Thesis (PhD)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Soilless production of crops relies on the addition of high concentrations of nutrients with the irrigation water. The drained nutrient solution should be re-used to reduce the risk of pollution and to increase the water- and nutrient use efficiency of the system. Besides the risk of pathogen build-up, one of the main impediments of a wider application of this method is the frequent analysis required to maintain optimum nutrient concentrations and ratios in the rootzone. Yield reductions may be caused by an unbalanced nutrient solution. Alternatively the addition level of nutrients can be calculated through the use of nutrient uptake models that simulate the change in the re-circulated nutrient solution. To simulate crop water and nutrient demand necessary for model based regulation it was necessary to quantify the key factors affecting nutrient uptake by plants. The nutrient solution concentration and ratios between the macro-nutrients affected the uptake of water and nutrients. The total nutrient uptake per root dry weight increased and more specifically the nitrate (NO3 -), phosphate (H2PO4 -), potassium (K+) and sulphate (SO4 2-) uptake increased with an increase in nutrient solution electrical conductivity (EC) from 0.8 to 4.0 mS cm-1 while water uptake decreased. Except for Ca2+ uptake there was no correlation between nutrient and water uptake. Nutrient uptake can thus not be calculated based on water uptake. Instead a mechanistic high-affinity Michaelis-Menten based model can be used to estimate macro-nutrient uptake (Un, mg m-2 hr-1). Water and nutrient uptake was also affected by the solar radiation levels. Since nutrient uptake is related to the growth rate, solar radiation levels can be expected to influence nutrient uptake. The uptake of all ions increased with an increase in the solar radiation levels and for NO3 -, K+ and H2PO4 - the uptake rate was higher at higher nutrient solution concentrations. The Michaelis-Menten based model was adjusted to incorporate the effect of solar radiation levels on nutrient uptake. Water uptake (Wu, L m-2 day-1) was simulated as a function of crop transpiration and crop leaf area using a linear regression model, but since leaf area development was affected by solar radiation levels this was additionally incorporated into the estimation of the leaf area index (LAI). The composition of the nutrient solution also affected the biomass allocation of the crop which can again affect nutrient use as well as the fruit yield. There was also a direct effect of nutrient solution composition on fruit yield and quality with higher EC’s resulting in smaller fruit but an increase in fruit dry matter %, total soluble solids (TSS), titratable acidity (TA) and lycopene content. The results in this thesis make a valuable contribution to our understanding of the effect of nutrient availability (concentration and ratios) and nutrient requirement for growth (solar radiation levels) on nutrient uptake. Incorporating these into nutrient uptake models resulted in the development of a handy tool to simulate changes in composition of re-circulating nutrient solutions ultimately resulting in an improvement of the water and nutrient use efficiency of soilless systems. / AFRIKAANSE OPSOMMING: Die grondlose verbouing van gewasse is afhanklik van toediening van voedingselemente teen hoë peile in die besproeiingswater. Die voedingsoplossing wat dreineer moet hergebruik word om die risiko van besoedeling te verminder en ook om die water en nutriënt verbruik doeltreffendheid van die sisteem te verbeter. ʼn Ongebalanseerde voedingsoplossing kan ʼn verlaging in opbrengste veroorsaak. Benewens die risiko van patogene wat opbou, is die gereelde analises nodig word vir die handhawing van optimale nutriënt konsentrasies en verhouding tussen elemente in die wortelsone een van die hoof faktore wat ʼn meer algemene gebruik van die metode verhoed. Alternatiewelik kan die nutriënt toedieningspeile bereken word deur voedingstof opname modelle en simulasie van die verandering in water en nutriente wat dreineer. Om ʼn model gebaseerde reguleringsmetode daar te stel was dit nodig om die belangrikste faktore wat nutriënt opname beïnvloed te kwantifiseer. Beide die konsentrasie van die voedingsoplossing en die verhouding tussen elemente het ‘n effek gehad op die opname van water en nutriënte. Die totale nutriënt opname per wortel droë massa het toegeneem. Terwyl water opname afgeneem het met ‘n toename in die elektriese geleding (EG) van die voedingsoplossing vanaf 0.8 tot 4.0 mS cm-1 het die nitraat (NO3 -), fosfaat (H2PO4 -), kalium (K+) en sulfaat (SO4 2-) opname verhoog. Behalwe vir Ca2+ opname was daar geen korrelasie tussen water en nutriënt opname nie. Nutriënt opname kan dus nie bepaal word gebaseer op wateropname nie. Alternatiewelik is die gebruik van ʼn meganistiese hoë-affiniteit Michaelis-Menten-gebaseerde model voorgestel om die opname van makro-nutriente (Un, mg m-2 hr-1) te bepaal. Water- en voedingstofopname is beinvloed deur die ligintensiteit vlakke. Voedingsopname word bepaal deur die groei van die plant, daarom is dit verwag dat ligintensiteit vlakke die opname van voedingstowwe sal beïnvloed. Die opname van al die ione het toegeneem met 'n toename in die ligintensiteit vlakke en die tempo van NO3 -, K+ en H2PO4 - opname was hoër by 'n hoër voedingsoplossing konsentrasie. Die Michaelis-Menten gebaseerde model is aangepas om die effek van ligintensiteit vlakke op nutriënt opname te inkorporeer. Opname van water (Wu, L m-2 dag-1) is gesimuleer as 'n funksie van transpirasie en blaaroppervlakte met behulp van 'n lineêre regressiemodel en aangesien die blaaroppervlak ontwikkeling ook deur ligintensiteit vlakke beïnvloed word, is dit opgeneem in die skatting van die blaaroppervlakte-indeks (LAI). Die samestelling van die voedingsoplossing het die biomassa verspreiding beïnvloed. Dit kan nutriënt gebruik en vrug opbrengs beïnvloed. Die voedingsoplossing samestelling het vrug opbrengs en - kwaliteit beinvloed met kleiner vrugte, maar 'n toename in droëmateriaal %, totale oplosbare vastestowwe (TOVS), titreerbare suur (TA) en likopeen inhoud by ʼn hoër EG. Die resultate in hierdie tesis lewer 'n waardevolle bydrae tot ons begrip van die effek van nutriënt beskikbaarheid (konsentrasie en verhoudings) en voedingstof behoefte vir groei (ligintensiteit vlakke) op voedingsopname. Deur die inligting te inkorporeer in voedingsopname modelle het gelei tot die ontwikkeling van 'n handige instrument om die veranderinge in die samestelling van hersirkulerende voedingsoplossings te simuleer. Dit lei gevolglik tot die verbetering van die water en voedingstof gebruik doeltreffendheid van grondlose stelsels.

In order to predict how terrestrial ecosystems will respond to global change, there is growing recognition that we need to better understand linkages between plant and soil processes. Previously the factors and processes with potential to influence the terrestrial carbon (C) cycle have been investigated in isolation from each other. This study investigated the interactions of nutrient availability and warming in controlling the soil carbon dynamics, with regards to the fate of already sequestered carbon in soil, under conditions of increasing atmospheric temperatures. The project objectives were addressed by three independent experiments designed to explain specific components of the carbon-nutrient cycle interactions, and the findings brought together to describe the implications for future soil carbon storage. The main measurements collected throughout this project included soil carbon dioxide (CO2) fluxes, partitioned into autotrophic and heterotrophic components, net ecosystem exchange and respiration fluxes, and background soil moisture and temperature data, backed by gas, soil and biomass analyses. In the two field experiments, these measurements were taken from plots with or without any inorganic nutrient additions or in the presence or absence of legumes providing biological nitrogen addition to the ecosystem. In the laboratory, temperature and nutrient availability were manipulated within the ecosystem. The reduction in decomposition rates, without reduction of productivity as a result of inorganic nutrient additions, indicated the potential for increasing C storage. There was also evidence that nutrient availability controls the strength of the link between plant and soil processes in semi-natural grasslands. The yields, decomposition rates and soil C fluxes recorded in the presence and absence of legumes provided some evidence of N2 fixation, improving ecosystem productivity and soil properties while reducing soil C effluxes, in a managed grassland. In the laboratory, the warming of soils from lysimeters with and without plants, receiving or not receiving fertiliser, supported the findings from field experiments regarding the importance of the soil-plant link in controlling C fluxes. However, C stocks and δ13C analyses showed that over a year’s worth of warming and nutrient manipulations made little difference to the amount of C stored in the soil, indicating that edaphic factors have greater control over the response of C dynamics to increased temperatures.

Florestas tropicais tem um papel relevante nos ciclos globais do carbono e nitrogênio. A Floresta Atlântica é um bioma tropical ameaçado da América do Sul, de imensa complexidade estrutural, mas ainda pouco investigado quanto ao seu funcionamento. Neste contexto, o objetivo principal desse estudo foi investigar a estrutura da floresta através da quantificação da biomassa viva acima do solo e da queda anual de folhas. Paralelamente, as concentrações de carbono e nitrogênio foram determinadas no solo, e na vegetação, permitindo a determinação de estoques e fluxos desses elementos em áreas de diferentes fitofisionomias localizadas nas escarpas da Serra do Mar, na costa nordeste do Estado de Santa Catarina, sul do Brasil. As áreas de estudo localizam-se no município de Joinville, onde a temperatura média anual é de 20°C e a precipitação média acumulada anualmente é de aproximadamente 2200 mm. Os sítios amostrais foram estabelecidos em elevações de 200 m anm (Submontana) e 800 m anm (Montana) Em cada área, quatro parcelas de aproximadamente 0,25 ha foram delimitadas, e, então, foram determinadas a estrutura florestal, biomassa acima e abaixo do solo (viva e morta) e produção de serrapilheira. Concentração e composição isotópica de carbono e nitrogênio em cada compartimento (folhas, tronco, serrapilheira e solo) foram determinadas e usadas para converter biomassa em estoques e fluxos desses elementos, e compreender as complexas interações dentro desses ecossistemas. A AGLB foi de aproximadamente 300 Mg.ha-1 na floresta Submontana e 380 Mg.ha-1 na floresta Montana, AGDB, por sua vez, foi de 6,8 e 6,6 Mgoha-1 e BGB (raízes) foi de 54 e 66 Mg.ha -1, respectivamente. Não foi encontrada diferença significativa entre a produção de serrapilheira entre as fitofisionomias, sendo encontrada uma produção média de 6,4 Mg.ha-1 ano-1 na floresta Submontana e 6,9 Mg.ha-1 ano-1 na Montana. As folhas foram o componente majoritário da serrapilheira, contribuindo com ~68% em ambas as áreas. Os estoques totais de carbono e nitrogênio foram significativamente maiores na fitofisionomia Montana, aumentando de 208 MgC.ha-1 para 390 MgC.ha-1 e de 7 MgN.ha-1 para 16 MgN.ha-1, respectivamente. Fluxos de carbono e nitrogênio via serrapilheira nas florestas Submontana e Montana foram, respectivamente, igual a 3,0 e 3,2 Mg.ha-1 e de 0,12 e 0,14 Mg.ha-1. Valores de ?13C em cada compartimento foram similares entre as áreas, e típico de plantas que seguem o ciclo fotossintético C3. Conforme esperado, um enriquecimento em 13C e 15N ao longo do perfil do solo foi encontrado em ambas as áreas. Valores de ?15N foram maiores em folhas, serrapilheira e solo na floresta Submontana, possivelmente devido à maior disponibilidade de nitrogênio neste sítio; e/ou ao baixo tempo de residência do solo e serrapilheira empobrecida em 15N no sítio Montana. De maneira geral, posições latitudinal (em termos de disponibilidade de luz) e altitudinal (em termos de precipitação e temperatura) foram determinantes na composição, alocação e dinâmica do carbono e nitrogênio nestes ecossistemas / Tropical evergreen forests have a key role in the global carbon and nitrogen cycles. The Atlantic Forest is a vanishing South American tropical biome of immense structural complexity. The structure and functioning of these forests are relatively unknown. In this context, the main objective of this study was to investigate the forest structure by estimating aboveground live biomass (AGLB), belowground biomass (BGB) and litterfall. The determination of carbon and nitrogen concentrations in soils and vegetation allowed to quantify stocks and fluxes of these two elements. Four 0.25 ha-plots were established in two elevations (200 m asl and 800 m asl) in forests located on the slopes of the Serra do Mar, on the northeast coast of the State of Santa Catarina, southern Brazil. The sampling areas are located in the municipality of Joinville. The historical average annual temperature is 20°C and rainfall is approximately 2200 mm. The AGLB varied along the sites from approximately 300 Mg.ha-1 (submontane) to 380 Mgoha-1 (montane), AGDB varied from 6.8 to 6.6 Mgoha-1 and BGB (roots) varied from 54 to 66 Mg.ha -1, respectively. . The average litterfall production of 6.4 Mg.ha-1 ano-1 and 6.9 Mg.ha-1 ano-1 were found in the submontante and montane study sites, respectively. Leaves were the major component of litterfall contributing ~68% in both sites. Total carbon and nitrogen stocks were higher in the montane site. Total carbon stock (AGLB + BGB) increased from 208 Mg.ha-1 in the submontane site to 390 Mg.ha-1 in the montane. While, total nitrogen stock increased from 7 Mg.ha-1 to 16 Mg.ha-1, respectively. Fluxes of carbon and nitrogen via litterfall in the submontane and montane site varied from 3.0 to 3.2 Mg.ha-1 and from 0.12 to 0.14 Mg.ha-1, respectively. ?13C values in each compartment were similar between the two sites, and representative of C3 plants. As expected, there was an enrichment of ?13C and ?15N values in depth in each studied site. ?15N values where higher in leaves, litter and soil of the submontane site, possibly due higher nitrogen availability in this site; or low residence time of soil and depleted 15N in litterfall of the montane site. Overall, latitutinal (in terms of light availability) and altitudinal (in tems of precipitation and temperature) position were determinant in the nitrogen and carbon composition, allocation and dynamics in these ecosystems

ABSTRACT SOIL AND WATER ANALYSIS TECHNIQUES FOR AGRICULTURAL PRODUCTION Maral, Nuh M. Sc., Department of Chemistry Supervisor: Prof. Dr. G. inci G&ouml / kmen May 2010, 108 pages In Turkey, usage of increasing amounts of fertilizers and pesticides by some unconscious farmers cause soil pollution and soil infertility for the crop production. Usage of water in excessive amounts and/or in poor quality for irrigation creates problems during the plant production. So in this study, soil and water samples were analyzed by using simple and reliable techniques for the soil and water quality in laboratories of METU and Soil Fertilizer and Water Resources Central Research Institute Laboratory in Ankara. The soil and water samples were collected using the standard techniques from Ankara, Bolu, &Ccedil / orum and Kirikkale. According to the soil test results, the textures of the soil samples are found as loam and clay loam. The total salt content of the soil samples are between 0.033 &ndash / 0.063 % (w/w), meaning they are low salinity soils (total salt less than 0.15 % w/w). The pH of the soil samples are between 7.86&ndash / 8.15, they are slightly alkaline. The phosphorus concentrations of soil samples are in a range 4.95 to 35.45 P2O5 kg/da. Some of the soil samples have too high phosphorus content (greater than 12 P2O5 kg/da). The potassium content of soil samples are found between 141&ndash / 286 K2O kg/da, so the soil is efficient for crop production. Lime content of the soil samples is between 1.04&ndash / 2.67 % (w/w) CaCO3. It means all of the soil samples are calcareous but it is not too high for the agricultural production. Organic matter content of soil samples are found between 0.83&ndash / 2.04 % (w/w). This means the soils are limited in their organic matter content for the crop production. Analysis of 22 water samples yielded EC values between 0.384 &ndash / 1.875 dS/m. Water samples have moderate to high-salinity (if EC values between 0.205 and 2.250 dS/m), yet these can be used for the irrigation of the crops. pH values of water samples are found between 7.18-8.10, meaning that they are slightly alkaline. Bicarbonate concentrations of 19 of the water samples are greater than 200 mg/L. These waters may not be suitable for irrigation of ornamental plants. All of the water samples, except water samples from G&ouml / lbaSi, have sodium absorption ratio (SAR) values between 1 and 9. Water samples with low SAR values, except water samples from G&ouml / lbaSi, can be used for irrigation of almost all soils with little danger of developing harmful levels of sodium. The Residual Sodium Carbonate (RSC) values of water samples Ankara G&ouml / lbaSi and Sincan-1 are greater than 2.50 meq/L and these water samples are not suitable for the irrigation. RSC values of Etimesgut, Sincan-2 and Kazan water samples are positive and lower than the value 2.00 meq /L. All the other water samples have negative RSC values so they are the safe to use for irrigation. It has been observed that development of practical field analysis techniques for all soil and water quality parameters may be possible with exception of micronutrient determination. For determining soil and water quality parameters in the rural areas there is a need to establish a small laboratory with necessary equipment and apparatus and training one or two farmers. With the experience gained in this study, some of these techniques may be adapted to the rural field applications, so soil and water may be tested by the farmers for better yields.

In British Columbia, humus form identification is widely use to infer the level of plant-available soil nutrients. This identification is based on field-observable (morphological) features. We recognize three major humu forms: Mors, Moders, and Mulls - which are differentiated according to the type of F horizon, and the presence/absence of organic matter-enriched Ah horizons. Mors represent humus forms where decomposition is dominated by fungi, with slow decomposition rates and accumulations of organic matter on the soil surface. Mors are characterized by the presence of a Fm (m - mycogenous) horizon. In contrast, Mulls represent humus forms with high rates of decomposition and faunal activity resulting in organic matter being intimately incorporated into the upper mineral soil layer instead of accumulating on its surface. Intermediate on the humus form gradient from Mors to Mull are the Moders. Moders are similar to Mors in that they have accumulations of organic matter on the surface of the mineral soil but decomposition is not fungus dominated, so they lack the diagnostic Fm horizon. The central concept of the Moder is represented by the Leptomoder, which is characterized by a Fz (z - zoogenous) horizon with an active population of soil meso- and microfauna, fungal mycelia are not present or present in small amounts. When both fungal mycelia and faunal droppings can be found, but neither clearly predominate over the other, an Fa (a - amphimorphic) horizon results. Fa horizons are characteristic of Mormoders, an integrade between Mors and Moders. Considering the prevalence of Mormoders and Leptomoders in B.C. and the difficulties in identifying Fa horizons, the aim of this study was to determine whether the morphological features used to differentiate these two Moder humus forms reflects differences in their physical, chemical and biotic properties.

Florestas tropicais tem um papel relevante nos ciclos globais do carbono e nitrogênio. A Floresta Atlântica é um bioma tropical ameaçado da América do Sul, de imensa complexidade estrutural, mas ainda pouco investigado quanto ao seu funcionamento. Neste contexto, o objetivo principal desse estudo foi investigar a estrutura da floresta através da quantificação da biomassa viva acima do solo e da queda anual de folhas. Paralelamente, as concentrações de carbono e nitrogênio foram determinadas no solo, e na vegetação, permitindo a determinação de estoques e fluxos desses elementos em áreas de diferentes fitofisionomias localizadas nas escarpas da Serra do Mar, na costa nordeste do Estado de Santa Catarina, sul do Brasil. As áreas de estudo localizam-se no município de Joinville, onde a temperatura média anual é de 20°C e a precipitação média acumulada anualmente é de aproximadamente 2200 mm. Os sítios amostrais foram estabelecidos em elevações de 200 m anm (Submontana) e 800 m anm (Montana) Em cada área, quatro parcelas de aproximadamente 0,25 ha foram delimitadas, e, então, foram determinadas a estrutura florestal, biomassa acima e abaixo do solo (viva e morta) e produção de serrapilheira. Concentração e composição isotópica de carbono e nitrogênio em cada compartimento (folhas, tronco, serrapilheira e solo) foram determinadas e usadas para converter biomassa em estoques e fluxos desses elementos, e compreender as complexas interações dentro desses ecossistemas. A AGLB foi de aproximadamente 300 Mg.ha-1 na floresta Submontana e 380 Mg.ha-1 na floresta Montana, AGDB, por sua vez, foi de 6,8 e 6,6 Mgoha-1 e BGB (raízes) foi de 54 e 66 Mg.ha -1, respectivamente. Não foi encontrada diferença significativa entre a produção de serrapilheira entre as fitofisionomias, sendo encontrada uma produção média de 6,4 Mg.ha-1 ano-1 na floresta Submontana e 6,9 Mg.ha-1 ano-1 na Montana. As folhas foram o componente majoritário da serrapilheira, contribuindo com ~68% em ambas as áreas. Os estoques totais de carbono e nitrogênio foram significativamente maiores na fitofisionomia Montana, aumentando de 208 MgC.ha-1 para 390 MgC.ha-1 e de 7 MgN.ha-1 para 16 MgN.ha-1, respectivamente. Fluxos de carbono e nitrogênio via serrapilheira nas florestas Submontana e Montana foram, respectivamente, igual a 3,0 e 3,2 Mg.ha-1 e de 0,12 e 0,14 Mg.ha-1. Valores de ?13C em cada compartimento foram similares entre as áreas, e típico de plantas que seguem o ciclo fotossintético C3. Conforme esperado, um enriquecimento em 13C e 15N ao longo do perfil do solo foi encontrado em ambas as áreas. Valores de ?15N foram maiores em folhas, serrapilheira e solo na floresta Submontana, possivelmente devido à maior disponibilidade de nitrogênio neste sítio; e/ou ao baixo tempo de residência do solo e serrapilheira empobrecida em 15N no sítio Montana. De maneira geral, posições latitudinal (em termos de disponibilidade de luz) e altitudinal (em termos de precipitação e temperatura) foram determinantes na composição, alocação e dinâmica do carbono e nitrogênio nestes ecossistemas / Tropical evergreen forests have a key role in the global carbon and nitrogen cycles. The Atlantic Forest is a vanishing South American tropical biome of immense structural complexity. The structure and functioning of these forests are relatively unknown. In this context, the main objective of this study was to investigate the forest structure by estimating aboveground live biomass (AGLB), belowground biomass (BGB) and litterfall. The determination of carbon and nitrogen concentrations in soils and vegetation allowed to quantify stocks and fluxes of these two elements. Four 0.25 ha-plots were established in two elevations (200 m asl and 800 m asl) in forests located on the slopes of the Serra do Mar, on the northeast coast of the State of Santa Catarina, southern Brazil. The sampling areas are located in the municipality of Joinville. The historical average annual temperature is 20°C and rainfall is approximately 2200 mm. The AGLB varied along the sites from approximately 300 Mg.ha-1 (submontane) to 380 Mgoha-1 (montane), AGDB varied from 6.8 to 6.6 Mgoha-1 and BGB (roots) varied from 54 to 66 Mg.ha -1, respectively. . The average litterfall production of 6.4 Mg.ha-1 ano-1 and 6.9 Mg.ha-1 ano-1 were found in the submontante and montane study sites, respectively. Leaves were the major component of litterfall contributing ~68% in both sites. Total carbon and nitrogen stocks were higher in the montane site. Total carbon stock (AGLB + BGB) increased from 208 Mg.ha-1 in the submontane site to 390 Mg.ha-1 in the montane. While, total nitrogen stock increased from 7 Mg.ha-1 to 16 Mg.ha-1, respectively. Fluxes of carbon and nitrogen via litterfall in the submontane and montane site varied from 3.0 to 3.2 Mg.ha-1 and from 0.12 to 0.14 Mg.ha-1, respectively. ?13C values in each compartment were similar between the two sites, and representative of C3 plants. As expected, there was an enrichment of ?13C and ?15N values in depth in each studied site. ?15N values where higher in leaves, litter and soil of the submontane site, possibly due higher nitrogen availability in this site; or low residence time of soil and depleted 15N in litterfall of the montane site. Overall, latitutinal (in terms of light availability) and altitudinal (in tems of precipitation and temperature) position were determinant in the nitrogen and carbon composition, allocation and dynamics in these ecosystems

Organism succession during ecosystem development has been well studied for aboveground plant communities while the associated pattern of change in microbial communities remains largely unknown. A study was conducted along developmental sand-dune chronosequences bordering Lake Michigan at Wilderness State Park and Altamaha river valley of southeast Georgia with the hypothesis that soil bacterial communities will follow a pattern of change that is associated with soil, plant, and ecosystem development. This study site included 5 replicate sites along 14 dunes ranging in age from 105 to 500,000 years since deposition. The microbial composition and diversity in the soil was studied using bacterial tag-encoded FLX amplicon pyrosequencing of the 16S rRNA gene. As hypothesized, Bray-Curtis ordination indicated that bacterial community assembly changed along the developmental gradient at both sites. However, there was no seasonal effect at Michigan sites despite likely differences in plant carbon inputs. At the Michigan site, soil Ca, Mg levels and pH showed a significant log-linear correlation with soil development (r = 0.83, 0.84 and 0.81, respectively). Bacterial diversity represented by Simpson’s reciprocal index (Simpson’s 1/D) showed a steady decline from the youngest to the oldest dunes with the largest decline (212 to 58) during the initial stages of soil development (105 to 450 years). The change in plant species abundance was higher in the youngest sites than the older sites. This change was significantly correlated with the change in microbial community distribution (p < 0.0001; r = 0.56). Similarly, at Georgia sites, soil development showed significant log-linear correlation with soil base cations (Ca and Mg) (r = 0.93and 0.95). However, diversity indices and PLFA failed to show any particular change in trend across the developing chronosequences. When the results from both sites were used to study bacterial spatial patterns, local geochemical features were found to be a dominant factor in driving bacterial community structure, while geographic distance as a single factor could contribute to some community variation at a scale (50 – 1700 km). The results suggest that soil nutrients and plant community could be a strong driving force in shaping microbial community assembly across a developing soil ecosystem.

No description available.

Increased floodplain and wetland restoration activity has raised concerns about potential impacts on the release of greenhouse gases (GHGs) to the atmosphere due to restored connectivity between aquatic and terrestrial ecosystems. Research has shown GHG fluxes from hydrologically active landscapes such as floodplains and wetlands vary spatially and temporally in response to primary controls including soil moisture, soil temperature, and available nutrients. In this study, we performed a semimonthly sampling campaign measuring GHG (CO2, CH4, and N2O) fluxes from six locations within a third-order stream floodplain. Site locations were based on dominant landscape positions and hydrologic activity along a topographic gradient including a constructed inset floodplain at the stream margin, the natural levee, an active slough, the general vegetated floodplain, a convergence zone fed by groundwater, and the upland area. Flux measurements were compared to abiotic controls on GHG production to determine the most significant factors affecting GHG flux from the floodplain. We found correlations between CO2 flux and soil temperature, organic matter content, and soil moisture, CH4 flux and pH, bulk density, inundation period length, soil temperature, and organic matter content. But minimal correlations between N2O flux and the measured variables. Spatially, our results demonstrate that constructed inset floodplains have higher global warming potential in the form of CH4 than any other site and for all other GHGs, potentially offsetting the positive benefits incurred by enhanced connectivity. However, at the reach scale, total CO2 flux from the soil remains the greater influence on climate since the area covered by these inset floodplains is comparatively much smaller than the rest of the floodplain. / Master of Science

Made available in DSpace on 2016-04-28T14:58:59Z (GMT). No. of bitstreams: 1 2007- Jose Eduardo Dias.pdf: 7576584 bytes, checksum: f462c307c38711b0d35ca1237ce04396 (MD5) Previous issue date: 2007-05-18 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / Agroecological production systems are more complex than conventional ones, since a series of crops follow in sequence or are intercropped to achieve high diversity and maximize soil nutrient recycling. The full understanding of these systems require systemic analyses and therefore, adequate tools. To understand soil fertility evolution, heterogeneity should also be considered. This thesis was performed at Integrated Agroecological Production System, located at Serop?dica, Rio de Janeiro, an area that has been organically managed for 14 years. The first chapter consisted of studying land use from 2003 to 2005. Each land unit use was annotated every three months to generate 12 maps. The information gathered was processed to generate three maps: (i) cultivated species richness; (ii) fallow intensity and (iii) green manure use. High cultivated species richness was observed. In some land units up to 40 plant species were recorded. However, this diversity was not uniformly distributed throughout the terrain. A high intensity of land use, mostly with annuals was also observed in a large part of the area. In most cases, fallow periods were up to 3 months in 3 years. The use of legumes was less frequent on the land units used for annual crops. They were not uniformly distributed throughout the terrain. The objectives of the second chapter were: (i) to generate maps of spatial distribution of soil nutrients (Ca++, Mg++ K++ and P), pH and organic matter; at two soil depths 0-20 and 20-40 cm, in 2004 and 2005 (ii) evaluate the difference of those soil attributes in two years (iii) evaluate the difference of soil fertility between the layers of 20-40 cm and 0-20 cm, (iv) generate two maps of fertility potential , one for each year, for the 0-20 cm layer; (v) to combine maps of all attributes in order to generate maps of potential of gain/loss of nutrients . In general, soil nutrient levels were high and very high, indicating that the 14 years of organic management has built up soil fertility. There was a trend of decreasing soil fertility from 2004 to 2005. The area with higher loss potential coincided with a high intensively used portion of the system. However, on the area used as a control, losses also occurred, in lower intensity. From all nutrients studied, phosphorus was the most spatially and temporally variable. / Sistemas de produ??o com base agroecol?gica s?o mais complexos que os convencionais. A heterogeneidade ? uma regra nesses sistemas, pois muitas culturas se sucedem no tempo e no espa?o, compondo assim, alta biodiversidade. A compreens?o desses sistemas requer an?lises de car?ter sist?mico e, portanto, necessita de ferramentas compat?veis. Da mesma forma, estudos de fertilidade do solo, que consideram a heterogeneidade espacial e temporal s?o imperativos para se compreender a din?mica dos nutrientes. O presente estudo foi conduzido no Sistema Integrado de Produ??o Agroecol?gica - SIPA, uma ?rea que vem sendo manejada organicamente por catorze anos, situada no munic?pio de Serop?dica RJ. O primeiro cap?tulo constou do estudo do uso da terra, entre 2003 e 2005. Foram feitos registros trimestrais do uso das glebas, gerando doze mapas de uso. As informa??es contidas nesses mapas foram processadas de forma a gerar tr?s novos mapas: (a) riqueza de esp?cies cultivadas; (b) intens idade de pousio e (3) utiliza??o de leguminosas. Foi verificada alta riqueza de esp?cies cultivadas, havendo algumas glebas com at? 40 esp?cies vegetais. Por?m, esta diversidade encontra-se desuniformente distribu?da no terreno. Houve uma alta intensidade de uso da terra, sendo o per?odo de pousio, em grande parte, relativamente curto, ou seja, de at? tr?s meses, em tr?s anos. Nesses casos, predominaram cultivos anuais. O uso de leguminosas, no sistema, foi menor nas glebas de cultivos anuais e estava desuniformemente distribu?do no terreno. O segundo cap?tulo teve como objetivos: (1) gerar mapas da distribui??o espacial de MO, pH. Ca++, troc?vel, Mg++ troc?vel e K++ troc?vel em quatro glebas selecionadas do SIPA, para as camadas de 0 a 20 e de 20 a 40 cm, nos anos 2004 e 2005; (2) avaliar a evolu??o dos teores daqueles elementos nos dois anos; (3) avaliar a diferen?a de teores entre as camadas 0- 20 cm e 20-40 cm; (4) gerar dois mapas de fertilidade potencial qu?mica , um para cada ano, combinando-se os mapas de todos os nutrientes, para a camada de 0 a 20 cm e (5) gerar dois mapas de potencial de perda/ac?mulo de nutrientes, um, pela combina??o dos mapas de evolu??o da fertilidade e outro, pela combina??o dos mapas de potencial de 2005 e 2004. Os resultados indicaram n?veis dos nutrientes, na faixa de alto a muito alto, na maioria da extens?o da ?rea de estudo, nos dois anos, indicando que o manejo do SIPA, nos catorze anos, desde a sua cria??o, tem constru?do a fertilidade do solo. No entanto, detectou-se uma tend?ncia de diminui??o dos n?veis de nutrientes de 2004 para 2005. O potencial de perda coincide com ?rea de alta intensidade de uso, entretanto a ?rea controle, de uso m?nimo, tamb?m indicou perdas, contudo em propor??es menores.