Benthic macroinvertebrates and degradation of phytomass as indicators of ecosystem functions in flooded rice cropping

The objective of this work was to evaluate the ecosystem functions of a natural wetland and of artificially flooded rice areas, managed under organic and conventional systems, by phytomass degradation and by the colonization of this material by benthic macroinvertebrates. The experiment was carried out in a natural wetland area, and in two flooded rice areas managed under organic and conventional systems. Twenty-five decomposition bags filled with 10 g of dry vegetation were installed in each site. At 14, 28, 42, 56, and 70 days after the beginning of the experiment, five bags from each site were collected. Macroinvertebrates were identified and classified by functional trophic group. The number of species of benthic macroinvertebrates increased: natural wetland > organic system > conventional system. The Chironomidae group was present in all areas, confirming its food plasticity and adaptability to different substrates and environmental stress situations. The Amphipoda group was present only in the artificially flooded rice area, and the Acari, only in the natural wetland. The diversity of species in the natural wetland area was higher than in the artificially flooded rice area. Nutrient cycling, provided by phytomass decomposition, is affected by the management system, and the delay in this process causes a reduction of the ecosystem functions in the conventional system.


Introduction
Remaining flooded rice straw and post-harvest sprouts (traditionally developed in rice regrowth, in the southern end of Rio Grande do Sul) are important as fodder for livestock (Monks et al., 2002).Straw degradation ensures a gradual supply of significant amount of nutrients to crop in succession in well managed systems (Massoni et al., 2013).In addition, if the area is replanted, a rapid elimination of straw is necessary in rice cultivation, to allow of planting to be carried out in a timely manner within the recommended period (Massoni et al., 2013).
Irrigated soils for rice cultivation, comprising humid or waterlogged environments, have a poor drainage (Monks et al., 2002) which persists after harvest.In those wet environments, decomposition can be divided into the three following phases: leaching, which releases water-soluble organic and inorganic compounds that are important for microbial and phytoplanktonic metabolism; conditioning or catabolism, which involves colonization by microorganisms, mainly bacteria and fungi that will macerate, metabolize, and embed leaves for secondary production, as well as increase of the palatability and nutritional value of litter; and fragmentation that can be carried out biologically by detritivorous macroinvertebrates, other microorganisms, or physical fragmentation (Cunha-Santino & Bianchini Jr., 2006;Gimenes et al., 2010).
In wetlands, benthic macroinvertebrates are responsible for sediment biorevolving, releasing nutrients into water, and accelerating the cycling process.They have an important role in the trophic dynamics, and constitute the link between the grower and the consumer, by performing filtering, fragmentation, and scraping of food, processing and converting it into phytomass available for other organisms, as well as assisting in the decomposition process (Nin et al., 2009;Oliveira & Callisto, 2010).These organisms are components of the biological diversity of natural as well as artificial wetlands, which include agroecosystems of flooded rice fields.They perform important ecosystem functions (Melo et al., 2015), by carrying through the ecosystem services in these areas.According to Groot et al. (2002), an ecosystem function is the ability of natural process and components to provide goods and services that satisfy human needs either directly or indirectly.Therefore, the nutrient cycling process is linked to the provision of ecosystem regulation services.Studies conducted in different regions have shown the biodiversity importance of soil organisms on the maintenance and distribution of other ecosystem services, which requires the understanding of occurrence patterns of these mechanisms, in different ecosystems, and the types of functional microorganisms (Bardgett & Putten, 2014;Handa et al., 2014).
For Silveira et al. (2012), irrigated rice production in Rio Grande do Sul is characterized by highly technicalized production; therefore, pesticides are used to prevent invasive plants, insects, and diseases from damaging the production.Chemicals cause ecological risks to different organisms, and the use of conjugate products, which is common in cultural practices, results in additive actions between them, even for less toxic ones (Nakagome et al., 2007).This technicalization arises from conventional rice growing systems, in contrast to organic production systems of irrigated rice, which not only are performed without the use of pesticides and synthetic fertilizers, but also are based on crop rotation (Wilson et al., 2008).
Species diversity and functional redundancy of the microorganisms responsible for providing these processes are ecological strategy keys to bring stability to the system and sustainability to the crop (Thrupp, 2000).Thus, the degradation process can provide important information on ecosystem functionality (Telöken et al., 2011), and can help in the definition of a sustainable management.
The objective of this work was to evaluate the ecosystem functions of a natural wetlands and of artificially flooded rice areas, managed under organic and conventional systems, by phytomass degradation and by the colonization of this material by benthic macroinvertebrates.

Materials and Methods
The experiment was carried out in the municipality of Santa Vitória do Palmar, RS, at 10 m altitude, Brazil (32º45'00"S; 52º38'54"W), one of the largest flooded rice growing areas in the country.The climate of the region is characterized by a Cfa (humid subtropical), according to the classification by Köppen-Geiger, with an average annual temperature of 16°C, and precipitation between 1,800 and 2,200 mm (Kottek et al., 2006).
Rice is sown between September and November in the region.A minimum cultivation is applied under conventional systems, which involves the previous desiccation of vegetation by a nonselective herbicide for direct seeding.The seeding occurred in October 2013, and seed were treated to prevent infestation by Oryzophagus oryzae.Organic farming requires soil preparation in the planting period, which eventually causes delays, due to the possible entry of machinery in the field.In the studied site, sowing occurred in late November 2013.In the conventional system, 60 kg ha -1 P 2 O 5 and K 2 O were added in the planting, and, for fertilization, 150 kg ha -1 N was applied in the differentiation of the flowering event.There was no fertilization in the organic system, except for the residual fertilization used on pasture, which consisted of 200 kg ha -1 natural phosphate.Irrigation started for both systems at the three-to-four leaf stage (V3/V4), and water blade was kept around 10 cm.Rice under conventional system was harvested in February 2014, and under the organic system, in March 2014.
To evaluate the proposed systemic functions (Table 1), 20x15 cm decomposition bags were made with a combination of mesh spacing.One side was made of 0.5 cm mesh between knots, and the other, of 0.1 cm mesh, as proposed by Carvalho & Ueida (2009).The bags were filled with 10 g rice straw, collected in the experimental station site Terras Baixas of Embrapa Clima Temperado, and straw was oven-dried at 60°C until constant weight was obtained.
The experiment was conducted between May 26 and August 4, 2014, in three different management systems: organic rice farming; conventional farming; and a natural wet site (natural wetland vegetation) in the surroundings of the cultures (Figure 1).Seventyfive decomposition bags were introduced -25 bags each system -, which were fixed at stakes and equally distributed in five points in each system.
At installation and during the collections of the decomposition bags, four physical and chemical water parameters were monitored: pH (pHmetro Digimed DM2P), conductivity (conductivity meter Digimed DM3P), temperature (mercury thermometer), and height of water blade (graduated ruler).
At 14, 28, 42, 56, and 70 days, five bags were individually removed from each site, and were warped with a 0.05 mm mesh to prevent loss of plant material and organisms.In the field, the bags were stored in plastic bags, and sent to a laboratory at Embrapa Clima Temperado, where they were opened and washed in running water in 850 µm and 250 µm sieves.
Retained biomass in the 850 µm sieve was stored in paper bags, which were labeled and oven-dried at 60°C until constant weight was attained; then, Table 1.Ecosystem services obtained by processes and functions from organisms in the soil-water interface, assessed according to Groot el al. (2002).they were weighed on an analytical balance.Benthic macroinvertebrates sorted from the material retained in the sieves were identified under a stereomicroscope using a suitable identification key (Mugnai et al., 2010), and after this procedure, they were recorded, and stored in 70% alcohol.Functional trophic groups (FTG) of the sampled benthic macroinvertebrates were categorized according to the classification proposed by Cummins et al. (2005) and to supplementary bibliography (Mormul et al., 2006;Oliveira & Callisto, 2010;Barbola et al., 2011;Santos, 2014;Telöken et al., 2014;Santos & Rodrigues, 2015).
The diversity index of Shannon-Wiener was calculated by the program Dives 3.0 (Rodrigues, 2014), and the occurrence constancy was calculated by the equation C = [(number of samples with species/total number of samples) × 100].Constant species occurred in 50% or more of the samples, accessory ones in the range between 25 and 50% of samples, and incidental species occurred in less than 25% of the samples.
The model of Petersen & Cummins (1974) was used to calculate the percentage of remaining biomass (%R) and biomass decay coefficient (k), according to the respective equations % R = (Wt/W0) × 100, and k = -(1/t) × ln (Wt/W0), in which: Wt is the biomass weight at the time of observation; W0 is the initial biomass weight, and the decay coefficients are classified as slow, when k<0.005 per day, or fast, when k>0.010 per day.
The correlation analysis between the physicochemical parameters and the organisms in the studied sites was conducted using the Mantel test, with the Multiv program (Pillar, 2001), and the rope distance was considered as a similarity measure.The randomization test (Pillar & Orlóci, 1996), carried out with the Multiv program, was used to evaluate the differences (p<0.05) of the physicochemical data of water, biomass degradation, abundance of benthic macroinvertebrates, as well as the diversity index between the studied sites.The Euclidean distance was used as a similarity measure for biotic parameters, and, the Gower index, for the abiotic parameters, without performing data transformation.The correlations between biomass decay and the diversity index, biomass decay and abundance, and biomass decay and the number of functional trophic groups were also evaluated for straw from organic and conventional farmings, and from the natural wetland vegetation (p<0.05).

Results and Discussion
There was no difference (p>0.05) for water physicochemical parameters between the three sampled sites.Electric conductivity showed a great variation, and the highest values were observed in the conventional farming (Table 2).Chemical inputs used in conventional management may have affected this difference; the organic farming does not make use of these inputs (Wilson et al., 2008).No correlation was observed between the physicochemical and biotic parameters (p>0.05)(Figure 2).
Using decomposition bags in a lentic environment, in the southern end of Rio Grande do Sul state, Telöken et al. (2011) found a greater abundance of organisms than that found in the present study.However, despite the proximity of the sampled sites, this work was conducted in the winter, while Telöken et al. (2011) performed theirs in the autumn with an average temperature of 23.95°C, which may have affected these differences, as well as the differences between the evaluated plant species (Carvalho & Uieda, 2009;Gimenes et al., 2010).
The abundance of benthic macroinvertebrates was higher (p<0.05) in the natural wetland than in the cultivated fields, which is opposite to what was reported by Pires et al. (2015), who evaluated irrigated rice farmings during the crop season, and reported drought problems at the time of the study.In the present work, abundancy in the site of the organic farming was higher (p<0.05)than in the conventional one, in accordance with the results obtained by Wilson et al. (2008) during the crop season.
The diversity index showed variation over the period, and its were higher (p<0.05) at days 14, 28, and 56 than those observed at days 42 and 70.The management systems showed that the diversity index in the conventional cultivation was lower (p<0.05)than that observed for the natural wetland.However, the diversity index for the organic systems did not differ from that in the other sampled sites.The results were different from those observed in rice fields in Australia, where the organic cultivation showed higher diversity indices throughout the experimental period (Wilson et al., 2008).Keeping diversity is a feature of more sustainable systems (Thrupp, 2000), which increases redundancy of microorganisms performing the same systemic functions (for example, more organisms of a single trophic group); this allows of organisms to carry out the processes, and ensure maintenance of the ecosystem functions, even if some of these organisms decrease, due to their sensitivity to some environmental factors.
According to Pires et al. ( 2015), the fauna recorded in the conventional system is usually rich and diversified; however, it is characterized by a high ecological dynamism, and by its composition of environmental disturbance-tolerant groups, with a high capacity for colonization.During the sampling period, sensitive organisms to environmental changes, such as Ephemeroptera (Kerans & Karr, 1994) were not found in the conventional systems, which corroborates data of Pires et al. (2015).
The occurrence frequency (Table 4) showed that the organisms were distributed in the fields as follows: in the conventional system, Chironomidae and Amphipoda; in the organic system, Hirudinea, Chironomidae, Nematoda, and Amphipoda; and, in the natural wetland, Hirudinea, Acari, Chironomidae, Nematoda, and Collembola.Chironomidae were present in all studied areas.Colonization in decomposition bags by this family has been observed by various authors, underscoring its dominance both in lotic and lentic environments (Moulton & Magalhães, 2003;Mormul et al., 2006).This situation probably occurs because Chironomidae show great food plasticity, and high adaptive power to different substrates, and to different situations of environmental stress.This group is important for the leaf degradation processes in the litter with low nutritional value, and absence of shredders (Moretti et al., 2007).
The Amphipoda group was constant only in the conventional and organic systemss, while Acari occurred only in the natural wetland.For Nin et al. (2009), colonization by amphipods starts in the first hours of exposure, and it is later replaced by other groups, especially insects.Santos & Rodrigues (2015) observed that over time there is an increase of colonization and succession of smaller organisms, such as mites.Considering that the colonization by shredders in the natural wetland occurred before the first collection, these herbivores (shredders) were replaced by detritivores (collectors), and these, by carnivores (predators).From these predators, Acari were the major representatives in the natural wetland, while Nematoda were predominant in the cultivated fields.Probably, this distribution is due to Acari sensitivity to physical disorders (Di Sabatino et al., 2000), according to the sensitivity range of macroinvertebrate families of Costa Rica (Rizo-Patrón V. et al., 2013).The comparison of the occurrence frequency (Table 4) with this scale showed that in the conventional rice cultivation, accessory and constant families are the only ones classified as sensitivity 3.In the conventional management, in addition to these families, Haliplidae with sensitivity 4 occurred.In the wetland, accessory and constant families were Acari and Caenidae, with sensitivity 4.
Among the functional trophic groups (FTG), removed from the bags at the day 14, the presence of shredders was observed in the conventional system, and predators were prevalent, followed by collectors (Figure 2).During this period, the number of shredders in the conventional system increased, and FTG were dominant in this management system, in comparison to the others.FTG succession was apparently faster in the organic farming and in the natural wetland than in the conventional system.In the natural wetland, detritivores appeared in a greater proportion at the day 14; and, in the organic management, they appeared at the day 28, gradually increasing after this period.For Telöken et al. (2011), despite the high correlation of shredders Table 2. Mean values and standard deviation of physical and chemical parameters by the time of sample collection (1) .Values followed by equal letters, in the lines, show no significant difference, at 5% probability.The standard deviation is described in parentheses.with foliar degradation rate in temperate regions, the contributions of this FTG for the litter decomposition in subtropical areas have been smaller than in areas of higher latitudes.This was also observed in the present study; however, the first evaluation occurred only at the day 14.Shredders might have occurred in the first two weeks, mainly in the organic cultivation and in the natural wetland; however, in the late collection period, they were not detected.After this period, shredders were replaced by carnivores and omnivores in the natural wetland and in the organic cultivation, while detritivores reached a higher proportion between FTG in the conventional system only at the day 70.
The great presence of predators and parasites occurs probably by the availability of potential prey (Mormul et al., 2006).The number of functional trophic groups showed no significant correlation with decay of foliar biomass, which is probably due to this dynamic in colonization.
The different management systems were equivalent as for biomass loss; however, there was interaction between time and management systems.There was difference between the conventional and the other management system at 14 and 28 days after incubation (p<0.05)(Figure 3).According to the decay coefficient, proposed by Petersen & Cummins (1974), the results in all periods and management systems were greater than 0.010 per day, which expresses a quick biomass decay.
In the present study, there was a fast biomass decay, which was more expressive at the day 40, mainly in the natural wetland that showed a greater richness during this period, indicating that the action of these organisms was effective for biomass decomposition.There was no correlation between diversity and biomass degradation in the organic cultivation and in the natural wetland, while in the conventional system, this correlation was significant.Probably the redundancy between groups makes diversity greater than the minimum required to keep this ecosystem functioning (Moulton et al., 2003), both in the organic system and in the natural wetland.However, for the conventional system, this diversity is smaller, and any variation is related to losses in the ecosystem function.Our results on degradation agree with those reported by Telöken et al. (2011), which had greatest biomass losses in the initial phase, with a trend to stabilization that extends from the day 32 until the end of the experiment, at day 71.collector; filtering-collectors; scrapers; shredders; predators/ collectors-gatherers; collectors) found in the sampled sites at 14, 28, 42, 56, and 70 days after the beginning of incubation.CT, conventional system; OF: organic system; and NW, natural wetland.
Abundance did not correlate with biomass decay in the systems, but it was significant for the natural wetland, showing that the benthic macroinvertebrates in cultivation areas were probably affected, having Table 3. Abundance of benthic macroinvertebrates found in bags of foliar degradation in the areas of conventional (CT) and organic systems (OF) residues, and on the area of the natural wetland (NW), at 14, 28, 42, 56, and 70 days after the introduction of degradation bags, and their respective functional trophic group (FTG).
Taxon FTG (1)  DAY   The systems studied had a lower abundance of benthic macroinvertebrates than natural wetlands; and the diversity of these organisms was also reduced in the conventional system.The evaluation of phytomass degradation was sensitive to this differentiation of the environmental management, showing its effects on the functionality of this agroecosystem.
The reduction of ecosystem functions causes the declining of nutrient availability, affecting the natural processes and other services, as the maintenance of plant and animal biodiversity, thus increasing crop production costs; this way, the use of the organic management or other less impactful systems, as water management for invasive plants and integrated pest management, are recommended to provide the maintenance of the evaluated processes that may improve gains from cultivation sustainability.

Conclusions
1.The management systems promote differences in the colonization of rice straw by benthic macroinvertebrates.
2. The need to maintain diversity of benthic macroinvertebrates becomes clear when there is diversity loss, as observed in rice straw biomass in the site of conventional system.
3. Nutrient cycling, provided by phytomass decomposition, is affected by the management system, and the delay in this process causes a reduction of the ecosystem function in the conventional system, thus affecting its sustainability; nevertheless, 70 days after the phytomass incubation, the remaining straw is similar in all treatments, which shows a resilience ability.

Figure 1 .
Figure1.Map of the experimental sites in Santa Vitória do Palmar, state of Rio Grande do Sul, Brazil: NW, natural wetland area; CT, conventional systems; and OF, organic systems, and the points from which five units were sampled, in each site, are described according to the collecting day after the incubation, at 14, 28, 42, 56, and 70 days.

Figure 3 .
Figure3.Decay coefficient of phytomass degradation and amount of remaining phytomass during the experiment of the studied management systems.CT, conventional system; OF, organic system; and NW, natural wetland.

Table 4 .
Occurrence frequency of benthic macroinvertebrates found in foliar degradation in the areas of conventional (CT) and organic (OF) systems, and natural wetland (NW).species (present in less than 25% of the collections); Ac, incidental species (present in less than 50%, and more than 25% of the collections); and Co, constant species (present in more than 50% of the collections).