungal taxa, this occurrence pattern was not so clear,  although the total number of occurrences was also higher in the 2 nd and 3 rd collections.
It is remarkable the high occurrence of Endophragmiella sp. that has not been previously observed by Moreira (2006) or by Schoenlein-Crusius et al. (2009) in PEFI, but has been frequently observed in urban waters, together with Margaritispora and Pyramidospora, as was verified by Moreira (2011) on leaves of Caesalpinia echinata Lam. and Campomanesia pheae (O.Berg.)Landrum submerged in a lake in the park Parque Municipal Alfredo Volpi in the city of São Paulo.
Table 12 presents the fungal taxa isolated during the 3 rd experiment, initiated in the dry and cold season, using mixed leaf litter confined in nylon net litter bags.Twelve taxa were obtained, distributed in 42 occurrences, with the greatest number of taxa -8 (15 occurrences) in the 5 th collection and the lowest-two (two occurrences), in the first collection.The point 4 presented the highest number of taxa (8) and occurrences (11), whereas the opposite occurred at point 2, where only three occurrences of two fungal taxa were registered.The predominant taxa were Triscelophorus mnonosporus (11 occurrences), followed by Camposporium pellucidum (6 occurrences) and Varicosporium elodeae (5 occurrences).
The similarity index of Sörensen corresponded to 54% in the comparison of the 1 st with the 3 rd experiments, which were performed in similar seasons, but with different substrates (table 16).
Table 13 presents the fungal taxa isolated during the 4 th experiment, started in the rainy and hot season, also using mixed leaf litter confined in nylon net litter bags.Thirteen taxa, distributed in 40 occurrences were obtained, with predominance of Margaritispora aquatica (12 occurrences), followed by Triscelophorus monosporus (8 occurrences) and Lemonniera aquatica (4 occurrences).The greatest number of taxa was obtained in the first collection (10 occurrences) and the lowest in the 6 th collection (three occurrences).At point 1 and point 5, ten occurrences, the highest number of taxa were registered, while at point 6, only three occurrences.The highest index of similarity (56%) were verified comparing the mycota of the 4 th and 3 rd experiments, where the same substrate was used, but in different seasons (table 16).
Excepting the 3 rd experiment, a tendency of fungi to be present with larger number of taxa and occurrences in the first collections was observed.In the 3 rd experiment these numbers were greater in the 5 th collection, perhaps due to the impact of heavy rains after two months of drought.
Table 14 presents the fungal taxa that were isolated from free mixed leaf litter, collected concomitantly to the 3 rd experiment, initiated in the dry and cold season.Eight fungal taxa were isolated, totalizing 38 occurrences.In the free leaf litter samples, the predominance of Triscelophorus monosporus (16 occurrences), especially in the 2 nd and 3 rd collections was observed.One more species of Triscelophorus sp.1 (species not identified) occurred seven times, as well as Anguillospora crassa.Seven taxa (11 occurrences) were registered at point 1 and only one taxon (two occurrences) at points 5 and 6.Comparing the mycota of the free leaf litter with the one of the confined leaves from the 3 rd experiment, a similarity of 50% was verified.
Table 15 presents the fungal taxa obtained from free mixed leaf litter during the 4 th experiment, started in the rainy and hot season.Thirteen taxa (61 occurrences) were isolated, with predominance of Anguillospora longissima (13 occurrences), Triscelophorus monosporus (12 occurrences), followed by one not identified species of Pyramidospora sp.(9 occurrences).The greatest number of taxa -nine (19 occurrences) was registered in the 5 th collection and the lowest -three taxa (three occurrences) in the 3 rd collection.The points 1, 4 and 5 presented equal total number of taxa (seven), being the highest number of occurrences (thirteen), registered for points 1 and 4.An index of 48% of similarity (table 16) was found comparing the mycota of the free leaf litter with the one of the confined leaves during the 4 th experiment, as well as in the comparison with the 3 rd experiment (mixed confined leaf litter, during the dry and cold season).
The index of similarities presented on table 16 varied from 39 to 56% and could be considered low when compared to the ones cited in studies about fungal succession such as in Schoenlein-Crusius & Milanez (1998) and Moreira (2006).Besides, according to Christensen (1989), in studies about fungi in tropical environments, similarity indexes around 70% are expected.Similarities below this value could be considered low.
Although the total number of fungal taxa obtained in each of the experiments in the present study may be considered modest in comparison to other studies, which includes the citation of 24 taxa of aquatic Hyphomycetes in the entire area of the Botanical Garden (Schoenlein-Crusius et al. 2009), the quantity of taxa in the different experimental situations was similar.The composition of the mycota, however, varied considerably in each situation, presenting the predominance of certain taxa in distinct collection and points.
The interpretation of the results of the present study is complex, because the confined leaves remained submerged and were gradually removed from the water, so that the isolated mycota followed a temporal sequence, whereas this was not the case for the free mixed leaf litter samples.
Added to this, the four experiments were conducted while the stream was in a regeneration phase, approaching increasingly to an situation of a natural water body, with new plants flanking the margins, along with new organisms as fishes and amphibians occupying new spaces in the forming environment.
The first two experiments were conducted with leaves of Tibouchina pulchra.The mycota of this leaf species was studied for the first time by Grandi & Gusmão (2002) in the Reserva Biológica do Alto da Serra de Paranapiacaba, in São Paulo State.In that study the authors obtained 22 fungal taxa employing the method of incubating leaf litter fragments in moist chambers.From their fungal list, only Beltrania rhombica is cited here for the T. pulchra leaves of the 2 nd experiment, during the rainy and hot season.
Also in the Atlantic rainforest of Paranapiacaba, Moreira (2002) submerged leaves of T. pulchra confined in nylon net litter bags, in a stream, obtaining 40 fungal taxa, including six aquatic Hyphomycetes such as: Lunulospora curvula Ingold, Triscelophorus monosporus Ingold, Lemonniera aquatica de Wild., Anguillospora crassa Ingold, Tripospermum myrti, and Margaritispora aquatica Ingold.The presence of the taxa was more intensively observed after some weeks of decomposition of the leaves.
In the study of Moreira (2002), among the aquatic Hyphomycetes, Margaritispora aquatica was the most abundant taxa in T. pulchra leaves.In the present study, this taxon was the most abundant in the mixed leaf litter samples during the 4 th experiment, not being the most common in T. pulchra leaves, although it was present mainly in the two initial collections of the two experiments.
Triscelophorus monosporus, in general, is one of the most common taxa mentioned in Brazilian studies about ingoldian fungi (Schoenlein-Crusius & Grandi, 2003).In the present study, the species showed a high number of occurrences during the 2 nd experiment, was abundant during the 4 th experiment and was one of the most frequent species in the free mixed leaf litter.The species was not the most frequent one in the study conducted by Moreira (2002).In a similar way, Lunulospora curvula, which is also very common in Brazilian studies, was not so frequent in the T. pulchra leaves in the present study and in the experiments mentioned by Moreira (2002).
The fungal succession and decomposition rate of submerged leaves of T. pulchra were studied by Moreira (2006) in the PEFI, at the Lago das Ninféias, which waters supplies the Pirarungaua stream.The mycota found in the lake, formerly characterized as meso-oligotrophic was compared to a hypertrophic lake called Lago das Garças.In the first lake, Moreira (2006) found 79 taxa, including two aquatic species.In that study and now, in the Pirarungaua stream, the following fungal taxa were found in common: Anguillospora crassa, Dictyochaeta, Pestalotiopsis and Fusarium oxysporum.
The fungal succession studied by Moreira (2006) in the PEFI included aquatic facultatives, zoosporic organisms and few ingoldian fungi, much less than occurred in the Atlantic rainforest of Paranapiacaba (Moreira 2002).The mycota composed by aquatic Hyphomycetes in the T. pulchra leaves in the present study approaches more the one found in Paranapiacaba than in the Lago das Ninféias in the Botanical Garden of São Paulo.
Considering that the methodology used to isolate the fungal species in these studies was similar, as well as the plant species used as substrate, one may assume that the type of aquatic environment and its intrinsic abiotic factors exert great influence on the composition of the mycota associated to decomposing plant substrates, what is in agreement with the literature (Bärlocher 1992).
However, Moreira (2006) verified a slower decomposition rate in the meso-oligotrophic lake than in the hypereutrophic lake, attributing this result to the massive presence of aquatic plants and probable presence of humic substances, which could have had a hampering effect on the diversity of ingoldian fungi.
Concomitantly to the study of fungal succession in the Lago das Ninféias (Moreira 2006), a survey of aquatic Hyphomycetes on submerged mixed leaf litter was conducted in several water bodies in the entire area of the Botanical Garden, revealing 24 taxa, with predominance of Anguillospora crassa, Lunulospora curvula, Tetrachaetum elegans and Camposporium pellucidum (Schoenlein-Crusius et al. 2009).In the present study, the first three fungal taxa occurred almost in all experiments.However, the occurrence of Tetrachaetum elegans was registered only in the confined T. pulchra leaves in the 2 nd experiment and in the free mixed leaf litter during the last experiment.
From the 24 total fungal taxa obtained in the survey mentioned above (Schoenlein-Crusius et al. 2009), 12 (50%) were no more observed in the present study.On the other hand, from the 35 total taxa registered in the experiments in the Pirarungaua stream, 12 were already cited for the Botanical Garden, but the remaining 23 are novelties for the area, including three new citations to Brazil.
Although the Pirarungaua stream is small and being in revitalization processes it is a strategic location for ecological studies of fungi, because the diversity is expressive.As the artificial environment will approach gradually the natural, probably other taxa may be revealed in the future, adding valuable information to the geographical distribution of fungi.Example of that is Ulocoryphus mastigophorus from the free mixed leaf litter (table 15), whose presence was only observed in Agathis australis leaves in New Zealand (Michaelides et al. 1982).
Initially much was asked how a taxon observed in such a distant continent suddenly was present in a small stream in São Paulo, but it has to be taken in account that the drastic modification of the watercourse from channeled to an open system that receives all the influence of environment, provides new niches and the opportunity for previously unknown species to become expressive in the coming taxonomic surveys.
A stream, even being small and designed primarily to compose an aesthetic landscape element, is still a part of the ecosystem where the biodiversity is established according to the biotic and abiotic conditions, deserving to be recovered, preserved and conserved for future generations.
and selection of sampling points at the Pirarungaua Stream -The present study was divided into four experiments conducted in Pirarungaua stream in the stretch phase of revitalization and comprising a length of approximately 229 m, with 2.80 to 3.00 m in width and minimum depth of 15 cm.The distance between the input stream and the first waterfall corresponds to 62 m.From this place to the second fall comprises 59 m, and until the third fall more 55.7 m.The distance of 25.4 m separates the third and fourth waterfall and more 27 m has to be transverse from this point until the river mouth.The following six collection points were marked with wooden stakes: Point 1 -water entrance in the stream, Point 2 -under the first bridge after entry, Point 3 -near the first waterfall; Point 4 -near the second waterfall, under the third bridge; Point 5 -near to the third waterfall, under the fourth bridge and Point 6 -near the river mouth of the channel.Characterization of climatic conditions during the experiments -Data regarding rainfall, average relative humidity and air temperature during the experiments was provided by the Astronomical and Geophysical Institute, University of São Paulo, located within the Parque Estadual das Fontes do Ipiranga -PEFI.

Table 1 .
Monthly total rainfall (mm), monthly mean values of temperature ( o C), and relative air humidity (%) from July 2010 to April 2012.Data provided by the Astronomical and Geophysical Institute, University of São Paulo.

Table 2 .
Individual values and means of the water temperature ( o C) measured during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brasil.

Table 3 .
Kruskal-Wallis test related to water temperature ( o C) measured during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brazil.Values in gray shades indicate significance (p < 0.05).

Table 4 .
and Schoenlein-Individual values and means of the water pH measured during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brasil.

Table 5 .
Kruskal-Wallis test related to water pH measured during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brazil.Values in gray shades indicate significance (p < 0.05).

Table 6 .
Individual values and means of the dissolved oxygen (mgO 2 L -1 ) content of the water measured during the four experiments in the Pirarungaua stream, in the Botanical Garden in São Paulo, Brazil.aquaticfacultatives.New citations for Brazil are Helicoon spirale Boedjin, Triscelophorus curviramifer Matsushima and Ulocoryphus mastigophorus Michaelides, L. Hunter & W.B. Kendr.Tables 10 to 15 present the occurrences of the aquatic Hyphomycetes in the four experiments and table 16 summarizes the similarity indexes of Sörensen calculated between the different situations analyzed in the present study.Table10presents the fungal taxi of the 1 st experiment, initiated during the dry and cold season, with leaves of T. pulchra confined in nylon net litter bags.During the 1 st experiment 14 taxa were obtained, totalizing 48 occurrences.The highest number of taxa (8) was registered in the second and the lowest in the first collection.Point 1 presented the highest number of occurrences, whereas in point 2 the highest number of taxa (8).Endophragmiella sp.occurred

Table 7 .
Kruskal-Wallis test related to the dissolved oxygen (mgO 2 L -1 ) during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brazil.Values in gray shades indicate significance (p < 0.05).

Table 8 .
Individual values and means of the conductivity (µScm -1 ) of the water measured during the experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brazil.

Table 9 .
Kruskal-Wallis test related to the conductivity (µScm -1 ) during the four experiments in the Pirarungaua stream in the Botanical Garden of São Paulo, Brazil.Values in gray shades indicate significance (p < 0.05).

Table 15 .
Occurrences of conidial fungi in the free submerged leaf litter in the Pirarungaua stream during the 4 th experiment,