Mitigating fertility effects of flooding with variable rate amendment

Final Report for ONE12-158

Project Type: Partnership
Funds awarded in 2012: $14,999.00
Projected End Date: 12/31/2013
Region: Northeast
State: Vermont
Project Leader:
Dr. Josef Görres
University Of Vermont
Expand All

Project Information


Climate change predictions forecast more extreme and frequent precipitation events. This can lead to higher frequencies of and more prolonged floods. This research aimed to assess the changes in soil fertility distribution patterns with the goal of providing nutrient management recommendations for flood-prone areas on a gradient. In partnership with the Intervale Community Farm, a field depression was identified on currently cultivated, flood-prone farm land. Soil samples were taken on seven sampling dates in three categories (Figure 1). The Post-Irene categories contained the sampling dates after 2011 in May 2012 after snowmelt and in October 2012 after rye cover crop was seeded. The Peak category contained three sampling dates taken within a month after a peak rainfall, snowfall, or river height event on two dates after snowmelt in May 2013 and May 2014 and after the wettest two consecutive months in Vermont history in June 2013. The Non-Peak category contained the sampling dates in August 2013 and June 2014. Soil fertility parameters pH, and modified Morgan extracted phosphorus, potassium, calcium, magnesium, manganese, iron, sulfur, organic matter, and cation exchange capacity were analyzed for correlation with small changes in elevation and means were compared among categories.

There was a clear inverse linear relationship between elevation within the depression and soil moisture on 6 out of seven sampling dates (Table 1 of attached file). Similarly, on peak events P was inversely related to elevation (Figure 3 and Table 2). Iron was also inversely related to depression elevation (Figure 2) on 4 out of 7 sampling dates. Manganese concentrations were most consistently inversely related to elevation across the four depressions in May 2014 but not in June 2004 (Tables 6,7). Only Peak events had inverse relationships between elevation and MME-manganese (Table 5).

Farmers who have flood plain fields should consider that depressions have greater P, Fe and Mn availability after seasonal flooding events. However, we also observed that depressions seem less productive after seasonal flooding events. The depression don’t have to be very deep to have a negative effect. The discrepancy between post flood soil fertility levels as measured by modified Morgan method and crop productivity points to potential other problems such as lower infectiousness of mycorrhizae, conditions that are wetter for longer, and  that the modified Morgan nutrient concentrations do not reflect availability in the field under these wet conditions. One of the recommendations for post flood syndrome (i.e. lower productivity of flooded soils) is to plan for floods by using resilient cover crops. In theory this will provide sufficient hosts for mycorrhizae to overcome reduced productivity. Alternatively, less intense use of small flood plain depressions may improve a sustainable production.


In the northeastern USA flooding is projected to become more frequent as climate change is forecast to bring more frequent and intense rain storms (NECIA, 2006). Historically, flooding brought nutrient inputs, with beneficial effects on fertility. However, in the managed, modern agroecosystems, flooding can reduce soil fertility. Erosion, leaching, compaction and oxygen depletion reduce fertility (Manitoba Food, Agriculture and Rural Initiatives). Soil compaction caused by flooding may make it difficult for plant roots to penetrate the soil. Flooding may shift the microbial community towards anaerobic physiology increasing denitrification, reducing available N and reducing decomposition and mineralization. This will improve sustainability of organic agriculture by reducing nutrient inputs, thus lowering cost and environmental impacts.

Organic farmers are particularly affected by flooding because most organic amendments release nutrients slowly. Fertility amendments to remedy plants needs immediately are expensive and some are no longer certified as organic. But even the slow release of nutrients may be hampered because flooding favors anaerobic soil organisms.

Recommendations based on soil tests have been widely accepted as best farming practices for ecological and economic management of soil fertility. A single soil test is usually done for fields several acres in size from composite sample of 10 or more randomly taken samples (Darby, Hills and Bosworth, 2009). A uniform recommendation is given for that area based on the composite soil sample. This level of spatial detail is not sufficient to take into account the spatial variability caused by flooding. Raised water tables affect low lying areas more frequently and for longer than higher elevations thus fertility impairments will differ across the landscape. At $14 per test, fertility soil tests are too expensive to obtain fertility information with sufficient resolution. One of our objectives was to evaluate the spatial distribution of fertility with the goal to identify zones of fertility related to water tables/topography so that amendments could be applied at variable rates to counterbalance the effect of flooding.

The recent multiple flooding events of the farms in the Intervale in Burlington, Vermont, are an opportunity to assess flood damage and its effect on next year’s crop. They are also an opportunity to experiment with variable rate application of organic amendments that consider the fertility differential between areas that were inundated and those that were not. While different types of flooding events occurred we will only address problems associated with raised water tables affecting depressions during the spring when large rainfalls/snowmelts are expected. Catastrophic floods during extreme river stages like caused by extreme events such as Hurricane Irene trigger food safety concerns beyond the scope of this grant.

Project Objectives:

Goal 1: Correlate spatially referenced soil fertility tests (soil tests) with

  1. electrical conductivity measurements as a potential, more affordable fertility assessment tool. This method was shown to be unreliable in overly saturated soils as the EC meter was unable to accurately measure during super saturated conditions and impossible to measure EC under conditions with moisture contents of less than 15% by volume.
  2. elevation and seasonal high water table as a proxy for flooding likelihood. For this site, the height of the gage where Winooski River meets Essex Junction (12 feet) is an accurate predictor for when the Intervale floods.
  3. nutrient deficiency symptoms as a biological indicator of fertility needs. The crops planted during the duration of this study did show visual signs of stunted growth at the bottom of the depression in the beginning of the season that often was imperceptible mid-season. This initial delay may become problematic if weather extremes worsen in intensity or frequency.


Goal 2: Explore spatially referenced fertility data by combining the data to give averages for

  1. an entire depression
  2. elevation referenced zones, such as bottom, mid-slope and top of depression
  3. by contour intervals

The above goals were completed with linear regression analysis (see Results and Discussion section) and soil maps produced used geostatistics (Figures 1 -3 of attached File).

 Goal 3: Identify spatial patterns of nutrient needs. Specifically, we will address the question whether proxies for soil tests (Goal 1) can guide fertility recommendations. Overall, the changes within the depression were complex and would be time and labor intensive to manage individually. However, the concentrations of some elements did widely fluctuate with severity of water event (see Results and Discussion section and he Tables in the File uploaded to support the Summary)

 Goal 4: Prescribe and apply amendments in accordance with the spatially distributed nutrient needs. The soil test results within a depression on a given sampling date were often complex making it difficult to recommend amendments in accordance with spatial distribution. Further analyses still in progress appear to predict P based on the distribution of organic matter, elevation and redox active elements such as Mn and Fe whose minerals are closely associated with MME-P.


Click linked name(s) to expand/collapse or show everyone's info
  • Lindsey Ruhl


Materials and methods:

Site Description

The field site was selected to meet the following criteria: currently under organic agricultural vegetable management, located on a floodplain, flood annually, and have naturally occurring depressions at least six inches deep with a minimum of 60 feet in diameter. A field matching this criteria was located in Burlington, Vermont at the Intervale Community Farm (ICF) in an area colloquially known as The Intervale. This particular ICF field was on Limerick silt loam (mesic Fluvaquentic Endoaquepts). This soil series is a poorly drained alluvial soil, is in a broad depression, on a 0-3% slope, is coarse-silty, and is superactive (high CEC). Limerick silt loams have a seasonal high water table 0.0-1.5 feet deep, flood briefly and frequently, and therefore have very limited land use.

The average precipitation for this area is 44 inches. However in 2011, the research site (Burlington, Vermont) received above average levels of precipitation with a total of 59.92 inches of rainfall and 112.10 inches of snowfall (NOAA, 2014). The growing season was shortened on both ends. In late April, snow melt and record rainfall delayed the growing season with flooding. Four months later, in August, Tropical Storm Irene deposited enough rainfall to be the second largest flooding event of the 21st and 20th centuries. The flooding from Tropical Storm Irene was caused by the the Winooski River itself jumping the banks, which is a rare event. The study site is unique in that it is in a flood-prone field where the most flooding is caused by a high water table, the source of which is water seeping up from the lower lying wetlands.

The data was categorized into three types of events: Post Irene, Peak and Non-Peak. Post-Irene category includes sampling dates in the agricultural season following prolonged snow melt and flooding from Tropical Storm Irene of 2011. Peak water events were defined as one of the top 12 greatest rainfalls, snowfalls, or Winooski River Gage height at Essex Junction in the 30 month sampling period between January 1rst, 2012 and June 15th, 2014. The Advanced Hydrological Prediction Service defines flood categories as action stage at 10 feet and flood stage at 12 feet for The Intervale in Burlington (NWS, 2014).

The Peak category included sampling dates that fell within a month after a peak water event (April 2013, June 2013, and May 2014). The Peak sampling dates all fall within a month of a gage height peak. April 2013 and May 2014 followed snow melt events. Sampling in June 2013 occurred during the two consecutive wettest months recorded in Vermont. Non-Peak sampling dates (August 2013 and June 2014) were after a month of a Peak event. The Post-Irene data set includes samplings dates in May 2012 and October 2012. Table 1 lists the for dates and their categorization. Figure 1 shows weather conditions, agronomic practices, and sampling dates. Throughout this paper the remaining tables will have a date column with superscripts that indicate category, ◊ is First Year After T.S. Irene, ‡ is Peak, and † is Non-Peak. In the mean column, superscript indicates an ordered letter report where means sharing common letters are statistically similar.

Field Management, Environmental Conditions, and Sampling Dates

        During peak harvest in the fall of 2011, the floodwaters of Tropical Storm Irene inundated the research site for over 24 hours. As a result of field conditions, minimum management occurred. Cover crops were not planted. The following year, in late April 2012, the research site was chisel plowed and on May 30th, a variety of storage squash was transplanted. May 2012 samples were taken after the field was plowed, but before squash was transplanted. The field was mechanically cultivated throughout the season as needed and on September 20th winter rye was seeded as a cover crop. Samples were taken in October 2012 when the cover crop was just emerging with a height of two to three inches.

After snow melt in April 2013, soil samples were collected. In mid-June 2013, the winter rye was cut, harvested, and then disced. Sampling in June occurred after the rye was cut, but before it was disced. The June 2013 samples followed the wettest two consecutive months recorded in Vermont. May and June received a higher amount of precipitation than any other two months in Vermont’s recorded history. Another set of soil samples were taken under dry conditions in August 2013. Between September 27th and October 4th, stands of wheat and rye were seeded, leaving a section on the end of the field free of seed. On April 3rd 2014, the research site was frost seeded with Alsike clover at 12-15 pounds per acre. Thirty-five days later, the May soil sampling initiated. One month from the start of the May sampling, the June 2014 soil samples were collected. Despite the 2013 and 2014 cover crop seeding, the research site remained dominated by knot weed during both 2014 sampling dates.



Experimental Procedures

Sampling dates, number of samples, and other parameters surrounding soil sample analysis are given in Table 2.0. During each sampling date, soil samples were collected from randomly selected locations within a single depression. During the last two sampling dates in May 2014 and June 2014, soil samples were collected in three additional depressions and randomly selected points between depressions (Scattered set). The undisturbed soil samples were taken by forcing 5-cm diameter X 5-cm long plastic core sleeves into the soil. Cores were extracted from the soil with a trowel, placed in a bag and put in a 4°C walk-in freezer and and dried within three days with the exception of October 2012 which was kept in the cooler for 16 days, June 2013 which was kept in the cooler for 12 days, and August 2013 which was kept in the walk-in cooler for 40 days. Topographic survey was used to determine relative position and elevation of each sampling point. Soil samples were analyzed for pH, moisture, organic matter, and fertility with Modified Morgan’s extract with the ICP (inductively couple argon plasma spectrometry). Nutrient results were used to calculate cation exchange capacity. Soil samples in cores were extracted with a trowel, placed in a bag, and put in a 4°C walk-in freezer


Laboratory Analysis


June 2013 samples were measured for pH with the same method the Maine Laboratory utilized. Five (5) cm3 of ground soil was placed in 10 mL of 1M CaCl2 in deionized water and stirred. After 30 minutes, the Accumet Basic AB15 pH meter was utilized to measure pH. There was not enough soil to remeasure April 2013 and August 2013 with the salt buffered method. Instead, a Mettler Toledo SevenEasy pH meter was used to measure pH. A 1:10 ratio of dry soil to water was used. A standard weight of 4.0 grams of soil was placed 40 mL of reverse osmosis (R.O.) water, stirred, and then measured after 10 minutes. Both pH meters was calibrated using 4.0 and 7.0 standards. Soil was dried, ground, and sieved through 2 mm mesh.

Moisture and Organic Matter

Mass moisture and organic matter were measured gravimetrically as a percentage of oven-dry weight. To determine mass moisture, fresh soil samples were weighed and then oven-dry-weight of the soil samples were determined by drying at 55°C in a General Signal Blue M Electric oven until dry. The dry weight was recorded and the following equation was used to determine moisture: : .

 Subsequently, soil organic matter (SOM) was measured by loss on ignition (LOI), following the same standard procedure recommended by the The Northeast Coordinating Committee for Soil Testing in both laboratories (Schulte, 2011). Between five and ten grams of dry soil were placed in crucibles in a furnace for two hours at 110°C and weighed. Then heated at 375°C for two hours, cooled to 105°C, then weighed. Percent SOM was calculated with the following equation: .

Soil Fertility

        Extraction methods followed procedures recommended by The Northeast Coordinating Committee for Soil Testing (Wolf, 2011). Hereafter, concentrations of elements are reported in terms of results produced by Modified Morgans extraction. Soil samples were dried at 55°C in a General Signal Blue M Electric oven until dry, ground with pestle and mortar, then sieved through 2 mm mesh. Soil was weighed on a Metler Toledo PL303 scale to 4.000 grams. Soil was placed on a rack of 12 Erlenmeyer flasks. Modified Morgan’s extract (ammonium acetate, pH 4.8 +/- 0.05) was added at a volume of 20 mL. The extracts were then shaken on an Eberbach reciprocal shaker for 15 minutes.

The extracts were filtered through 9m Ahlstrom Filter Paper into funnel tubes. Extracts were filtered a second time if the extract contained sediment. Filtered extracts were placed into ICP tubes to measure concentrations of macro and micro nutrients (available phosphorus, potassium, magnesium, aluminum, calcium, zinc, sulfur, manganese, boron, copper, iron, sodium) on the inductively couple argon plasma spectrometry (ICP). To get average mg/kg of an element in the soil, the ICP measurement was multiplied by five (the dilution factor of the extract). Soil fertility for sample sets in 2013 were measured in this manner. Cation exchange capacity was calculated from the ICP analysis results using the following equation: (Ross and Ketterings, 2011). The denominator is the equivalent weight and is determined by dividing the atomic weight by number of valences.


Analytical Analysis

Surveying Calculations

        A Northwest Instrument Inc. NSL500B transit level was used to measure location (x,y) and elevation (z). There are three sets of measurements that the surveying equipment employs, using the instrument position as origin. The difference between high and low stadia on the rod, multiplied by 100 measures the distance from origin (instrument position). The angle (degrees and minutes), is turned into radians to help determine the x, y coordinates. The following equation was used to calculate the x coordinates: distance*cosine(radians). The following equation was used to calculate y coordinates: distance*sine(radians). The center stadia on the rod measures elevation (z). These measurements were recorded for sampling dates in 2012 and 2013. In 2014, a tape measure was used to measure distance. In July 2014, sampling locations were not resurveyed. Instead, soil samples were taken within 5 cm along similar elevation of May 2014 sampling locations.

Data Analysis

        All statistical calculations were performed using JMP Pro 11.0 (SAS Institute Inc.). Normality of distribution of moisture, pH, various elements, organic matter, and CEC by elevation for each sampling date were not normal as determined by testing the residuals by the Shapiro-Wilk W Continuous Goodness-of-Fit test (p<0.0001). As such, to determine if data could be defined by categories, Wilcoxon/Kruskal-Wallis, was first used, followed by nonparametric comparison each pair using the Wilcoxon method to identify mean concentrations by dates that may be similar to each other. Linear regression analysis was employed to calculate correlation of pH, various elements, organic matter, and CEC by elevation concentrations.  Significance of correlations were determined at p<0.0001.

            Field Patterns were analyzed using semivariograms and kriged maps of soil properties

Research results and discussion:


Moisture correlated with elevation on six of the seven sampling dates (Table 1, File uploaded with summary). The overall trend was an inverse relationship with higher percent moisture at the bottom of the depression than at the top. Various soil fertility parameters were tested for correlation with elevation and not moisture because moisture measurements do not necessarily measure length of saturation. However, elevation can serve as a proxy to distance from water table and therefore length of saturation which can influence soil chemistry such as loss of nutrients through leaching and increasing the cycling of redox reactions as the soil fluctuates between dry and wet conditions.

Although mean percent moisture was significantly different among the categories, mean percent moisture does not necessarily indicate length of saturation. For example, mean percent moisture may be similar in events that cause soil saturation, such as a flood that lasts 10 days or a heavy down pour lasting four hours. Overall, as there was such strong correlation between percent moisture and elevation, elevation can serve as proxy for length of saturation and mean moisture was lowest in the Non-Peak category.



Limitations on plant-available elements can be set by pH concentrations. Overall, there was little correlation of pH with elevation as pH correlations with elevation only one date, June 2014 (Table 4). However, pH was influenced by severity of event. The data suggests that the large snowmelt of 2011 followed by the flooding from Tropical Storm Irene, influenced pH by increasing it by an average of 0.8 pH. The change in pH among the categories indicates both that consecutive extreme events can influence pH and that pH may decrease under more typical weather conditions and usual agronomic practices.



Phosphorus is a macronutrient used for synthesis of energy. Plants require large amounts of macronutrients than micronutrients like trace metals. Phosphorus correlated with elevation on four of the seven dates with as much as 24% of the variation in phosphorus explained by changes in elevation (R2), regardless of severity of weather event as there were correlations in each of the three categories (Table 2, Summary). The statistically significant linear relationship of phosphorus with elevation was inverse with higher concentrations of phosphorus at the bottom of the depression than at the top.

Although phosphorus concentrations were statistically lower in the Post-Irene category, the difference was 0.46 ppm, a negligible amount in terms of changing agronomic practices to accommodate the difference. Given the inconsistent trends of the data (correlations with elevation or among categories), it appears that other soil properties also influence  phosphorus concentrations more than small changes in elevation. for example in June 2014, 88% of the variation of P was explained by pH, Ca, Mg, and Mn when executing a step wise regression with Mg and Ca explaining most of it. Only Magnesium had a (direct) linear relationship with elevation. P relationship with Ca was strongly positive (r2 = 0.74, p< 0.001) and with Mg it was negative (r2 = 0.30, p = 0.0012).


Concentrations of potassium correlated with elevation  on the same four dates that concentrations of phosphorus correlated with elevation . Like phosphorus, the statistically significant linear regressions were in all categories, suggesting that severity of weather event did not significantly different from each other, as is also evidenced by lack of statistical significance among the categories using the nonparametric Wilcoxon each pair comparison method.



Magnesium (ppm) had direct relationships with elevation on three of the sampling dates. There was statistically significant relationship in each of the categories (Table 7). Given that magnesium correlated with elevation on less than half the sampling dates and the dispersal of linear correlation into all three categories, there was relatively low or consistent correlation of magnesium with elevation or severity of water event.



Calcium concentrations correlated with elevation on two of the seven sampling date with higher calcium concentrations at the bottom of the depression than at the top (Table 8). The mean calcium concentration was significantly lower in the Post-Irene category. However, the lowest concentration of calcium was in the Peak category. Overall, the data suggests that there are other parameters that more directly influence calcium concentration than small changes in elevation or severity of water event.



Manganese concentrations correlated with elevation on three of the seven sampling dates with higher manganese at the bottom of the depression than at the top (Table 9). Furthermore, the mean manganese concentrations among the categories were statistically different from each other. The mean manganese concentration in the Post-Irene category was statistically lower. This may be due to redox potential or the sensitivity of manganese to fluctuations between saturated and unsaturated conditions. These low levels were potentially due to leaching during the 2011 heavy spring snow melt and Tropical Storm Irene floodwaters. The highest levels of mean manganese concentrations were in the Peak category perhaps from the shorter saturation periods causing ‘hotspots’ or statistical outliers of reduced (plant-available) manganese. Overall, manganese concentrations showed the most sensitivity to severity of saturation event.



Iron concentrations correlated with elevation (Table 10) on the same three sampling dates that manganese correlated with elevation (Table 9). However, there was less fluctuation of iron means among sampling dates or categories than manganese. Iron is also a redox sensitive element. In saturated or reducing conditions when the oxygen supply is limited, there may be more plant available iron. The iron data reinforces what the manganese data suggested, that there were reducing conditions changing the soil chemistry by influencing concentrations of plant available manganese and iron.



Sulfur concentrations correlated with elevation on two of the sampling dates (Table 11). Although mean sulfur concentration was statistically different among categories, the means of sulfur within categories was inconsistent i.e. the lowest means were not in the same category nor were the highest means in the same categories. The data indicates that sulfur concentrations may be influenced more strongly by other factors than small changes in elevation or severity of weather event.



Cation exchange capacity correlated with small changes in elevation on two of the seven sampling dates with lower CEC at the bottom of the depression than at the top (Table 12). The mean CEC was significantly lower in the Post-Irene category indicating that there may have been some loss of Ca, K, or Mg. Overall, there is not sufficient evidence in this study to suggest that CEC consistently correlated with small changes in elevation or severity of water event.


Organic Matter

Percent organic matter correlated with small changes in elevation on two of the sampling dates with less organic matter at the bottom of the depression than at the top (Table 13). This may be due to longer durations of saturation delaying decomposition. However, the average mean of each category were not significantly different from each other, indicating that severity of water event is not the main influence in distributing organic matter. However, there are higher amounts of organic matter in 2013 than during the other sampling dates. This may be due to the additions of organic matter from the rye cover crop.


Spatial Patterns of Modified Morgan Extractable P, Fe and Mn Relative to Elevation with Field Depressions

Generally, elevation and extractable P had large scale patterns with range values (distances over which the values were autocorrelated) of 10 – 15 feet. In contrast extractable Fe and Mn had range values between 4 to 8 feet. This is expressed in the maps shown in Figure 2 of the Maps upload.


Work Under Objective I

Originally two farms agreed to partner with us on this project. However, due to the high flooding potential, one farmer unexpectedly chose not to cultivate that area. In effect, a replicate was lost with this decision. At the remaining site, soil samples were taken on five sampling dates (after snow-melt in 2012, autumn of 2012, after snow-melt in 2013, after the May/June rains in 2013, and in August of 2013) with 5 cm X 5 cm cores. Surveyor equipment was used to record the location of each soil sample.

 The first set of samples was sent to Maine for analysis. In an effort to extend funds for more testing, the remaining soil fertility tests were carried out in UVM’s AETL lab with KCl extract on the Lachet for inorganic nitrogen, other soil fertility elements with ammonium acetate extract on the inductively coupled argon plasma mass spectrometer (ICP), and organic matter by loss on ignition. The August 2013 samples will be analyzed by the ICP in 2014. Active carbon was completed for the first set of samples and the remaining have yet to be analyzed. Electrical conductivity was taken on two sampling dates. Dates omitted are in October 2013 due to equipment malfunction and June 2013 due to time constraints. Measuring pH will be finished in 2014.


In the summer of 2012, height measurements of the squash crops were recorded for one row along the depression’s gradient on two sampling dates. Leaf samples were taken for tissue analysis. That data has yet to be analyzed.


Although some analysis has been conducted on correlation of iron, phosphorus, and potassium with sampling date ie level of saturation, more is needed. In 2014, data collected will be analyzed using JMP software for ANOVA and GS+ to correlate fertility data with elevation.


Work Under Objective II

A section of analysis was included in the handout “Cover Crops to Cope with the Effects of Flooding on Soil Fertility” along with information on general soil fertility loss and using cover crops as a remediating tool for soil damage caused by flooding.

This brochure was handed out to 15 participants during the field day by the same name. This field day was in conjunction with NOFA’s new fall workshop series and was advertised on their website.1 We attemtped to obtain CIG, USDA-NIFA support for further studies of the intriguing effect of small agricultural depressions on soil fertility. We stand a better chance now that we have analyzed most of the data and will resubmit using alliances with Extension and NOFA . 1To see advertisement on NOFA webpage:


Work Under Objective III:

Visual comparisons were made between maps of elevation, available P, extractable Fe and Mn for four field depressions for the 2014 sampling dates (figures 1 to 3 in the Maps upload). These data will be presented in a journal paper.



May 2012: Case and Maden advise on suitable research areas. Soil sampling for fertility will be conducted early in May by Görres and Ruhl. An aliquot of each sample will be sent for soil testing, the remainder of the soil will be analyzed in the lab for organic matter, and active carbon (Ruhl). After these tasks are done, the fields will be surveyed and additional field tests such as electrical conductivity and moisture will be carried out by Görres and Ruhl. Farmers will receive fertility recommendations (from graduate student and PI) and plan their planting and fertilizer schedules accordingly. Farmers will apply nutrients.

One farmer unexpectedly did not cultivate the agreed upon area. This decision was not communicated in a timely manner to the researchers in order to make any possible adjustments like identifying another depression on the farm to study.


May 2012 – April 2013: Statistical analysis will be conducted (Ruhl, Görres)

Some analysis has been completed, namely elevation, element, and sampling date correlations for iron, potassium, phosphorus. Some inconsistencies with data collection make analysis difficult. For example, the Watermark Sensor with one sensor that records temperature and seven sensors that record saturation six inches underground, were not present or working properly throughout the entire duration of the study. In other words, some dates have temperature and soil moisture tension data to correlate with soil fertility and others do not. Precipitation data has been used as a proxy for saturation.  

Furthermore, different types of crops were present and at different stages of growth which can alter length of saturation. If there are more roots present, infiltration rate may increase and water uptake may increase resulting in a reduced length of saturation. Decomposition may also minimize the effect of elevation and length of saturation on fertility distribution patterns by adding or reducing an amount substantial enough to have a leveling effect on the amount of available nutrients. Tilling and other cultivation practices mixed soil and may have had a homogenizing effect on soil fertility patterns.

Furthermore, the first sampling date had 60 samples and was analyzed at a lab outside of UVM. The other sets of soil samples had 30 soil samples per sampling date and were analyzed in UVM’s AETL lab by Ruhl. The method of analysis between labs uses a soil standard to test for comparable results. Both labs are consistent with measurements. However, having more samples allows for a greater potential of finding more correlations. As a result, it appears that the first sampling date has a higher rate of correlation with elevation, sampling date, and element when in fact this could be more of a reflection of the inequality in sample size.


June 2012: Farmers will plant their crops.

The Intervale Community Farm transplanted squash by the end of May 2012. Originally, this land was to be left fallow. However, in the October of 2012, rye was planted as a cover crop. This cover crop remained until harvest in June 2013. The land was disced, and planted with winter wheat and winter rye in fall 2013.


June 2012 – October 2012: Case, Maden, Gorres and Ruhl will assess crop health using visual cues for nutrient deficiencies. Yield of crop in plots will be assessed.

Plants were visually assessed for health. One row of squash plants was measured for height at two sampling dates and leaf tissue was taken of analysis. Analysis of this data is pending. Yield was not assessed due to incompatible farmer and researcher schedules during harvest time.


October 2012: Soil sampling and testing to assess residual nutrients after harvest (Gorres and Ruhl). See May 2012 for more detail.

Thirty soil samples were taken in October, locations recorded with surveyor equipment, and analyzed for nutrients.


October 2012 – April 2013: Görres and Ruhl will write grants to NIFA-AFRI, CIG and SARE to continue this work.

            Grants to the aforementioned institutions were written with fruitless results.


May 2012 – April 2013: Gorres and Ruhl will write Farmer’s Guide to Fertility Patterns in Flood Prone Fields

Some results have been shared in the Cover Crops to Cope with Effects of Flooding on Soil Fertility brochure. In lieu of creating the brochure, Lindsey Ruhl’s thesis will be uploaded to the SARE website in February 2014.


April 2013: Soil sampling and testing to assess carry over after harvest (Görres, Ruhl). See May 2012 for more detail.

Thirty soil samples were taken in April, locations recorded with surveyor equipment, and analyzed in UVM’s AETL for nutrients.


April 2013: Field day at an Intervale farm or the Intervale Center (Gorres, Ruhl, Case, Maden).

The field day was held in September of 2013. Fifteen participants attended including farmers and university personnel. This field day was a part of NOFA’s fall workshop series and also included information from NESARE Graduate Grant, Mitigating and Preventing Flood-Related Soil Quality Degradation Using Cover Crop Blends.


June 2013:

Thirty soil samples were taken in June, locations recorded with surveyor equipment, and analyzed for nutrients.


August 2013:

Thirty soil samples were taken in August, locations recorded with surveyor equipment. Soil analysis is complete.


October 2013:

Data was presented in a talk at UVM’s Plant and Soil Science Department Seminar Series.


November 2013:

A poster was submitted to the 2013 International Annual Soil Science Society of America Conference in Tampa, Florida.


February 2014

            Data presented to the Soil and Water Pollution and Bioremediation class at UVM


May 2014:

            Forty-three soil samples were taken in August, locations recorded with surveyor equipment, and samples shipped to main for analysis.


June 2014:

Thirty-three soil samples were taken in August, locations recorded with surveyor equipment, and samples shipped to main for analysis.


November 2014:

A poster was submitted to the 2014 International Annual Soil Science Society of America Conference in Los Angles, Florida.


December 2014

Data presented at Lindsey Ruhl’s thesis defense to an audience of 20 professionals at UVM.

Research conclusions:

This is difficult to assess because we presented part of this work in conjunction with another SARE grant at several workshops. For example Fifteen participants attended including farmers and university personnel. This field day was a part of NOFA’s fall workshop series and also included information from NESARE Graduate Grant, Mitigating and Preventing Flood-Related Soil Quality Degradation Using Cover Crop Blends.

After the presentation of the poster at the 2014 SSSE annual meetings I received several requests from students who wanted to continue this research as graduate students.

A poster was submitted to the 2013 International Annual Soil Science Society of America Conference in Tampa, Florida.

Data presented to the Soil and Water Pollution and Bioremediation class at UVM

Data presented at Lindsey Ruhl’s thesis defense to an audience of 20 professionals at UVM.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary:

Education/outreach description:

A field-season, weekly WordPress blog1 recorded field updates, analysis of results, and agricultural current events. Results were presented in a brochure entitled “Cover Crops to Cope with the Effects of Flooding on Soil Fertility” at a field day in NOFA’s fall workshop series. Results of the three element analysis were also presented at the 2013 and 2014 International Annual Soil Science Society of America Conference. The data was presented to the spring 2014 Soil and Water Pollution and Bioremediation class. Further analysis was presented at Lindsey Ruhl’s thesis defense to an audience of 20 professionals in December 2014.

Project Outcomes

Project outcomes:

Not required

Assessment of Project Approach and Areas of Further Study:

Areas needing additional study

The findings provided useful insight into the effect of water saturation on soil fertility in naturally established eld depressions. Few studies have categorized changes in nutrient availability by severity of water saturation event or within eld depressions. In particular, the timing of the study took advantage of a unique opportunity to measure the potential period of recovery in soil fertility after a major ooding event (Tropical Storm Irene). The data also suggested that deficiencies in soil fertility may be high in the beginning of the growing season and recover within a growing season without additions of fertilizers or amendments.

Additional analyses in particular of yields and economics would help set this study in a perspective appreciated by farmers.  An economic analysis of the impact field depressions can have in size or quantity and the requirements necessary to remediate field depressions would help farmers decide best management practices. This would necessitate further studies into preventative measures and recovery plans. Rate and type of amendments could be used to alleviate nutrient de ciency after a saturation event. Knowing length of saturation and soil type may help farmers make timely decisions of what to apply to elds without additional soil analysis. As a preventative and recovery plan, soils could be amended with mycorrhizae to aid in nutrient uptake to plants. Furthermore, cover crop treatments may reduce saturation impact through higher rates of in ltration and aid in amelioration of nutrient depleted elds. If a eld is pockmarked with eld depressions and repeatedly inundated throughout a season, it may be necessary to conduct a cost benefit analysis to determine if the eld is now on marginalized land where it is no longer economically advantageous to agriculturally manage.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.