Sixty berries per replication were then wrapped together in two layers of cheesecloth and squeezed with a hand press to obtain a composite juice sample. The juice was used to determine soluble solids concentration with a temperature-compensated handheld refractometer and expressed as a percentage. Twenty-one hundredths of an ounce of the same juice sample was used to determine titratable acidity with an automatic titrator and reported as a percentage of citric acid. Some samples that had a high viscosity were centrifuged with a super speed centri-fuge at 15,000 rpm for 5 minutes, in order to get liquid juice for soluble solids concentration and titratable acidity measurements . The ratio of soluble solids concentration to titratable acidity was calculated.Antioxidant capacity was measured in the 2005 and 2007 seasons. Eighteen hundredths of an ounce of berries per replication was used to determine the level of antioxidants by the DPPH free-radical method . Samples were extracted in methanol to assure a good phenolic representation, homogenized using a polytron and centrifuged for 25 minutes. The supernatant was analyzed against the standard, Trolox, a water-soluble vitamin E analogue, and reported in micromoles Trolox equivalents per gram of fresh tissue .An in-store consumer test was conducted on ‘Jewel’, ‘O’Neal’ and ‘Star’ blueberry cultivars in 2006, and on the six blueberry cultivars studied in 2007, using methods described previously . The fruit samples were held for 2 days after harvest at 32°F prior to tasting. One hundred consumers who eat fresh blueberries, representing a diverse combination of ages,black plastic nursery pots ethnic groups and genders, were surveyed in a major supermarket in Fresno County. Each consumer was presented with a sample of each blueberry cultivar in random order at room temperature, 68°F .
A sample consisted of three fresh whole blueberries presented in a 1-ounce soufflé cup labeled with a three-digit code. At the supermarket, the samples were prepared in the produce room out of sight from the testing area. For each sample, the consumer was asked to taste it, and then asked to indicate which statement best described how they felt about the sample on a 9-point hedonic scale . Consumer acceptance was measured as both degree of liking and percentage acceptance, which was calculated as the number of consumers liking the sample divided by the total number of consumers within that sample . In a similar manner, the percentage of consumers disliking and neither liking nor disliking the sample was calculated.Agricultural managed aquifer recharge is a recharge technique for groundwater replenishment, in which farmland is flooded during the winter using excess surface water in order to recharge the underlying aquifer . In California, for example, Ag-MAR is currently being implemented as part of the efforts to mitigate California’s chronic groundwater overdraft . Ag-MAR poses several risks for agricultural fields and groundwater that may influence its future adoption. This includes crop tolerance to flooding, soil aeration, bio-geochemical transformations, long-term impact on soil texture, leaching of pesticides and fertilizers to groundwater, and potential greenhouse gas emissions. Some of these issues have been addressed in recent studies of Ag-MAR, including soil suitability guidelines , nitrate leaching to groundwater , crop suitability and soil aeration . In the current study, we focused solely on the question of “how long can water be applied for Ag-MAR with minimal crop damage?”, while ignoring some of the above-mentioned challenges involving Ag-MAR implementation. Preferably, Ag-MAR flooding is done during fallow or dormant periods, when crop damage is potentially minimal, so agricultural lands can serve as spreading basins for groundwater recharge. Root zone residence time is defined as the duration that the root-zone can remain saturated during Ag-MAR without crop damage . RZRT is a crucial factor in Ag-MAR, as long periods of saturated conditions in the root-zone can damage crops due to oxygen deficiency or complete depletion of oxygen , which ultimately may result in yield loss . However, flood tolerance among crops varies considerably due to biotic and abiotic conditions , therefore only appropriate crops under specific conditions may be suitable for Ag-MAR application.
For example, Dokoozlian et al. have found that grapevine during dormancy can be flooded for 32 days each year without yield loss. Dahlke et al. recently investigated the effect of different Ag-MAR flooding schemes on established alfalfa fields. Results suggest a minimal effect on yield when dormant alfalfa fields on highly permeable soils are subject to winter flooding. On the other hand, some crops are sensitive even to short-period flooding. Kiwi vines for example, are highly sensitive to root anoxia with reported yield lost and vines death due to extreme rainfalls and/or shallow groundwater levels . In a study on peach trees, flood cycles of 12 h per day with 5 cm ponding, applied for two months, resulted in branches with lower diameter and length growth, as well as smaller, low-quality, fruits, compared to the control trees . The above examples demonstrate the need for an RZRT planning tool that can estimate Ag-MAR flood duration with minimal crop damage. Usually, when Ag-MAR water application starts, aeration of the rootzone will be quickly suppressed by a water-layer covering the soil surface, as it prevents oxygen transport to the root-zone in the gas phase. When water application ceases, re-aeration of the root-zone will depend on the soil’s drainage rate that controls the formation of connected air pores between the root-zone and atmosphere . Hence, proper estimation of the planned flood duration during Ag-MAR requires prior knowledge of both crop characteristics and soil texture. Only a few attempts for estimating RZRT during Ag-MAR were made, as Ag-MAR is a relatively new MAR technique. O’Geen et al. used a fuzzy logic approach to rate the RZRT during Ag-MAR, based on the harmonic mean of the saturated hydraulic conductivity of all soil horizons, soil drainage class, and shrink-swell properties. Their RZRT rating was combined with other factors generating a Soil Agricultural Groundwater Banking Index . Flores-Lopez et al. proposed a root-zone model that includes crop type, soil properties, and recharge suitability to estimate water application, flooding duration, and the interval between water applications.
Their model was integrated with a Groundwater Recharge Assessment Tool to optimize Ag-MAR water application. Here,greenhouse pot we propose a simple model to estimate the planned water application during Ag-MAR based on the following parameters: soil texture; crop saturation tolerance; effective root-zone depth; and critical water content. The concept of critical water content was proposed by several authors as it indicates a percolation threshold where the gas transport path is blocked by pore-water, which results in gas diffusivity and permeability of practically zero. Hence, when the water content is either below or above this threshold, gaseous oxygen transport into the soil is blocked or opened, respectively . As opposed to the previous Ag-MAR models mentioned above, our proposed model is physically based and includes explicitly the soil water content, that is used to infer the soil aeration status. Yet, thanks to its simplicity, this model can be integrated easily into various existing Ag-MAR assessment tools such as SAGBI or GRAT . In the following, we first describe the theory of the model and the methods used to test the model performance. Next, we present the model predictions and compare them with observations and numerical simulations. Last, we present an example of how to calculate Ag-MAR water application duration and we discuss the applicability of the model and its limitations. Plant tolerance to flooding or the duration of flooding with minimal crop damage is a very challenging parameter to estimate. A tremendous diversity of tolerance exists, which depends on several factors: soil texture and chemistry; degree and duration of hypoxia/anoxia; soil microbe and pathogen status; vapor pressure deficit , and root-zone and air temperatures; plant species, age, stage and season of the year; and plant adaption as a result of prior climate and soil conditions . An estimate of crop tolerance to flooding of common perennial crops is provided in Table 3, which is an extended version of a previous survey . Annual crops were not included in Table 3 because it was assumed that these fields usually would be fallow during winter and spring when excess surface water is available for Ag-MAR . Waterlogging tolerance in most fruit trees is mainly determined by the root stock and not by the scion , where tolerance is higher during dormancy, but more prone to damage during bud break and growth . The plant tolerance scales in Table 3 have different definitions, as some authors use plant survival as the tolerance criterion, while others consider economical damage as the tolerance criterion; these differences are indicated in Table 3. We note that the data provided in Table 3 should be used with caution because most of it is based on expert opinion or experiments with seedlings, while very few waterlogged experiments were conducted with bearing fruit trees. The fit between the predicted and observed effective saturation ranges from poor to excellent , and generally the fit is better for the RZRT models that underwent calibration followed by Hw1 and H5w. Obviously, a better fit of the predicted and observed water contents will lead to a more accurate estimation of twap.
Therefore, when possible, it is recommended to use the proposed RZRT model with site-specific hydraulic parameters. This is demonstrated in the Yolo silt loam soil, where a reasonable fit was not feasible without the use of site-specific hydraulic parameters . Note that site-specific parameters can vary considerably for the parameters obtained from the NCSS database. This is especially notable for the Ks values which can vary by more than one order of magnitude . The reason for this discrepancy is attributed to the low spatial representation of each soil series in the NCSS database, which is based on few soil pedons that are not always a good representation of the soil series where the field data was collected. In some cases, even when the overall effective saturation fit is poor, it is possible to estimate twap accurately given that the fit is good at the range of Sc. This is demonstrated in the Harkey loam for the H5w-fit . Note that for all soils H1w performs better than H5w, supposedly not as expected, because Rosetta3 is a hierarchical PTF where the highest hierarchy should perform better than lower hierarchies . As noted above, this is because each of the H5w parameters in this study was based on only one soil pedon sample from a specific location , which in this study was less representative compared to the Hw1 parameters that are based on averaging by soil texture a large number of soil samples. The fit of the effective saturation between the proposed RZRT model and the numerical model HYDRUS-1D ranges from good to excellent , and in all cases, HYDRUS fits better to the RZRT model than to the observed data. This indicates that the deviations between the RZRT model predictions and the observed water content data are probably due to soil layering, soil heterogeneity, and preferential flow, which cannot be captured by simplified homogenous one dimensional flow models. Another explanation for the observed and modeled water content deviations could be an inappropriate setting of the models’ boundary conditions. This mainly refers to the assumption of free drainage at the bottom boundary because the top boundary was controlled during the experiments. The performance of the RZRT model with a hard pan layer above or below the effective root zone was compared to HYDURS simulations with similar settings . According to our limited test, the RZRT model with the harmonic mean Ks is preferred during the infiltration period. During the drainage period the arithmetic or the harmonic mean Ks are preferred when the hard pan layer is above or below the effective root zone, respectively. As expected, when the hard pan layer is far below the effective root zone there is no impact on the effective root zone by the hard pan. The total water applied calculated with Eq. and the harmonic mean Ks is almost identical to the HYDRUS results . This demonstrates the impact of a hard pan on deep percolation, as the total water amounts applied were reduced by more than half when the hard pan layer was close to the root zone.