Modelling CO2 and energy exchanges in a northern semiarid grassland using the carbon- and nitrogen-coupled Canadian Land Surface Scheme (C-CLASS)

The development of carbon (C) and nitrogen (N) simulations is one of the ongoing efforts in the land surface schemes of climate models. The C- and N-coupled Canadian Land Surface Scheme (C-CLASS) was recently modified to better represent grassland ecosystems. Improvements include revised plant growth and senescence calculations that are driven by the plant C balance between fixation and respiration, and leaf-out and leaf-fall schemes that are regulated by the seasonal dynamics of C and N reserves. These revisions were developed to better simulate the stress-related senescence and regrowth of perennials. The model was tested with observations of surface carbon and energy fluxes, soil temperature and moisture, and plant growth during 3 years of declining precipitation at a northern semiarid grassland near Lethbridge, Alberta, Canada. The R² and standard deviations between the simulated and observed half-hourly fluxes were 0.95 and 22.5 W m⁻² for net radiation, 0.82 and 42.1 W m⁻² for sensible heat, 0.66 and 29.2 W m⁻² for latent heat, and 0.63 and 0.95 μmol C m⁻² s⁻¹ for net CO₂ exchange. The model and observations both showed a strong impact of declining precipitation on annual carbon budgets in this semi-arid grassland. In a wet year (1998, precipitation = 482 mm), the ecosystem acted as a strong C sink (92 g C m⁻² modelled and 109 g C m⁻² measured from June 20th to December 31st). In a near-normal year (1999, precipitation = 341 mm), a smaller C sink was indicated (24 g C m⁻² modelled and 21 g C m⁻² measured). In a dry year (2000, precipitation = 276 mm), the ecosystem acted as a small C source (−18 g C m⁻² modelled and −17 g C m⁻² measured).     

Three-dimensional modeling of the lower trophic levels in the Ria de Aveiro (Portugal)

The water and the ecosystem dynamics of the Ria de Aveiro, a shallow, multi-branch lagoon located on the northwest coast of Portugal, are simulated using a new fully coupled 3D modeling system. This model couples the hydrodynamic model SELFE (semi-implicit Eulerian-Lagrangian finite element) and an ecological model extended from EcoSim 2.0 to represent zooplankton dynamics. The model application is based on an unstructured grid spatial discretization, which is particularly appropriate for this system given its complex geometry. The baroclinic circulation is calibrated and validated for different environmental conditions, leading to velocity errors smaller than 5 cm/s across the lagoon. Ecological simulations, focused on zooplankton dynamics represented by a site-specific formulation, are then presented and compared against field data for two contrasting environmental conditions: Autumn 2000 and Spring 2001. Results show that the fully coupled model is able to reproduce the dynamics of the ecosystem in the Spring 2001, fitting the model results inside the range of data variation. During this period zooplankton differences between data and model results are of about 0.005 mg C/l (60%), while other ecological tracers’ differences are generally smaller than 20–30% along the several branches of the lagoon. In the Autumn 2000, the model tends to overestimate zooplankton by a factor of 10 and to underestimate phytoplankton and ammonium, with discrepancies of about 0.1 mg C/l and 4.8 μmol N/l, respectively. Factors like the ecological conditions imposed at the boundaries, the input parameters of the ecological model and the simplification of the ecosystem structure, since phytoplankton is the only primary producer considered, may explain the observed differences.     

Crop growth,soil water and nitrogen balance simulation on three experimental field plots using the Opus model—A case study

Over a period of 5 years, the agro-ecosystem model Opus was used to simulate soil water and nitrogen balance as well as crop growth for three experimental field plots. At these plots, different agricultural management practices were applied. The data set obtained from these plots consists of automatically recorded time series of daily volumetric soil water contents measured by TRIME-probes as well as daily pressure heads measured by tensiometer. Aboveground total biomass, yield, nitrogen-uptake by crops as well as nitrate contents in the soil were measured at 6–10 sample times per year. The objective of this study was an evaluation of the accuracy of Opus regarding the simulation of crop growth, soil water and nitrogen balance. The simulations of soil water contents and pressure heads correspond with the commonly measured trends in soil depths shallower than 60 cm. In depths deeper than 60 cm, some differences between measured and simulated soil water contents as well as pressure heads could be observed. Nitrate contents in the root zone and the aboveground total biomass were simulated satisfactorily. In contrast to that, simulated and observed yields show greater discrepancies. This indicates the need of a site specific calibration of crop growth parameters within the Opus model.     

Modelling seasonal dynamics of biomasses and nitrogen contents in a seagrass meadow (Zostera noltii Hornem.): application to the Thau lagoon (French Mediterranean coast)

Anumerical deterministic model for a seagrass ecosystem (Zostera noltii meadows) has been developed for the Thau lagoon. It involves both above- and belowground seagrass biomasses, nitrogen quotas and epiphytes. Driving variables are light intensity, wind speed, rain data and water temperature. This seagrass model has been coupled to another biological model in order to simulate the relative contributions of each primary producer to: (i) the total ecosystem production, (ii) the impact on inorganic nitrogen and (iii) the fluxes towards the detritus compartment. As a first step in the modelling of seagrass beds in the Thau lagoon, the model has a vertical structure based on four boxes (a water box on top of three sediment boxes) and the horizontal variability is neglected until now. This simple box structure is nevertheless representative for the shallow depth Z. noltii meadows, spread over large areas at the lagoon periphery.After calibration, simulation results have been compared with in situ measurements and have shown that the model is able to reproduce the general pattern of biomasses and nitrogen contents seasonal dynamics. Moreover, results show that, in such shallow ecosystems, seagrasses remain the most productive compartment when compared with epiphytes or phytoplankton productions, and that seagrasses, probably due to their ability in taking nutrients in the sediment, have a lower impact on nutrient concentration in the water column than the phytoplankton. Furthermore, in spite of active mechanisms of internal nitrogen redistribution and reclamation, the occurrence of a nitrogen limitation of the seagrass growth during summer, already mentioned in the literature, have also been pointed out by the model. Finally, simulations seems to point out that epiphytes and phytoplankton could compete for nitrogen in the water column, while a competition for light resources seems to be more likely between epiphytes and seagrasses.     

Nitrogen controls on ecosystem carbon sequestration: a model implementation and application to Saskatchewan,Canada

A plant–soil nitrogen (N) cycling model was developed and incorporated into the Integrated BIosphere Simulator (IBIS) of Foley et al. [Foley, J.A., Prentice, I.C., Ramankutty, N., Levis, S., Pollard, D., Sitch, S., Haxeltine, A., 1996. An integrated biosphere model of land surface process, terrestrial carbon balance and vegetation dynamics. Global Biogeochem. Cycles 10, 603–628]. In the N-model, soil mineral N regulates ecosystem carbon (C) fluxes and ecosystem C:N ratios. Net primary productivity (NPP) is controlled by feedbacks from both leaf C:N and soil mineral N. Leaf C:N determines the foliar and canopy photosynthesis rates, while soil mineral N determines the N availability for plant growth and the efficiency of biomass construction. Nitrogen controls on the decomposition of soil organic matter (SOM) are implemented through N immobilization and mineralization separately. The model allows greater SOM mineralization at lower mineral N, and conversely, allows greater N immobilization at higher mineral N. The model's seasonal and inter-annual behaviours are demonstrated. A regional simulation for Saskatchewan, Canada, was performed for the period 1851–2000 at a 10 km × 10 km resolution. Simulated NPP was compared with high-resolution (1 km × 1 km) NPP estimated from remote sensing data using the boreal ecosystem productivity simulator (BEPS) [Liu, J., Chen, J.M., Cihlar, J., Park, W.M., 1997. A process-based boreal ecosystem productivity simulator using remote sensing inputs. Remote Sens. Environ. 44, 81–87]. The agreement between IBIS and BEPS, particularly in NPP spatial variation, was considerably improved when the N controls were introduced into IBIS.     

A box model of carrying capacity for suspended mussel aquaculture in Lagune de la Grande-Entrée,Iles-de-la-Madeleine,Québec

An object-oriented model of environment–mussel aquaculture interactions and mussel carrying-capacity within Lagune de la Grande-Entrée (GEL), Iles-de-la-Madeleine, Québec, was constructed to assist in development of sustainable mussel culture in this region. A multiple box ecosystem model for GEL tied to the output of a hydrodynamic model was constructed using Simile software, which has inherent ability to represent spatial elements and specify water exchange between modelled regions. Mussel growth and other field data were used for model validation. Plackett–Burman sensitivity analysis demonstrated that a variety of bioenergetic parameters of zooplankton and phytoplankton submodels were important in model outcomes. Model results demonstrated that mussel aquaculture can be further developed throughout the lagoon. At present culture densities, phytoplankton depletion is minimal, and there is little food limitation of mussel growth. Results indicated that increased stocking density of mussels in the existing farm will lead to decreased mass per individual mussel. Depending on the location of new farm emplacement within the lagoon, implementation of new aquaculture sites either reduced mussel growth in the existing farm due to depletion of phytoplankton, or exhibited minimum negative impact on the existing farm. With development throughout GEL, an excess of phytoplankton was observed during the year in all modelled regions, even at stocking densities as high as 20 mussels m⁻³. Although mussels cultured at this density do not substantially impact the ecosystem, their growth is controlled by the flux of phytoplankton food and abundance of zooplankton competitors. This model provides an effective tool to examine expansion of shellfish farming to new areas, balancing culture location and density.     

A process-based model of nitrogen cycling in forest plantations: Part II. Simulating growth and nitrogen mineralisation of Eucalyptus globulus plantations in south-western Australia

We applied a new version of the G’DAY ecosystem model to short-rotation plantations of Eucalyptus globulus growing under a Mediterranean climate in south-western Australia. The new version, that includes modified submodels for biomass production, water balance, litter and soil organic matter (SOM) decomposition, and soil inorganic N balance, was parameterised and applied to three experimental eucalypt sites (Mumballup, Darkan and Northcliffe) of contrasting productivity. With a common base set of parameter values, the model was able to correctly reproduce observed time series of soil water content, canopy leaf area index and stemwood data at the three sites. The model's ability to simulate soil N supply under forest plantations was tested by simulating N mineralisation at each of the three sites over the duration of the experiment (10 years). Simulated annual net N mineralisation in the litter and top 20 cm soil layer ranged from 50 to 170 kg N ha⁻¹ across the sites as a result of differences in rates of litter production, SOM and litter decomposition, and microbial N immobilisation and (re-)mineralisation. Simulations of annual soil N mineralisation were similar to measured rates over a 3-year period, except for an overestimation in 1 year at Mumballup and 2 years at Darkan. Model results indicated the importance of fine root production and turnover for N supply. As plantations age, supply of N to trees increasingly originates from litter decomposition, while the contribution from decomposition of SOM decreases. Although major soil feedbacks associated with litter production, decomposition and N availability are adequately integrated into G’DAY, further work is required in some aspects of the model, including the utility of the C-allocation submodel over a wide range of site conditions and silvicultural treatments.     

Modeling and coupling of soil respiration and soil water content in fenced Leymus chinensis steppe,Inner Mongolia

Soil respiration was measured with the enclosed chamber method during 2 years in fenced Leymus chinensis steppe, Inner Mongolia, China. Soil water content at 0–10 cm depth was a major limited factor of soil respiration in semi-arid grassland, accounting for 76.4% of the variation. The temperature-dependent exponential function could only explain 38.7% of the variation in soil respiration. With 246 data over the entire experimental period, multiple linear stepwise regressions of soil respiration rate were analyzed with the influencing factors, including soil water content at 0–10 cm depth, air temperature, air pressure, air humidity, total radiation and their interactions. With soil water content at 0–10 cm depth (W) and air temperature (T_h) as combined factors, the twice linear regression (F = 1.68WT_h − 109.09) was simple and its coefficients were significant, accounting for 83.1% of the variation in soil respiration. Due to the lack of long-term and continuous soil water content, a water sub-model based on precipitation and evapotranspiration was introduced, which could provide better fits with the measured values (R² = 0.813). The magnitudes of soil respiration calculated from the twice linear regression equation and water sub-model were 439.58 and 463.06 g CO₂ m⁻² in 2001 (19 June–23 September) and in 2002 (1 June–24 September), respectively. The mean hourly soil respiration rates were in the range of the previous studies in the adjacent region and the world's major temperate grasslands.     

Modelling nutrient emissions from river systems and loads to the coastal zone: Po River case study,Italy

The nutrient emission model MONERIS (MOdelling Nutrient Emissions into River Systems) is applied to the Po catchment, a large (>70,000 km²), densely populated, highly agriculturally exploited and industrialized landscape. The catchment is located in northern Italy. The Po River discharges into the northwestern Adriatic Sea. Model runs cover the period 1991–2000. The purpose is to model the catchment in 2001, estimating nutrient emissions and natural background in the basin and loads to the coastal area. The model was calibrated with data for the period 1990–1995. After validation with data for the period 1995–2000, the model is used to evaluate future catchment management scenarios.MONERIS is a spatially distributed parameters steady state model with a time scale of 5 years. The emissions considered are originated from diffuse and point sources and delivered trough various pathways (groundwater, erosion, overland flow, atmospheric deposition, urban systems and WWTPs). In order to estimate nutrient loads to the river system, MONERIS includes a retention model.An overview of model input requirements, data needs and related problems and solutions adopted is presented in the paper. Simulated and measured data of several sections along the river are compared for calibration and validation. The relative importance of different nutrient generation pathways are evaluated. Finally, forecasted yearly nutrient loads at the outlet of PO basin for the years 2001, 2008 and 2016, consequence of different basin management scenarios, are presented. The results are ready to be supplied to a water quality Coastal Zone Model, allowing us to evaluate significant switches in trophic state conditions of the coastal ecosystem [see Artioli, Y., Bendoricchio, G., Palmeri, L., this issue. Defining and modelling the coastal zone affected by the Po River (Italy). Ecol. Model.].     

Modelling carbon storage in highly fragmented and human-dominated landscapes: Linking land-cover patterns and ecosystem models

To extend coupled human–environment systems research and include the ecological effects of land-use and land-cover change and policy scenarios, we present an analysis of the effects of forest patch size and shape and landscape pattern on carbon storage estimated by BIOME-BGC. We evaluate the effects of including within-patch and landscape-scale heterogeneity in air temperature on carbon estimates using two modelling experiments. In the first, we combine fieldwork, spatial analysis, and BIOME-BGC at a 15-m resolution to estimate carbon storage in the highly fragmented and human-dominated landscape of Southeastern Michigan, USA. In the second, we perform the same analysis on 12 hypothetical landscapes that differ only in their degree of fragmentation. For each experiment we conduct four air-temperature treatments, three guided by field-based data and one empirically informed by local National Weather Service station data. The three field data sets were measured (1) exterior to a forest patch, (2) from the patch edge inward to 60 m on east-, south-, and west-facing aspects, separately, and (3) interior to that forest patch. Our field-data analysis revealed a decrease in maximum air temperature from the forest patch edge to a depth of 80 m. Within-patch air-temperature values were significantly different (α = 0.01) among transects (c.v. = 13.28) and for all measurement locations (c.v. = 30.58). Results from the first experiment showed that the interior treatment underestimated carbon storage by ～8000 Mg C and the exterior treatment overestimated carbon storage by 30,000 Mg C within Dundee Township, Southeastern Michigan, when compared to a treatment that included within-patch heterogeneity. In the second experiment we found a logarithmic increase in carbon storage with increasing fragmentation (r² = 0.91). While a number of other processes (e.g. altered disturbance frequency or severity) remain to be included in future experiments, this combined field and modelling study clearly demonstrated that the inclusion of within-patch and landscape heterogeneity, and landscape fragmentation, each have a strong effect on forest carbon cycling and storage as simulated by a widely used ecosystem process model.     

Utilizing satellite imagery and GLOBE student data to model soil dynamics

General Purpose Atmosphere Plant Soil Simulator (GAPS), a menu-driven soil-vegetation-atmosphere transfer (SVAT) model, was used to simulate soil water dynamics from 1998 through 2001 for Greenville, PA, USA. GLOBE student data collected by students from Reynolds Junior and Senior High School, coupled with normalized difference vegetation index (NDVI) data derived from SPOT4 vegetation imagery, were used to parameterize and validate the model. Data from the National Weather Service Cooperative (NWSC) was used to evaluate the GLOBE dataset. Overall, there was a high index of agreement (d > 0.80) between field measurements and simulated soil water values from both datasets (GLOBE and NWSC). Simulations using the GLOBE climate data outperformed the NWSC data for the 1999, 2000, and 2001 growing seasons. In addition, the GLOBE simulations showed that NDVI could be utilized to predict transpiration periods (QI, QII, and QIII) for northern latitudes >35° with a distinct winter period. In phenological terms, QI reflects the onset of the growing season when vegetation is greening up (NDVI < 0.60) and transpiration is beginning (<2 mm/day) and QII reflects the end of the growing seasons when vegetation is greening down and transpiration is decreasing. QIII reflects the height of the growing season when transpiration rates average between 2 and 5 mm per day and NDVI is at its maximum (>0.60). Results of this study demonstrate that GLOBE student data, coupled with remotely sensed data, can provide an important source of input and validation information for capacitance SVAT models such as GAPS.     

A peaked function for modeling temperature dependence of plant productivity

We suggests that temperature response of plant productivity can be modeled by the Arrhenius function modified to describe the effect of temperature on enzyme activity: G_A(T) = 2f(T)/(1 + f²(T)), where f(T) = exp(E_a/RT_opt − E_a/RT), R the universal gas constant, E_a the activation energy and T_opt is the optimal temperature. In common with other functions used for modeling the temperature response of plant productivity, the curve of function G is almost symmetrical and bell-shaped. The special convenience of G_A is that it relates the width of the “bell” to thermodynamic concepts, such as activation energy of chemical reactions converting carbon dioxide and water to carbohydrates.     

A preliminary model of nutrient cycling in sediments of a Mediterranean lagoon

A model of nitrogen and phosphorus cycles in the sediment of a lagoon has been developed. This model was applied to the Thau lagoon (southern France). Sediment was sliced in three layers to reproduce the oxygen profile, which is simulated within the model. Following an equilibrium hypothesis, the model was calibrated against field data. State variables and fluxes were estimated in the sediment and across the sediment-water interface. A Monte Carlo sensitivity analysis was performed to determine the most sensitive parameters and sediment state variables. A dynamic simulation with varying oxygen concentrations then showed the influence of anoxia on the phosphorus and nitrogen fluxes between water and sediment.     

A simplified approach to implement forest eco-hydrological properties in regional hydrological modelling

The topic of this paper is a simplified model for simulating the hydrological properties of forest stands based on a robust computation of the temporal LAI (leaf area index) dynamics. The approach allows the simulation of all hydrologically relevant processes. It includes interception of precipitation and transpiration of forest stands with and without groundwater in the rooting zone. The model also considers phenology, mortality and simple management practice. It was implemented as a module in the eco-hydrological model SWIM (Soil and Water Integrated Model). The approach was tested on Scots pine (Pinus sylvestris) and common oak (Quercus robur and Q. petraea).The results demonstrate a good simulation of annual biomass increase and LAI and satisfactory simulation of litter production (annual mean value). A comparison of the date of May sprout for Scots pine and leaf unfolding for Oak (1980–1990) with observed data of the DWD (German Weather Service) shows a good reproduction of the temporal dynamic. The daily simulation of transpiration shows an excellent correlation of r = 0.81 for the year 1998 but only r = 0.65 for 1999. The interception losses were also simulated and compared with weekly observed data showing satisfactory results in the vegetation periods and annual sums, but worse agreement in autumn and spring time. A regional assessment study was done in the federal state of Brandenburg (Germany) to test the applicability and multi-criteria evaluation capabilities of the approach on the landscape and catchments scale using forest data, daily river discharge and regional water balance.     

Release thresholds strongly determine the range of seed dispersal by wind

The effect of the seed abscission process on the dispersal distance of seeds has never been studied explicitly and is often ignored in studies that aim to estimate the seed shadows of species. To examine the importance of the abscission process for the seed shadow we used a seed trajectory model that keeps track of the release threshold dynamics of the individual seeds on mother plant. We defined the release threshold as the critical wind speed that induces a mechanical force that is just large enough to release a seed from its mother plant. The model used real wind speed sequences and seed appearance over time on the mother plant.Several calculations were performed to investigate the effect of release thresholds dynamics on seed shadow of two herbaceous species with contrasting terminal velocity values (V_t): Centaurea jacea (V_t = 4.1 m s⁻¹) and Hypochaeris radicata (V_t = 0.49 m s⁻¹).Release thresholds were responsible for a two-fold increase of median dispersal distances in both species. Tails of the seed shadows, the fraction of seeds that travel furthest, were even more sensitive and increased with a factor 4.5 for Centaurea and 7.0 for Hypochaeris. Our work indicates that the abscission process appears to be very important and suggests that dispersal distance of plants is currently severely underestimated, which, in turn, has major consequences for our current understanding of the distribution, metapopulation dynamics and survival of plant species.     

Artificial neural network approach for modelling nitrogen dioxide dispersion from vehicular exhaust emissions

Artificial neural networks (ANNs) are useful alternative techniques in modelling the complex vehicular exhaust emission (VEE) dispersion phenomena. This paper describes a step-by-step procedure to model the nitrogen dioxide (NO₂) dispersion phenomena using the ANN technique. The ANN-based NO₂ models are developed at two air-quality-control regions (AQCRs), one, representing, a traffic intersection (AQCR1) and the other, an arterial road (AQCR2) in the Delhi city. The models are unique in the sense that they are developed for ‘heterogeneous¹’ traffic conditions and tropical meteorology. The inputs to the model consist of 10 meteorological and 6 traffic characteristic variables. Two-year data, from 1 January 1997 to 31 December 1998 has been used for model training and data from 1 January to 31 December 1999, for model testing and evaluation purposes. The results show satisfactory performance of the ANN-based NO₂ models on the evaluation data set at both the AQCRs (d = 0.76 for AQCR1, and d = 0. 59 for AQCR2).     

Analyzing and modelling spatial distribution of summering lesser kestrel: The role of spatial autocorrelation

In modelling spatial distribution of species, ignoring spatial autocorrelation (SA) and multicollinearity may lead to false ecological conclusions. Here we take into account both issues for examining and modelling the spatial pattern of abundance of the globally threatened lesser kestrel (Falco naumanni) during summer in a 38,400 ha area of northwestern Spain where large premigratory aggregations of the species occur. Spatial pattern was examined using Moran's correlogram, and models were built including geographical coordinates and autocovariate terms (which account for SA) in generalized linear models (GLM) and hierarchical partitioning (HP) models. HP models allow to alleviate multicollinearity. A grid-based approach was used by dividing the study area in 24 contiguous 4 km × 4 km squares where birds were counted in 2–3 visits per square (response variable). Environmental coarse-grained variables were extracted from a geographic information system (GIS) at three spatial extents. Moran's correlogram showed that lesser kestrel mean abundance per square was spatially autocorrelated up to 4–8 km. The results from both GLM and HP analyses were roughly compatible. The GLM models explained 80.0% of the variation in kestrel abundance and were the same at the three spatial extents. Lesser Kestrel abundance was not significantly explained by landscape variables, but was negatively related to both the distance to the nearest communal roost and distance to the nearest breeding colony with more of 10 breeding pairs of lesser kestrel. An autocovariate term added later in the GLM models improved both their explanatory power (from 74.5 to 80.0%) and model residuals, which were not longer spatially autocorrelated, fulfilling thus the statistical assumption of independent errors. Findings suggest that the spatial distribution of abundance of summering lesser kestrel is, at least, partially driven by endogenous causes, such as conspecific attraction. Exogenous causes such as finer-scale variables (e.g. type of crops and food available) are yet likely needed for lesser kestrel-environment relationships.     

Catastrophic-like shifts in shallow Turkish lakes: a modeling approach

A generalized logistic model (GLM) was developed to determine occurrence of submerged macrophytes in shallow Lake Eymir, and the model was tested independently on the upstream shallow Lake Mogan using the data collected fortnightly from both lakes during 1997–2002. The independent variables included concentrations of chlorophyll-a (chl-a), suspended solids (SS) and total phosphorus (TP), Secchi disc transparency and z-scores of water levels. The dependent variable was the binary index of submerged plant occurrence. We used bootstrapping to determine the maximum number of epochs to train the model and to execute training when the corrected average cross entropy (c-index) leveled off. The model predicted that SS concentration, z-scores of water levels and TP concentration were the most important variables for determining occurrence of submerged plants. Sensitivity analyses showed that the probability of submerged plant occurrence followed a strong hysterisis response to varying water levels and the concentrations of SS and TP, with the break points being ±50 cm, 12–17 mg l⁻¹ and 200–300 μg l⁻¹, respectively. This observed sensitivity was in accordance with the alternative stable states hypothesis of shallow lakes. For occurrence of submerged plants, chlorophyll-a concentration and Secchi disc transparency had low significance. This was in concert with both relevances of input variables and the field results. The model gave a good definition of the system since the c-index and corrected c-index on the training data were high (0.970 and 0.963, respectively). Testing the model on Lake Mogan produced a c-index of 0.815 with around 80% of the cases being correctly classified. This showed that the model had a high ability to generalize over a spatially independent test set; therefore, it had a great reliability as well. In addition, the predictive power of the model was indeed very high. Consequently, the model captured the relationships between the input and output variables successfully and consistently with alternative stable states hypothesis.     

Nutrient budget of the lagoonal waters in an open central South Pacific atoll (Tikehau,Tuamotu, French Polynesia)

Nutrient concentrations were measured in the lagoon and surrounding oceanic waters of Tikehau Atoll (Tuamotu Archipelago, French Polynesia) from 1984 to 1987. The atoll-mass effect alters the nutrient profiles: turbulent vertical mixing of the waters along the deeper slopes of the atoll induces nitrogen and phosphorus enrichment of the surface layer. Nutrient concentrations varied with year and month of sampling; except for ammonium, inorganic nutrient levels were lower inside the lagoon than in the surrounding oceanic waters. Nitrogen, phosphorus and silica budgets were calculated by mean differences in nutrient concentrations recorded between lagoon and oceanic surface waters and by the waterexchange rate through the passage linking the lagoon and oceanic waters and the reef-flat spillways. Particulate and dissolved organic nitrogen and ammonium are exported from the lagoon to the open ocean through the westward passage. The nitrogen budget is not balanced by the nitrate input from oceanic waters and the organic nitrogen and ammonium output from lagoonal waters. Nitrogen fixation would appear to constitute another source of nitrogen for lagoonal waters. The phosphorus budget is largely balanced by phosphate input from the oceanic waters and organic phosphorus output from the lagoon waters. The oceanic waters became impoverished in silicate during their crossing of the atoll reef edge and their residence in the lagoon. The atoll constitutes a source of nitrogen for the surrounding oceanic waters.     

The effects of summer heat on academic achievement: A cohort analysis

This paper analyzed the effect of summer heat on academic achievement. Summer heat can negatively affect student learning, as previous studies have shown that high temperatures in laboratory settings have a negative effect on cognitive abilities. For this analysis, the test scores of five different cohorts were combined with city-level daily temperature data. To control for unobserved heterogeneity, the test scores of students within the same school were compared over time (school-fixed effects estimation). Summer heat negatively affected student test scores. Specifically, an additional day with a maximum daily temperature exceeding 34 °C (93.2 °F) during the summer, relative to a day with a maximum temperature between 28 °C (82.4 °F) and 30 °C (86 °F), decreased the scores of math and English tests by 0.0042 and 0.0064 standard deviations, respectively. No significant effects were found on the reading test scores. In addition, these effects were larger in relatively cooler cities, but did not differ based on gender. Finally, the previous year's summer also had negative effects on the current year's test scores.