Search

Conventional cultivation of claypan soils leads to soil and water quality degradation because of high runoff and associated soil erosion. The Goodwater Creek Experimental Watershed, which is part of the USDA Agricultural Research Service Benchmark Conservation Effects Assessment Project, Watershed Assessment Studies, was established to address these issues. Plot studies have highlighted trade-offs between erosion control and herbicide or nutrient runoff. There is a need for long-term field-scale evaluation of combined practices that reduce sediment, nutrient, and herbicide losses by runoff. A 36 ha field located in Missouri was under a conventional corn (Zea mays L.)–soybean (Glycine max L.) system from 1993 to 2003 with fertilizer application and tillage prior to planting in the spring. A precision agriculture system defined by two main management zones was implemented from 2004 to 2014: wheat (Triticum aestivum L.) and soybean in 60% of the field, and corn and soybean in the remaining 40%. The system included no-till, cover crops, atrazine split-applications based on weed pressure, variable rates of nitrogen (N), and variable rates of fall-applied phosphorus (P). The objective of this study was to compare runoff water quality from the two management systems, based on flow and load duration curves, cumulative distribution functions, and conclusions from replicated plot studies. The precision agriculture system did not affect annual runoff, but it did increase the frequency of low flows. Sediment losses were reduced by 87% as a result of no-till and cover crops. Atrazine and P losses were lower than expected, despite the lack of incorporation into the soil. Atrazine losses were possibly lower because of the wheat area acting as a buffer, greater atrazine adsorption and retention in the field, and faster decay. Dissolved P losses as a fraction of applied remained the same, likely because of greater adsorption and lower runoff risk when applying P. Finally, nitrate-N (NO3-N) losses decreased and resulted in an overall decrease of N losses despite a slight increase of ammonium-N (NH4-N) losses. Explanations included a greater soil water content, a different timing of N applications, and N uptake by cover crops. Building on these successes, an aspirational management system is proposed to further improve on the performance and practicality of the precision agriculture system.

The Soil Vulnerability Index (SVI) was developed by the USDA Natural Resources Conservation Service (NRCS) to identify inherent vulnerability of cropland to runoff and leaching. It is a simple index that relies on the SSURGO database and can be used with basic knowledge of ArcGIS. The goal of this study was to investigate a relationship between constituent (sediment and nutrient) loadings and fraction of the watershed in each SVI class. The SVI maps were developed for each of the seven subwatersheds of the Mark Twain Lake watershed in Missouri, which were similar in soil conditions and climatic variability. The SVI assessment was performed by investigating if the distribution of the SVI for cropland in each subwatershed could help explain measured 2006 to 2010 sediment and nutrient loads better than crop distribution alone. Regression analyses were performed between annual loads of sediment and nutrients exported from the watersheds and a composite number that included either cropland distribution alone, or cropland distribution combined with the SVI. Coefficients of determination and p-values were compared to assess the ability of land use and SVI distributions to explain stream loads. Integrating the SVI in the land cover variable improved the ability to explain constituent loads in the watersheds for sediment, total nutrients, and dissolved nitrogen (N). Regression results with and without the SVI were identical for dissolved phosphorus (P), potentially indicating that SVI was not indicative of dissolved P transport at the current site. Overall, the application of the SVI at watershed scale was not perfect, but acceptable at correctly identifying cropland of greatest vulnerability and linking with transported constituent loads.

Evaluation of the sustainability of grain production systems requires knowledge of long-term crop yields, and edge-of-field runoff volumes and water quality. When measured data are limited, computer simulation models —calibrated and validated with the available data— are well suited for this type of inquiry. In this study, the Agricultural Policy Environmental Extender (APEX) model was used to investigate how management affects crop yields, runoff volume, and water quality on claypan soils. Specific objectives were to determine the ability of APEX to simulate the variability of corn and soybean yields on claypan soils and the transport of dissolved-phase atrazine and nutrients via surface runoff; and how management affects long-term runoff and associated loss of these constituents. In spite of satisfactory model performance for corn yields and acceptable relative differences for soybean yields, the model was not capable of simulating the effects of management or landscape position on crop yields. Model performance was very good for runoff, and mostly satisfactory for dissolved constituents. Long-term simulation along with uncertainty analysis indicated that a no-till 3-year corn-soybean-wheat rotation with cover crops provides the most benefits to water quality by reducing nutrient transport in runoff. This system appeared to mitigate the negative effects of no-till on the surface transport of non-incorporated dissolved chemicals because of improved fertilizer management in terms of rate, placement, and timing of application. Cover crops appeared to reduce nitrate leaching compared to a 2-year no-till system. This research highlights the need for further research in crop growth modeling as inconsistent crop yield results undermine the usefulness of the model for making recommendations about management. Overall, the long-term simulations support the concept of multiple practices addressing tillage, fertilizer management, longer crop rotations, and use of cover crops to achieve desired environmental and production goals.

713 The USDA-ARS Long-Term Agro-Ecosystem Research (LTAR) network was established in 2012 as an agroecosystem assessment, complementary to the National Science Foundation’s Long-Term Ecological Research and National Ecological Observatory Network. Th e goal of the LTAR network is to “to sustain a land-based infrastructure for research, education, and outreach that enables understanding and forecasting of our capacity to provide agricultural commodities and other agro-ecosystem goods and services under changing conditions” (Walbridge and Shafer, 2011). Th e Goodwater Creek Experimental Watershed in northeastern Missouri was included in the Central Mississippi River Basin LTAR site due to its rich history (>40 yr) of agricultural productivity and environmental monitoring research. Comprehensive documentation of the establishment, overview, and available data from this LTAR site exists in Sadler et al. (2015a). A signifi cant component of baseline and future eff orts of the LTAR site in the Central Mississippi River Basin is a replicated, plot-scale long-term CS study established in 1991. Some distinctive features of this study are that it is situated on a claypan soil toposequence, includes variable input annual crop and perennial grass systems, has been and is instrumented with water quality monitoring infrastructure, and is linked to adjacent research fi elds where CS practices from the plots have been implemented at a larger scale. Previous studies have published fi ndings from this long-term study to document the impacts of CS and LP on environmental parameters, including investigations on soil quality and health (Jung et al., 2007, 2008, 2010; Jiang et al., 2007a; Chaudhary et al., 2012; Veum et al., 2015a, 2015b) and agrichemical loss and water quality (Kitchen et al., 1998; Ghidey et al., 2005; Mudgal et al., 2010). Results of these studies have highlighted some of the challenges and opportunities to sustainable grain crop production on claypan soils. Findings include: (i) greatly impaired productivity and ecosystem function due to historic topsoil loss; (ii) strong spatial variability in soil properties, crop growth, and crop yield; (iii) high propensity for agrichemical loss through surface runoff and denitrifi cation, especially with no-till; and (iv) improvements in soil quality and ecosystem function over time in conservation CS, but not in a conventional system. While many of the Crop Economics, Production & Management

The Conservation Effects Assessment Project (CEAP) was initiated in 2002 to quantify the potential benefits of conservation management practices throughout the nation. Within the Central Claypan Region of Missouri, the Salt River Basin was selected as a benchmark watershed for soil and water quality assessments. This study focuses on two objectives: (1) assessing soil quality for 15 different annual cropping and perennial vegetation systems typically employed in this region, and (2) evaluating relationships among multiple measured soil quality indicators (SQIs). Management practices included annual versus perennial vegetation, and varying grass species composition (cool-season versus warm-season), tillage intensity (no-till versus mulch-till), biomass removal, rotation phase, crop rotation (corn [Zea mays L.]–soybean [Glycine max L. Merr] versus corn–soybean–wheat [Triticum aestivum L.]) and incorporation of cover crops into the rotation. Soil samples were obtained in 2008 from 0 to 5 cm (0 to 2 in) and 5 to 15 cm (2 to 6 in) depth layers. Ten biological, physical, chemical, and nutrient SQIs were measured and scored using the Soil Management Assessment Framework (SMAF). Across SQIs, biological and physical indicators were the most sensitive to management effects, reflecting significant differences in organic carbon (C), mineralizable nitrogen (N), β-glucosidase, and bulk density. In the 0 to 5 cm layer, perennial systems demonstrated the greatest SMAF scores, ranging from 93% to 97% of the soil9s inherent potential. Scores for annual cropping systems ranged from 78% to 92%: diversified no-till, corn–soybean–wheat rotation with cover crops (92%) > no-till, corn–soybean rotation without cover crops (88%) > mulch-till corn–soybean rotation without cover crops (84%). Conversely, in the 5 to 15 cm layer, no-till cropping systems scored lower for overall soil function (58% to 61%) than mulch-till systems (65% to 66%). In the 0 to 5 cm layer, biological soil quality under the diversified no-till system with cover crops was 11% greater than under no-till without cover crops, and 20% greater than under mulch-till without cover crops. The effect of rotation phase was primarily reflected in 64% lower mineralizable N following corn relative to soybean. Additionally, soil nutrient function was significantly affected by biomass removal. The results of this study demonstrate that the benefits of conservation management practices extend beyond soil erosion reduction and improved water quality by highlighting the potential for enhanced soil quality, especially biological soil function. In particular, implementing conservation management practices on marginal and degraded soils in the claypan region can enhance long-term sustainability in annual cropping systems and working grasslands through improved soil quality.

Various land management decisions are based on local soil properties. These soil properties include average values from soil characterization for each soil series. In reality, these properties might be variable due to substantially different management, even for similar soil series. This study was conducted to test the hypothesis that for claypan soils, hydraulic properties can be significantly affected by long-term soil and crop management. Sampling was conducted during the summer of 2008 from two fields with Mexico silt loam (Vertic Epiaqualfs). One field has been under continuous row crop cultivation for over 100 years (Field), while the other field is a native prairie that has never been tilled (Tucker Prairie). Soil cores (76 × 76 mm (3.0 × 3.0 in)) from six replicate locations from each field were sampled to a 60 cm (24 in) depth at 10 cm (3.9 in) intervals. Samples were analyzed for bulk density, saturated hydraulic conductivity (Ksat), soil water retention, and pore-size distributions. Values of coarse (60 to 1,000 μm (0.0024 to 0.039 in) effective diameter) and fine mesoporosity (10 to 60 μm (0.00039 to 0.0024 in) effective diameter) for the Field site (0.044 and 0.053 m 3 m -3 (0.044 and 0.053 in 3 in -3 )) were almost half those values from the Tucker Prairie site (0.081 and 0.086 m 3 m -3 (0.081 and 0.086 in 3 in -3 )). The geometric mean value of Ksat was 57 times higher in the native prairie site (316 mm h -1 (12.4 in hr -1 )) than in the cropped field (5.55 mm h -1 (0.219 in hr -1 )) for the first 10 cm (3.9 in) interval. Differences in Ksat values were partly explained by the significant differences in pore-size distributions. The bulk density of the surface layer at the Tucker Prairie site (0.81 g cm -3 (50.6 lb ft -3 )) was two-thirds of the value at the Field site (1.44 g cm -3 (89.9 lb ft -3 )), and was significantly different throughout the soil profile, except for the 20 to 30 cm (7.9 to 12 in) depth. These results show that row crop management and its effect on soil loss have significantly altered the hydraulic properties for this soil. Results from this study increase our understanding of the effects of long-term soil management on soil hydraulic properties.

Streams and drinking water reservoirs throughout the claypan soil region of Missouri and Illinois are particularly vulnerable to herbicide contamination from surface runoff during spring. This study follows a plot-scale study conducted on claypan soils to quantify and compare edge-of-field herbicide losses from a corn–soybean rotation under mulch tillage and no-tillage systems. The objectives of the present study were to confirm at field scale (34.4 ha [85 ac] and 7.8 ha [19.3 ac]) the plot-scale findings (0.37 ha [0.92 ac]) on the effects of tillage and herbicide incorporation on herbicide transport and to evaluate the applicability of plot-scale exponential models in calculating atrazine and metolachlor concentrations as a function of application rate, runoff volume, and days after application at the field scale. Herbicide transport to surface runoff was studied (1997 to 2001) from two fields with cropping systems similar to those on the plots. Field 1 (F1) was a mulch tillage corn–soybean rotation system with surface-applied herbicides, which are then incorporated. Field 2 (F2) was a no-tillage corn–soybean rotation system with surface-applied herbicides that were not incorporated. During each event, runoff volumes were measured, and water samples were collected and analyzed for atrazine and metolachlor concentrations. The percentages of applied atrazine and metolachlor transported to surface runoff from no-tillage (F2) were 3.2 and 2.0 times those from mulch tillage (F1), respectively. Throughout the study period, 1.0% and 3.2% of total atrazine and 1.0% and 2.0% of total metolachlor applied to F1 and F2 were lost to surface runoff, respectively. Similar to the results from the plot study, the model performed well in calculating field atrazine concentrations from both mulch and no-tillage systems with coefficient of determination ≥ 0.70 and Nash and Sutcliffe efficiency ≥ 0.64. However, model performance in calculating metolachlor concentrations was poor for both tillage systems (Nash and Sutcliffe efficiency

The Mark Twain Lake/Salt River Basin was selected as one of the USDA Agricultural Research Service benchmark watersheds for the Conservation Effects Assessment Project because of documented soil and water quality problems and broad stakeholder interest. The basin is located in northeastern Missouri within the Central Claypan Region, and it is the source of water to Mark Twain Lake, the major public water supply in the region. At the outlet to Mark Twain Lake, the basin drains 6,417 km 2 (2,478 mi 2 ), including 10 major water- sheds that range in area from 271 to 1,579 km 2 (105 to 609 mi 2 ). The basin is character- ized by flat to gently rolling topography with a predominance of claypan soils that result in high runoff potential. The claypan soils are especially vulnerable to soil erosion, which has degraded soil and water quality throughout the basin, and to surface transport of herbicides. Results from cropping system best management practice studies showed that no-till cropping systems did not reduce surface runoff compared to tilled systems, and no-till led to increased transport of soil-applied herbicides. A major challenge is the need to develop cropping systems that incorporate herbicides yet maintain sufficient crop residue cover to control soil ero- sion. Results of the Soil and Water Assessment Tool model simulations showed that the model was capable of simulating observed long-term trends in atrazine concentrations and loads and the impact of grass waterways on atrazine concentrations. Current and future research efforts will continue to focus on best management practice studies, development of needed tools to improve watershed management, and refinements in the calibration and validation of the Soil and Water Assessment Tool model.

In nature and in the most primitive agricultural systems, seed distribution is broadcast across the landscape. Such a distribution results in nearly uniform spacial interaction of the developing phytomass. With the development of agrarian civilization has come an implement-dependent systemization of crop- planting patterns. This has brought about the planting of crops in uniform rows—from the drilling of small grains at inter-row spacings of 0.1 to 0.2 m and plant intra-row spacings of i to 5 cm, to the staking of horticultural and vine crops at 2- to 3—m inter-row spacings and typically 0.3- to l-m intra-row spacings. The implement dependence of agricultural cropping strategies has resulted in row cropping. The staple crops regarded as most suited to this approach are commonly called row crops, and this review will concentrate largely on how row crops interact with plant geometry, water, and nutrients to in?uence sustained productive capacity.

Proximal ground conductivity sensors produce high spatial resolution maps that integrate the bulk electrical conductivity (ECa) of the soil profile. For meaningful interpretation, variability in conductivity maps must either be inverted to profile conductivity or be directly calibrated to profile properties. Penetrating apparent electrical conductivity (ECa–P) sensors produce high depth resolution data at relatively fewer spatial locations. The objectives of this research were to (i) investigate the profile source of ECa in claypan soils via a detailed examination of ECa–P profiles; (ii) examine the potential for feature detection with ECa–P in claypan soils; and (iii) determine if ECa sensors can be calibrated to ECa–P features. Two study areas were chosen representing the claypan soils of north-east Missouri, USA. Profile conductivity was measured at high depth resolution on soil cores using a miniaturised Wenner conductivity probe and in the field using a conductivity penetrometer. Proximal ground conductivity was mapped with one direct contact sensor and two non-contact sensors, providing five distinct coil/electrode geometries. Increasing ECa–P was observed below the claypan, correlated with decreasing clay and water content and increasing bulk density. Depth to the claypan was successfully calibrated to derivative peaks on ECa–P profiles (R 2 = 0.72, p < 0.001). Relationships between ECa and ECa–P features were poor (R 2∼ 0.21) to good (R 2∼ 0.87) on a field-specific basis. Results show that ECa–P can be used for calibration of ECa to the depth to claypan.

Over the last three decades or more, researchers have estimated soil properties using visible and near-infrared (VNIR) diffuse reflectance spectroscopy (DRS), with varying results. Using VNIR DRS for estimating soil property variation within fields is particularly challenging, because in many cases the variation in the property of interest may be relatively small and because of the need to deal with spatially correlated data. In this study, we used VNIR DRS to estimate the variability in soil physical and chemical properties within a typical production field in north-east Missouri, USA. Soil samples were obtained to 15 cm depth on a 30 m grid spacing, plus at a number of random sampling locations. Laboratory analyses were conducted for sand, silt, and clay fractions, organic matter, pH, P, K, Ca, Mg, and cation exchange capacity (CEC). VNIR reflectance of dried and sieved samples was obtained in the laboratory using a spectrometer with a wavelength range from 350 to 2,500 nm. Partial least squares (PLS) regression was used to estimate soil properties from spectra, both for the full range and for subset wavelength ranges above and below 1,000 nm. A regression kriging approach was used to account for spatial dependence. We found that for these soils, an NIR-only instrument (1,000–2,500 nm) would be able to quantify CEC, organic matter, and texture with accuracy similar to that from a VNIR (350–2,500 nm) instrument. Some soil properties, including CEC and pH, were well estimated with PLS regression, while others, including organic matter, were not. Coupling regression kriging with PLS regression improved estimates in some cases, but was not as effective as has been reported in some other studies. More advanced calibration methods should be investigated for their ability to improve within-field VNIR DRS results on these soils.

Crop growth models that account for nutrient accumulation offer insight into soil fertility and plant nutrition interactions. This understanding provides opportunities to develop improved management practices. During the 1980s, several process‐level growth models were developed for soybean [Glycine max (L.) Merr.]. Model validation and application to different locations and weather require detailed, independent data sets. An extensive data set describing the nutrient status of a determinate soybean ('Bragg') was collected in 1979 on a Goldsboro (Aquic Paleudult) loamy sand near Florence, SC, USA. Because of its importance to subsequent model development, we concluded that providing this entire data set in a readily accessible form was a logical step in the course of this experiment. We report here, in tabular form, mean and standard deviation data for aerial accumulation of dry matter and eight nutrients (N, P, K, Ca, Mg, Mn, Fe, and Zn) for 10 dates, for four plant components (stems, leaves, pet...

The primary model selected for use in the Conservation Effects Assessment Project-Watershed Assessment Study (CEAP-WAS) was the Soil and Water Assessment Tool (SWAT) model. In this study, the scaling of the SWAT model from a small watershed to a larger watershed was evaluated. The model was first calibrated and evaluated for Goodwater Creek Experimental Watershed (GCEW), a 70 km2 watershed located in north-central Missouri, using SSURGO and STATSGO soil data sets. Then, the performance of the calibrated model in simulating stream flow from Long Branch watershed, a 462 km2 watershed that contains GCEW, was evaluated. Both GCEW and Long Branch watersheds had similar soil, land use, and cropping and management systems. The performance of the model in simulating stream flow from the Long Branch watershed was as good as that from GCEW. For Long Branch watershed, the 9-yr (1995-2003) average simulated annual stream flow was less than 4 % higher than that measured, and the ENS and r2 values were 0.97 and 0.94 for annual stream flow and 0.79 and 0.77 for monthly stream flow, respectively. The model, however, did not perform well in simulating daily stream flow. Overall, the model performed quite well in simulating annual and monthly stream flows. Future plans include (1) improving stream flow simulation on a daily basis, (2) improving sediment yield simulation, and (3) calibrating the pesticide component of the model.

The United States Department of Agriculture Conservation Effects Assessment Project (CEAP) was initiated to quantify the benefits of conservation practices on a national scale. This paper will address water quality benefits from conservation practices implemented at the field level using measured water quality data from the Goodwater Creek watershed in the claypan area of north central Missouri. Eleven years of hydrologic and climatic data from 1993-2003 were analyzed to identify trends and possible effects of best management practices (BMPs), including grassed waterways and terraces, on atrazine stream loadings and concentrations. During this period, area implemented with BMPs increased by 12%. Trends were identified using regression models. Over the time period, atrazine levels have been affected by drier, warmer springs, and increased no-till practices causing earlier application of atrazine, increased levels of atrazine in April, and decreased levels as the season progressed. Atrazine loading was reduced significantly over the entire eleven year time period (P

The Soil and Water Assessment Tool (SWAT) model was the primary model selected for use in the Conservation Effects Assessment Project-Watershed Assessment Study (CEAP-WAS). In this paper, the performance of SWAT in simulating streamflow, sediment yield, and atrazine loss from Goodwater Creek, a 72 km2 watershed located in the claypan soil region of north-central Missouri, was evaluated. When the model was run using default parameters, it overestimated average annual streamflow by 32%, underestimated average annual sediment yield by 23%, and overestimated average annual atrazine loss by 8%. The Nash-Sutctliff coefficient (ENS) values were < 0.35 when annual estimated and measured values were compared. The model was calibrated for streamflow using data collected from 1992 through 1996. After calibration, the difference between measured and estimated average annual flow was < 5%, and the ENS values for annual and monthly simulation results were 0.90 and 0.85, respectively. After calibration, the model did not perform well in estimating streamflow on a daily basis (ENS=0.3). The calibrated model was also validated using data collected from 1997 through 2002 from the Goodwater Creek watershed and the performance of the model was good with ENS=0.88 for annual streamflow and ENS=0.68 for monthly streamflow. Overall, the annual and monthly streamflow predictions of the model were good. Future plans include (1) improving streamflow estimation on a daily basis, and (2) calibrating the sediment and pesticide components of the model.

International audience

A side-by-side comparison was made in Avignon, France, during 1980, of two greenhouse types: a standard greenhouse with forced air heating and forced ventilation, and a fluid-roof house in which the temperature was controlled by water flowing over a roof made of infrared absorbing glass. Heat exchange with groundwater at 14°C kept the circulating water between 12 and 23°C. Both systems worked essentially as predicted by a simulation model on a cold day in April and a warm day in June. The model used measured meteorological data to calculate the control action and the actual conditions prevailing in both greenhouses. The predictions were adequate, except for a tendency to over-estimate interior humidities. The model was useful in suggesting that, in spite of essentially equal air temperatures, the conditions for plant growth were substantially different in the two greenhouses. In the fluid-roof house, there would be less light, a higher humidity, a much reduced transpiration, and, during the day, a somewhat lower CO 2 level and lower leaf temperature, all as compared to the standard glasshouse.

A 22-m 2 greenhouse was fitted with a hollow, channeled plastic roof through which a 3% solution of CuCl 2 was pumped. This roof was a spectrally discriminating filter that transmitted 50–60% of photosynthetically-active radiation and 0–13% of nonactive radiation. During 5 periods of 1–4 days from September, 1979, to July, 1980, data were collected on environmental conditions: solar irradiance, air temperature and humidity, and windspeed; and on greenhouse conditions: temperature and humidity of the air, temperature of the crop, the roof, the fluid storage tank and at four depths in the soil, and net radiation both above the greenhouse and above the crop in the greenhouse. The environmental conditions were used as boundary value inputs in a deterministic, mechanistic, and dynamic simulation model of the greenhouse system. To test the model, its outputs were compared to the corresponding measured values. The model simulated the various temperatures to within 2–3°C and the net radiation to within 20–30 W m −2 . For one of the five tests, the humidity was simulated almost exactly, but the reliability of the humidity data in the other tests was inadequate. The greenhouse tanks stored energy equivalent to 32% of the daily total of solar radiation. Among the 12 test days, this ratio ranged from 23 to 43%. This contrasts with a figure of 15% for energy storage by a conventional greenhouse, which we have found by simulation of the test structure with an empty roof.

A simple, rapid, semi-automated digitization method was developed for a microcomputer to aid in the collection and analysis of data obtained from an inexpensive, manual analogue recording penetrometer. Soil strength measurements were taken in a Norfolk loamy sand at Florence, SC using a hand-operated recording penetrometer with a 30°, 12.5 mm diameter cone tip. Three probings were taken at each position within each plot and recorded by pen tracings on an index card. Soil strength readings were taken at 0.1-m spacings across planted rows. Data were read from the cards for each 50-mm depth interval by placing them under the control of a microcomputer and the program “DIGIT”. Once initialized for the soil surface, the program automatically moved an eyepiece in the horizontal direction to the position on the card for a specific depth. The operator then moved the eyepiece vertically over the tracings and recorded the data via the digitization process using a flat-bed plotter and microcomputer. These data were averaged over replicates and plotted as isostrength contours using a locally written interpolation plotting routine.

This paper describes a simulation model of the energy flow and of the corresponding temperatures and humidities in a glasshouse and in a fluid-roof solar greenhouse, compared under several representative daily weather patterns. The model includes an account of soil heat storage and of the action of a simplified crop canopy in regard to radiation absorption, heat dissipation, transpiration and stomatal movement. The fluid-roof greenhouse concept embodies a selectively absorbing aqueous solution that slowly flows through a transparent, hollow roof to a storage tank and back. It has been suggested that such a device could be used to store unneeded day-time solar energy for later use, thus reducing day-time heating of plants and external energy needs for both night-time heating and day-time ventilation of the greenhouse space. Our calculations show this to be true for both winter and summer in South-Central Texas and in South-East France, near the Mediterranean. Reductions in heating requirements of 20–40% and virtual elimination of the need for forced ventilation are predicted. The theoretical calculations justify additional engineering and horticultural research to exploit the fluid-roof solar greenhouse concept in a practical way.