11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote




Название11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote
страница12/56
Дата22.10.2012
Размер1.55 Mb.
ТипДокументы
1   ...   8   9   10   11   12   13   14   15   ...   56


Key Findings and Recommendations


The SAB Panel finds that 20th century changes in the hydrologic regime of the Mississippi and Atchalafaya Rivers and the timing of freshwater inputs to the Louisiana-Texas continental shelf have likely increased the shelf area with potential for hypoxia, although these changes occurred mostly from the 1920s to the 1960s, before the measured onset of hypoxia in the mid-1970s. Additional work is needed to advance the understanding of the relative importance of physical factors in the formation of hypoxia in the NGOM. The SAB Panel therefore provides the following recommendations.


  • The development of a new suite of models that integrate physics and biogeochemistry should be encouraged and supported. This suite should include multiple types of models [i.e., relatively simple models such as those developed by Scavia et al. (2003) as well as more complex three-dimensional types such as Hetland and DiMarco (2007)].




    • A comparative impact study of past, present, and future river flow diversions and scenarios of altered nutrient supply to the river mouths should be encouraged and supported. Three-dimensional hydrodynamic modeling studies are needed to compare the spatial distribution of salinity and stratification with 15% (early 1900’s) and 30% (post-1970’s) Atchafalaya River contributions to the combined Atchafalaya-Mississippi River outflow. Coupling of this three-dimensional hydrodynamic model with a biogeochemical model would allow quantification of the impacts of past river flow diversions on the spatiotemporal extent of hypoxia. In addition, to anticipate the possible effects of proposed future effluent diversion plans via rerouted deltaic distributaries (CPRA, 2007), these three-dimensional biogeochemical and baroclinic shelf circulation models need to be applied to scenarios such as that shown in Figure 6 while also considering the effects of nutrient-rich Mississippi River waters discharged into local bays and estuaries.


  • Emerging coastal ocean observing and predicting systems in the Gulf of Mexico (http://www.gcoos.org) should be encouraged to measure and disseminate information needed by hypoxia modelers and those charged with adaptive management. Direct measurements of physical and biogeochemical parameters as well as direct time series measurement of dissolved oxygen in the bottom boundary layer should be routinely provided by the next generation of shelf moorings.




  • Studies of turbulent mixing processes involving the effects of stratification over the Louisiana-Texas shelf with instruments and techniques capable of quantifying turbulent dissipation rates of velocity, salinity, and temperature fluctuations should also be encouraged. Studies of the importance of lateral mixing processes should be encouraged.



      1. Role of N and P in Controlling Primary Production



Nitrogen and phosphorus fluxes to the NGOM--background


Excessive nutrient loading, dominated by discharge from the MARB, enhances planktonic primary production in the shallow near-shore receiving waters of the NGOM (Lohrenz et al., 1990, 1992; Turner and Rabalais, 1994; Rabalais et al., 1999a). The nutrients of concern are nitrogen (N), phosphorus (P), and silicon (Si) in the form of silicate. Both primary productivity and phytoplankton biomass are stimulated by these nutrient sources (Lohrenz et al., 1990, 1992; Ammerman and Sylvan, 2004; Sylvan et al., 2006). The spatial and temporal extent and magnitudes of this stimulation vary significantly, and their patterns and size appear to be related to 1) amounts of freshwater discharge and their nutrient loads; 2) the nature and frequencies of discharge (i.e., acute, storm- and flood-based versus more gradual, chronic, seasonal discharge); and 3) the direction and spatial patterns of discharge plumes as they enter and disperse in the NGOM (Justic et al., 1993; Lohrenz et al., 1994; Rabalais et al., 1999b). The Integrated Assessment concluded that N loading from the MARB was the primary driver for hypoxia in the NGOM. Since the Integrated Assessment, however, considerable knowledge has been gained concerning the processes that influence primary production and the relative importance of elements other than N as is discussed below.


A proportion of the freshwater discharge transits via freshwater and coastal wetlands and coastal groundwater aquifers, which modify the concentrations and total loads of nutrients entering the NGOM (Day et al., 2003; Turner, 2005). The extent to which wetlands alter nutrient loads and the effects wetland losses have had on changes in nutrient processing and loading are subjects of considerable debate (Mitsch et al., 2001; Day et al., 2003; Turner, 2005). Nutrients can also enter this region from deeper offshore sources, by advective transport over the shelf, a modified form of “upwelling” (Chen et al., 2000; Cai and Lohrenz et al., 2005), although this input is estimated to be only 7% of the nitrogen coming down the Mississippi River (Howarth, 1998). Lastly, nutrients can be derived from atmospheric deposition directly onto nutrient-sensitive NGOM waters (deposition onto the MARB and subsequent downstream export to the Gulf is considered in later sections). For nitrogen, this direct deposition is estimated to be 13% of the amount of nitrogen that flows down the river (Howarth 1998).


Historic analyses indicate a great deal of variability in seasonal, interannual and decadal-scale patterns and amounts of freshwater and nutrient discharge to the NGOM (Turner and Rabalais, 1991; Rabalais et al., 2002a). As a result, primary productivity and phytoplankton biomass response can vary dramatically on similar time scales, which poses a significant challenge to interpreting trends in nutrient-driven eutrophication in the NGOM as in other systems (Harding, 1994; Boynton and Kemp, 2000; Paerl et al., 2006b). Furthermore, in the turbid and highly colored waters (containing colored dissolved organic matter or CDOM) of the river plumes entering the NGOM, nutrient and light availability strongly interact as controls of primary production and biomass. These interactive controls modulate the relationships between nutrient inputs and phytoplankton growth responses in this region (Justic et al., 2003a, 2003b; Lohrenz et al., 1994). Ultimately these interactions affect the formation and fate of autochthonously-produced organic carbon that provides an important source of the “fuel” for bottom water hypoxia in this region.


N and P limitation in different shelf zones and linkages between high primary production inshore and the hypoxic regions further offshore


Physically, chemically and biologically, the NGOM region is highly complex, and nutrient limitation reflects this complexity. Along the freshwater to full-salinity hydrologic continuum representing the coastal NGOM influenced by river discharge, ratios of nutrient concentrations vary significantly, both in time and space. For example, depending on the season, specific hydrologic events and conditions (storms, floods, droughts), molar ratios of total N to P (N:P) supplied to these waters can vary from over 300 to less than 5 (Turner et al., 1999; Ammerman and Sylvan, 2004; Sylvan et al., 2006; Turner et al., in press). Furthermore, additional environmental factors, such as flushing rate (residence time), turbidity and water color (light limitation), internal nutrient recycling, and vertical mixing strongly interact to determine which nutrient(s) may be controlling primary production (Lohrenz et al., 1999b). Compounding this complexity is the frequent spatial separation between high nutrient loads, the zones of maximum productivity and hypoxia (e.g., Figure 7). Conceivably, primary production and algal biomass accumulation limited by a specific nutrient in the river plume region near shore may constitute the “fuel” for hypoxia further offshore in the next zone, where productivity in the overlying water column may be limited by another nutrient. Limitation by different nutrients in different areas appears to be the case during the spring to summer transitional period, when primary production in the river plume region near shore is P limited (Lohrenz et al., 1992, 1997; Ammerman and Sylvan, 2004 Sylvan et al., 2006), but offshore productivity is largely N limited (Lohrenz 1992, 1997; Dortch and Whitledge, 1992). The relevant questions concerning causes of hypoxia are what are the relative amounts of inshore river plume (largely P-limited) versus offshore (largely N-limited) productivity and what roles do these different sources of productivity play in “fueling” hypoxia.


Early work on NGOM nutrient limitation tended to focus on the waters overlying the hypoxic zone; typically, these waters are over the shelf but farther offshore than the river plume waters. Stoichiometric N:P ratios indicated that, during summer months when hypoxia was most pronounced, N should be the most limiting nutrient (Justić et al., 1995; Rabalais et al., 2002a). This work has been the basis for the general conclusion that N is most limiting, and that reductions in N loading would be most effective in reducing “new” carbon (C) fixation and resultant phytoplankton biomass supporting hypoxia (Rabalais et al., 2002a, 2004). This conclusion, coupled with the nutrient loading trend data over the past 40-50 years, which showed N loading increasing more rapidly than P loading, has formed the basis for arguing that N input reductions would be most effective in reducing the eutrophication potential and hence formation of “new” C supporting hypoxic conditions. The 2000 report from the National Academy of Sciences’ Committee on Causes and Management of Coastal Eutrophication (National Research Council, 2000) concluded that nitrogen is the primary cause of eutrophication in most coastal marine systems in the U.S. at salinities greater than 5 – 10 parts per thousand (ppt), including the NGOM.


While it is likely that N limitation characterizes coastal shelf and offshore waters, more recent nutrient addition bioassays (Ammerman and Sylvan, 2004; Sylvan et al., 2006) and examinations of nutrient stoichiometric ratios have shown that river plume-influenced inshore productivity appears to be more P limited, especially during periods of highest productivity and phytoplankton biomass formation (Feb-May) (Figure 8) when freshwater discharge and total nutrient loading are also highest (Lohrenz et al., 1999a, 1999b; Sylvan et al., 2006).






Figure 8: Response of natural phytoplankton assemblages from coastal NGOM stations to nutrient additions, March through September. All experiments, except those done in September, indicate a strong response to P additions. Taken from Sylvan et al., 2006.


The strong P limitation during this period appears to be a result of the very high rates of N loading that have increased more rapidly than P loading over recent history (the past 50 years) (Turner and Rabalais, 1991; Turner et al., 1999). This situation is exacerbated during periods of high freshwater runoff, which typically contain very high N:P ratios. Primary productivity in the river plume region near shore tends to shift into a more N limited mode once freshwater discharge decreases during the drier summer-fall period (June-October). However, total primary production and phytoplankton biomass accumulation are far lower during this more N-limited period than during the earlier P-limited period. Overall, maximum “new” organic C formation in recent years tends to coincide with periods of highest N:P, which are P limited (Lohrenz et al., 1992, 1997, 1999a; Ammerman and Sylvan, 2004; Sylvan et al., 2006).


Field data and remote sensing imagery indicate that in situ phytoplankton biomass (as chlorophyll a) concentrations can be quite high in river plume-influenced inshore waters that have been shown to be P limited. This pattern is evident in Figure 9, an image provided by the National Oceanic and Atmospheric Administration Sea-viewing Wide Field-of-view Sensor Project (NASA-SeaWiFS, 2007). Therefore, the following question emerges. What is the spatiotemporal linkage of this P-limited high primary production and phytoplankton biomass accumulation to hypoxic bottom waters located further offshore? Furthermore, what are the relationships between N-limited production later in the summer and hypoxic conditions, which typically are most extensive during this period? These potential “relationships” are complicated by the fact that there are strong, co-occurring physical drivers of hypoxia, including vertical density stratification and respiration rates, which tend to be maximal during periods of maximum development of hypoxia (c.f. Rowe and Chapman, 2002;Wiseman et al., 2004; Hetland and DiMarco, 2007; DiMarco et al., submitted).



Figure 9: NASA-SeaWiFS image of the Northern Gulf of Mexico recorded in April, 2000. This image shows the distributions and relative concentrations of chlorophyll a, an indicator of phytoplankton biomass in this region. Note the very high concentrations (orange to red) present in the inshore regions of the mouths of the Mississippi and Atchafalaya Rivers.


There are likely to be periods when both P and N are supplied at very low levels and co-limit phytoplankton production. These periods occur during the transition from spring to summer. A similar condition is observed in large estuarine systems with a history of eutrophication, such as Chesapeake Bay (Fisher et al., 1992). Spatially, the upstream, freshwater segments of Chesapeake Bay tend to be most P limited, especially during spring runoff conditions, while the more saline down-estuarine waters tend to be most N limited. In Chesapeake Bay, the more turbid upstream freshwater component tends to exhibit interactive light and P limitation or N+P co-limitation (Fisher et al., 1992; Harding et al., 2002). Farther downstream, light limitation plays a less important role. This scenario could prove similar to the riverine-coastal continuum in the NGOM, where the most turbid upstream river plume waters are likely to exhibit the highest probability for light-nutrient interactive limitation of primary production (Lohrenz et al., 1999a, b).


While bioassay data tend to indicate P limitation during springtime in the lower salinity portions of this continuum and N and P co-limitation and N limitation in the more saline offshore waters during summer months, the bioassays do not account for sediment-water column exchange because sediments are excluded during the course of incubation. It is possible, although unlikely because of short incubation times, that sediment-water column P cycling in the shallow NGOM water column may minimize P limitation in situ. In order for this scenario to be operative, parallel N recycling would have to be far less efficient than P cycling, which numerous studies suggest is the case (Gardner et al., 1994; Bode and Dortch, 1996; Pakulski et al., 2000; Wawrik et al., 2004; Jochem et al., 2004; Cai and Lohrenz, 2005). Bioassay-based N limitation results might also be influenced by the elimination of “internal” sediment-water column N recycling, although this situation seems unlikely as well, especially if denitrification is operative (Childs et al., 2002). Sediment-based denitrification would lead to N “losses” from the system, thereby exacerbating N limitation. This influence would not be captured in bioassays, which isolate the sediments from the water column during incubation. The relatively short incubation times of bioassays probably preclude these potential artifacts. They offer a “snapshot” of nutrient limitation to complement longer-term, ecosystem-scale assessments.


The degree of N and P limitation can be calculated from bioassays, and the data can be used to create ratios of N and P limitation (Dodds et al., 2004). Interestingly, N and P limitation inferred from stoichiometric ratios of soluble (and hence biologically-available) inorganic or total N or P concentrations and inputs (loads) tends to confirm bioassay-based conclusions concerning specific nutrient limitations. For example, inshore, river-influenced waters exhibit quite high molar N:P ratios, often exceeding 50 [Nutrient Enhanced Coastal Ocean Productivity (NECOP) Reports, NOAA, 2007]. Nutrient addition bioassays initially conducted in these waters by Lohrenz et al. (1999a) and more recently by Sylvan et al. (2006), consistently revealed P limitation, especially during spring periods of maximum primary production and phytoplankton biomass accumulation. These same studies also indicated a tendency towards N and P co-limitation and exclusive N limitation during later summer months, when soluble and total N:P values dipped below 15. It should also be noted however that rates of primary production and phytoplankton biomass during this more N-limited period are at least five-fold lower than spring values, according the Gulf of Mexico NECOP data (Lohrenz et al., 1999a, 1999b). Sylvan et al. (2006) point out that P-limited spring production of “new” C may play a proportionately greater role than N-limited summer production as a source of “fuel” supporting hypoxia in the NGOM. The degree and extent to which C from this nutrient-enhanced elevated spring production is transported and accounts for summer hypoxia need to be quantified. Developing an understanding of processes that link zones and periods of high primary production and phytoplankton biomass to zones exhibiting bottom water hypoxia is a fundamentally important and challenging area of research. Such research is necessary to improve understanding of the linkage between nutrient-enhanced production and bottom water hypoxia in the NGOM. Extrapolation of C production to hypoxia data along the entire riverine-coastal shelf continuum, where zones and periods of maximum productivity and bottom water hypoxia do not necessarily coincide or overlap, depends on knowing C transport and storage (including burial), internal nutrient, and C cycling and C consumption (heterotrophic metabolism and respiration) processes along this continuum (Redalje et al., 1992; Cai and Lohrenz, 2005). Quantifying the links between locations and periods of specific nutrient limitation (or stimulation) of production and the fate of this production relative to hypoxia will contribute to long-term, effective nutrient management strategies for this region.



Key Findings and Recommendations


The SAB Panel finds that there is compelling evidence that the near shore Mississippi/Atchafalaya River plume-influenced waters are P limited and P-N co-limited during the spring periods of highest primary production. Nitrogen limitation of primary production prevails during summer periods. Recent research results indicate that the spring period of maximum primary production is P-limited in at least the plumes of the rivers, largely due to excessive N input. As a result of this man-made imbalance in nutrient loading during this crucial period, P availability plays an important role in contributing to the production of “new” organic carbon in the spring time and quite likely contributing in a major way to the “fueling” of summer hypoxia in the NGOM. However, as stressed elsewhere in this report, there is great uncertainty over the coupling in space or time of phytoplankton production and its decomposition leading to hypoxia. Therefore, a better understanding of the spatial extent and temporal patterns of these nutrient limitations is needed. The SAB Panel recommends that the following work be undertaken to advance knowledge of the importance of nutrient limitation and co-limitation as factors in the formation of Gulf hypoxia.


  • Research should be conducted to develop a more complete understanding of the spatial and temporal linkages between river plume-influenced inshore P (in spring) and/or N limited (in summer) primary production, and offshore coastal shelf, more N-limited production, as well athe fate of C produced in each zone throughout the year.




  • Research should be conducted to link near inshore river plume-influenced production in time and space to O2 depletion farther offshore. Green et al. (2006b) suggest that the small region that the central Mississippi River plume could supply is responsible for about 25% of the C necessary to fuel hypoxia. The role of the Atchafalaya plume and other riverine influenced, inshore high productivity regions in offshore hypoxia needs to be clarified.




  • Research should be conducted to address the following questions. How closely linked are the periods of high productivity and hypoxic events throughout the regions in which they occur? What is the lag between C production and its ultimate degradation?






      1. Other Limiting Factors and the Role of Si



While excessive N and P loading are implicated in eutrophication of the NGOM, these nutrients also play a role in the balance, availability and ecological manifestations of other potentially-limiting nutrients, most notably Si. In the Mississippi River plume region, N is supplied in excess of the stoichiometric nutrient ratios needed to support phytoplankton and higher plant growth (i.e., Redfield ratio, Redfield, 1958). If N over-enrichment persists for days to weeks, other nutrient limitations may, at times, result and seasonally dominate; the most obvious and important is P limitation, which has recently been demonstrated in bioassays (Ammerman and Sylvan, 2004; Sylvan et al., 2006). In addition to P limitation, N and P co-limitation and Si limitation (of diatom growth) have been observed in the fresh and brackish water components of riverine plumes that can extend more than 100 km into the receiving waters (Dortch and Whitledge, 1992; Lohrenz et al., 1999a; Dortch et al., 2001). A similar scenario is evident in the Chesapeake Bay, where elevated N loading accompanying the spring maximal freshwater runoff period increases the potential for P limitation (Fisher and Gustafson, 2004). The biogeochemical and trophic ramifications of such shifts are discussed below.


Can increased N:Si and P:Si fuel an increased microbial loop and exacerbate hypoxia?


With regard to nutrient primary production interactions, it is important to know who the dominant primary producers are, where they reside, what their contributions to new production are, and what their fate is. In NGOM waters downstream of the rivers, wetlands and intertidal regions, microalgae are by far the dominant primary producers (Lohrenz et al., 1992, 1997; Redalje et al., 1992; Rabalais et al., 1999a). The microalgal communities are dominated by phytoplankton (Redalje et al., 1994a, 1994b; Chen et al., 2000) although benthic microalgal communities can also be important sites of primary production and nutrient cycling, especially in near-shore regions (Jochem et al., 2004). As nutrient loads and limitations change over time and space, the proportions of planktonic versus benthic microalgae may also change; i.e., as nutrient inputs are reduced and planktonic primary production is reduced, the microalgal community may shift to a more benthic dominated one. This process could yield significant implications for biogeochemical (nutrients, carbon and oxygen) cycling and trophodynamics (Rizzo et al., 1992; Darrow et al., 2003).


Historic and contemporary evidence supports the contention that anthropogenically and climatically-induced changes in N and P loading have increased NGOM primary productivity and phytoplankton biomass and altered phytoplankton community composition. There are several reasons why phytoplankton community composition may have been altered by changes in nutrient loading: 1) competitive interactions among phytoplankton taxa based on varying nutrient supply rates and differing affinities for nutrient uptake and assimilation (i.e., varying nutrient uptake affinities and kinetics); 2) competitive interactions based on the relationships between nutrient supply rates and photosynthetically available light (i.e., low versus high light adapted taxa); 3) competitive interactions based on changes in N versus P supply rates (e.g., differential N versus P uptake capabilities and selection for nitrogen fixing cyanobacteria); 4) competition based on the ratios of N and P versus Si (silicious versus non-silicious taxa and heavily- versus lightly-silicified diatoms); 5) differential grazing on phytoplankton taxa (top-down controls); and 6) nutrient-salinity controls (interactive effects of changes in freshwater discharge on NGOM salinity and nutrient regimes due to climatic and watershed hydrologic control changes). Each set of controls can influence the amounts and composition of primary producers. These controls can also interact in time and space, greatly compounding and confounding the interpretation of their combined effects.


One important aspect of differential nutrient loading is the well-documented increase in N and P relative to Si loading. While N and P loads tend to reflect human activities in and alterations of the watershed, Si loads tend to reflect the mineral (bedrock and soil) composition of the watershed; a geochemical aspect that is less influenced by human watershed perturbations. Agricultural, urban and industrial development and hydrologic alterations in the MARB have led to dramatic increases in N and P relative to Si loading. In addition, the construction of reservoirs on tributaries of these river systems has further exacerbated this situation by trapping Si relative to N and P. This anthropogenic biogeochemical change has been shown to alter phytoplankton community structure (i.e., away from diatom dominance), with subsequent impacts on nutrient and carbon cycling and food web dynamics (Humborg et al., 2000; Ragueneau et al., 2006a, 2006b). The overall result has been an increase in N:Si and P:Si ratios that can influence both the amounts and composition of phytoplankton; including potential shifts from diatoms to flagellates and dinoflagellates (Turner et al., 1998; Rabalais and Turner, 2001; Justic et al., 1995). Diatoms are a highly desired food item for a variety of planktonic and benthic grazers, including key zooplankton species serving an intermediate role in the NGOM food web (Dagg, 1995). The dinoflagellates, cyanobacteria and even a few diatom species, while serving important roles in the food web, also contain species that may be toxic and/or inedible (Anderson and Garrison, 1997; Paerl and Fulton, 2006). Some of these species can rapidly proliferate or “bloom” under nutrient sufficient and enriched conditions, and thus constitute harmful algal bloom (HAB) species. Toxicity may directly and negatively impact consumers of phytoplankton as well as higher-ranked consumers, including finfish, shellfish and mammals (including humans). If non-toxic but inedible (due to size, shape, coloniality) phytoplankton taxa increase in dominance, trophic transfer may be impaired. Planktonic invertebrates, shellfish, and finfish consumers (whose diets are highly dependent on the composition and abundance of specific phytoplankton food species and groups) may then be affected (Turner et al., 1998). This could have consequences for C flux, with a relatively higher fraction of C being processed through microbial pathways (i.e., the “microbial loop”) or sedimented to the bottom. In either case, a greater fraction of the primary production would remain in the system, as opposed to being exported out of the system by transfer to higher trophic level and fisheries. The net result would be more C metabolized within the system, leading to enhanced oxygen consumption and increased hypoxia potentials.

1   ...   8   9   10   11   12   13   14   15   ...   56

Похожие:

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconBoard of Elections – Advisory Opinion 1398 – 1399

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconPanel Report on the draft Replacement London Plan

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconDraft: August 2July xxanuary x19, 20065 abstract [For presentation at the Advisory Group on apec financial System Capacity Building meeting at Seattle, February 28, 2007.]

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconDraft – do not quote or distribute

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconRanch hand advisory committee

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconMedical Devices Advisory Committee

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconPositions, training & advisory roles

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconVeterinary medicine advisory committee

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconAdvisory committee on immunization practices

11-19-07 Science Advisory Board (sab) Hypoxia Panel Draft Advisory Report Do Not Cite or Quote iconNational Vaccine Advisory Committee (nvac)

Разместите кнопку на своём сайте:
Библиотека


База данных защищена авторским правом ©lib.znate.ru 2014
обратиться к администрации
Библиотека
Главная страница