INTRODUCTION

In Chile, forested wetlands show a wide distribution (29° S- 42° S), though they mostly dominate in the Mediterranean zone (MZ) (33° S - 39° S). In that zone, most wetlands show some degree of human intervention (e.g. urbanization, wood extraction and introduction of exotic species) (^{Correa-Araneda et al. 2011}), which has led to their fragmentation, eutrophication and reduced ecosystem services. Arboreal vegetation of these ecosystems is fundamental for their functioning, because it causes a low penetration of light, limiting the primary activity of phytoplankton (^{Benfield 1996}) and it implies that their main source of organic matter is the material coming from the woody component (e.g. leaves, twigs, flowers and fruits) (^{Correa-Araneda et al. 2011}). This material has been mainly supplied by hydrophilic native species from the Myrtaceae family, such as Blepharocalyx cruckshanksii (Hook. & Arn.) Nied and Myrceugenia exsucca (DC) Berg (^{Correa-Araneda et al. 2012}). However, in the last four decades they have been replaced by exotic species, mainly Pinus radiata (D. Don) and Eucalyptus globulus (Labill), which have been steadily increased in the Chilean Mediterranean Zone (ChMZ) (Shulz et al. 2010, 2011; ^{CONAF 2011}). For these species, an increase of up to 53% in the period 1975 - 2000 has been registered, with an annual growth that reached 10.5% (^{Pizarro et al. 2006}; ^{Little et al. 2009}) and currently representing more than 17% of the forest area of the country (^{CONAF 2011}).

The replacement of native vegetation of the ChMZ by monocultures of exotic species could be influencing the processes that regulate flows of matter and energy to and within the system, as reported in lotic systems (^{Canhoto & Graça 1995}; ^{Canhoto et al. 2013}). One of the processes involved in the energy flows in the wetlands corresponds to the degradation of the leaf litter and includes several trophic levels (^{Solada et al. 1996}). This process starts with the release of soluble compounds (^{Solada et al. 1996}), followed by a colonization of bacteria and fungi (BF) that facilitates the subsequent fragmentation and consumption by benthic macroinvertebrates (BMI) (^{Abelho 2001}; ^{Weyers & Suberkropp 1996}; ^{Gessner et al. 1998}). This indicates that both communities (BF and BMI) play a critical role in maintaining the flows of matter and energy (^{Sylvestre & Bailey 2005}; ^{Boyero et al. 2011}).

Recent studies have determined that the decomposition of the leaf litter in freshwater ecosystems is influenced for both environmental variables (e.g. pH, temperature, water regime) and structural components of the leaf itself, such as lignin, cellulose, cutin and tannins (^{Richardson et al. 2004}; ^{Swan & Palmer 2004}; ^{Graça & Cressa 2010}). That is why any change in the type of vegetation that is incorporated into the water component of the forested wetlands would impact significantly on the functioning of the system, affecting the capacity of processing this material (^{Graça & Canhoto 2006}). However, the mechanisms are not entirely clear, since wetlands have been very poorly studied (^{Correa-Araneda et al. 2011}). Because of the abovementioned backgrounds, the present study is aimed to evaluate the degradation process of the leaf litter carried out by the communities of BF and BMI on the foliar component of native and non-native plants from the forested wetlands.

MATERIAL AND METHODS

RESULTS

The degradation of the foliar detritus was faster in the treatments with non-native species (Pin, E) in comparison to native species (T, Pit) (Fig. 2a, 2b). Leaves of E. globulus presented the highest degradation rate, leaving 61.9% of the leaves exposed to BF and 58.8% of the leaves exposed to BF and BMI at the end of the exposure period (60 days). The lowest degradation rate among native species was presented by Temu (B. cruckshanksii), where 86.9% of the leaves exposed to BF remained and 79.6% of the leaves exposed to BF and BMI. The greatest degradation of leaf litter was registered during the first 15 days for both non-native and native species (Fig. 2c, 2d).

Covariance analyses (ANCOVA) for the degraded plant species indicates that there are statistically significant differences (p < 0.05) between the exposure to BF and BF- BMI, as well as between types of leaf litter, but there is no co-variation between the types of tree of came from and the exposure of the different communities (Table 1).

Regarding the toughness of the leaves, the average of the initial toughness of the experiment was 387 g in M. exsucca, 219 g in B. cruckshanksii, 251 g in E.globulus and 293 g in P. radiata. No clear pattern of increase or decrease over time was observed (Fig. 3). Indeed, this pattern tended to maintain constant, without significant differences (p > 0.05).

However, this variable did presented significant differences between type of leaves (native- exotic) (p = 0.002). The Tuckey HSD test for toughness indicated that the differences between types of leaves are mostly given by the leaves of exotic origin, specifically of the species E. globulus (EPit vs EPin, p < 0.05).

Physicochemical variables measured in the water column presented a strong direct correlation between conductivity, dissolved oxygen and turbidity, as well as between temperature and pH. Only turbidity showed a clear trend to decrease at temporal scale (T₀ = 75.2 NTU - T₁ = 1.8 NTU). However, all variables presented a high dispersion respect to their mean value (C.V. > 0.05) (Table 2) all the variables with correlation, show significant differences (p < 0.05) less the correlation between turbidity and O.D.

FIGURE 1 Location of the study area corresponding at the wetland Petrenco located in the commune of Pitrufquén, Araucanía Region, Chile. / Ubicación del área de estudio correspondiente al humedal Petrenco de la comuna de Pitrufquén, región De la Araucanía, Chile. 

FIGURE 2 Temporal variation of remnant mass rate of the different species exposed to colonization by BF (A) and BF-BMI (B), and the degradation rate of the different species exposed to colonization by BF (C) and BF-BMI (D) during their 60 day exposure in the water column. / Variación temporal de la masa remanente de las diferentes especies expuestas a la colonización por BF (A) y BF-BMI (B) y tasa de degradación de las diferentes especies expuestas a la colonización por BF (C) y BF-BMI (D) durante los 60 días de exposición en la columna de agua. 

TABLE 2 Values of the physicochemical variables of the water column in the different sampling days, Coefficient of variation (CV) and correlation analyses between variables. Cond. = Conductivity (μS/cm), Temp. = Temperature (ºC), O.D. = Dissolved oxygen (mg L-1 ), * = strong correlation. / Valores de las variables fisicoquímicas de la columna de agua en los diferentes días de muestreo, coeficiente de variación (CV) y análisis de correlación entre las variables. Cond. = Conductividad (μS/cm), Temp. = Temperatura (ºC), O.D. = Oxígeno disuelto (mg L-1 ), * = Correlación fuerte. 

FIGURE 3 Temporal variation of the toughness (g) in the leaves of native (M. exsucca and B. cruckshanksii) and non-native species (E. globulus and P. radiata). / Variación temporal de la dureza (g) en las hojas de especies nativas (M. exsucca and B. cruckshanksii) y exóticas (E. globulus and P. radiata). 

DISCUSSION

The general phases of the degradation of the foliar material in aquatic ecosystems have been described as sequential, initiated by a lixiviation process of the soluble compounds, followed by a colonization of BF which facilitates the arrival of the BMI, especially shredders (^{Cheshire et al. 2005}). Our results showed that the greatest weight loss was registered during the first 15 days for both native and exotic species. This coincides with liberation phase of soluble compounds, described as the phase of greater weight loss in the foliar material (^{Solada et al. 1996}).

In the heads of the mountain lotic environments like in the wetlands, light penetration is scarce and they receive a high load of fallen leaves (^{Lugo et al. 1990}), and the decomposition process is mainly given by BMI (^{Vannote et al. 1980}). In forested wetlands BMI would also be a determinant factor in the processing of organic matter, because in containers with presence of these organisms together with communities of bacteria and fungi (BF) was observed a significantly higher decomposition than in the case where only BF were present. This is in agreement with what was reported in tropical forested wetlands by ^{Del Valle (2003)}, where the macroinvertebrate community was also relevant in the processing of organic matter. However, forested wetlands present important differences respect to mountain river environments, as well as in other lentic environments (e.g. freshwater marshes and exposed lagoons), both in physical, chemical and biological characteristics. The fact that there is no mechanical effect produced by the water movement and the benthic components must be highlighted (^{Casas et al. 2000}), because this would explain the high non- degraded percentage of organic material in these ecosystems. This implies that the total degradation of the plant material would last twice the time considered for the current study or even more. It is interesting to mention that the shredding organisms not showed preference by native or non-native leaf litter, so that differences registered in the decomposition rate were mostly influenced by species-specific features. This is the case of the toughness of the leaves which in native species was higher compared to exotic species, as reported by ^{Ramírez et al. (1980}), who indicated that these species have a leathery nature. In addition, none of the species presented significant differences in toughness at timescale, contrary to was observed in mountain river environments from the ChMZ, where these were negatively correlated to the exposure time (^{Ovalle 2012}). In the latter case, the mechanical effect that takes place in these environments was a determinant factor on the variations of the toughness of leaves. The presence of chemical compounds such as waxes or oils secreted by the leaves, described for B. cruckshanksii (^{Tucker et al. 1993}), would provide hydrophobicity to the leaf, affecting the capacity of the BF community to penetrate the tissue. This implies that the decomposition process by these microorganisms is dependent on the individual features of each leaf.

An interesting case about the interaction between BF, BMI and the physicochemical characteristics of the leaves is presented by the species E. globulus, which has phenolic compounds whose anaerobic degradation can initiate an increment in compounds such as quinones, which are highly toxic for BMI from aerobic systems. This would not occur equally in forested wetlands, given the low concentration of oxygen registered (^{Canhoto et al. 2013}). but compounds like quinones not be affecting the bacterial degradation, causing that forested wetlands eucalyptus are degraded faster than native species. Hence, under these conditions the BF community would be more relevant on the degradation of organic matter of forested wetlands in comparison to BMI. Another variable that can affect the degradation rate, though not assessed in the present study is the surface are of the leaves, which can be affecting the size of fungi colonies associated to the degradation of organic matter (^{Allen et al. 1991}). It would be interesting to incorporate this issue in further studies.

In most aquatic ecosystems, physicochemical variables such as temperature, conductivity, amount of dissolved oxygen and nutrients are fundamental to the development of the organisms (^{Haapala et al. 2001}). In this regard, the environmental variables measured in the water column presented timescale variations, as well as direct relations between conductivity, dissolved oxygen and turbidity; this would be associated to the rainfall events that decrease turbidity and the concentration of ionic compounds responsible for the conductivity (^{Arle 2002}). This is an indicator of the buffer effect that would be produced by the forest cover, intercepting the water to enter the system by means of streamflow, which decreases the contribution and re- suspension of fine sediments (^{McIvor et al. 2012}), implying a lower perturbation of the water column (^{FAO 2005}).

Dissolved oxygen is one of the most relevant variables for aquatic life. This level was registered in low concentrations in the forested wetlands (6.2 ± 1.38 mg L-1), in comparison to that observed in aquatic environments of the same area (8.7 - 11.6 mg L-1; ^{Habit et al. 2003}; ^{Figueroa et al. 2003}; ^{Correa-Araneda et al. 2010}, ²⁰¹⁴), which implies the dominance of resident and generalist organism, such as the BF community (^{Bärlocher & Kendrick 1974}; ^{Figueroa et al. 2003}). Such organisms would be the main responsible of the degradation and changes in the toughness of the fallen trees in this kind of aquatic ecosystems.

Regarding the temperature, this variable affects the dissolution of oxygen in the water (^{Alvarado & Aguilar 2009}), the metabolism of the organisms and many other factors are directly or indirectly related to the processing of organic matter (^{Hauer & Hill 2007}). The low temperatures recorded (5.9 - 9.3°C), prevent the descent of the levels of oxygen to hypoxic levels and foster an ideal environment to the development of psychrophilic bacteria (0 - 30ºC) (^{Morita 1975}), which have a slower reproduction rate respect to mesophilic or thermophilic bacteria (Elliot & Michener 1965).

Forested wetlands present differential characteristics with other types of wetlands, highlighting the low penetration of light and explaining the low temperatures of the water, the constant supply of fallen leaves and the low degradation rate of organic matter. These features make forested wetlands particularly vulnerable against the excessive input of organic matter as a result of the diverse productive activities carried out in the area. In addition, the loss of forest material as a result of its extraction for human consumption (^{Correa-Araneda et al. 2011}) can facilitate the input of light, generating changes in the thermal regime and in the related variables (^{Döll & Zhang 2010}). The synergistic effect of these modifications might lead to a greater bacterial colonization (^{Molinero et al. 1996}), accelerating biological processes (^{Cortes 1992}, ^{Belovsky & Slade 2000}) and in turn, causing a greater consumption of oxygen. This would modify many of the physical and chemical characteristics of these ecosystems. These processes have not been yet clarified and the interest on them from a limnological point of view is emerging. Therefore, delving into these processes is of great importance to encourage and support future actions of both restoration and preservation.