On-farm Cyanobacterial Bio-fertilizer Production to Reduce the Carbon Footprint of Organic Fruit and Vegetable Production

Objective 1) Optimize the yield and efficiency of an on-farm cyano-fertilizer production system

Task 1.1 – Determine if CO2 bubbling will enhance growth and N-fixation of cyanobacteria grown in outdoor raceways.
Nitrogen-fixing cyanobacteria are attractive as a nitrogen (N) fertilizer because they are ubiquitous in nature and have minimal nutrient requirements. Our lab is scaling up production of a local strain of the N-fixing cyanobacterium Anabaena sp. in on-farm open raceways to determine its economic potential as a N fertilizer for horticultural crops. Our goal is to increase productivity in an organically certifiable growth medium above the current two week batch production levels of 30 mg L-1 total Kjeldahl N. To improve production, we tested delta wing arrays to improve mixing and mass transfer of nutrients. We also supplemented production raceways with CO2 bubbling to maintain a pH of 9.5 or lower. Using delta wings to improve mixing produced no significant differences in biomass or nitrogen production in CO2 limited cultures. Supplementation of CO2 to limit pH at 9.5 produced small improvements in biomass but no difference in N concentration.

During the CO2 trial, beginning day 7 until the end of the batch, most days showed significant differences in pH between treatment (with CO2 bubbling) and control (no bubbling) raceways. Overcast weather (notably day 13) reduced photosynthesis and pH in both control and CO2 supplemented raceways. Beginning on day 11, there were statistically significant differences in optical density (a measure of cyanobacterial biomass). However, there was no difference between total N in control raceways compared with pH controlled raceways. The increase in optical density in supplemented raceways holds little practical significance, and it would be difficult to justify the costs of bubbling CO2 to maintain a pH of 9.5.

Other nutrients may be limiting growth and N fixation. Increasing pH above 9.0 can render some nutrients, such as iron, unavailable to cyanobacteria. Using raceway production ranges of 9 g m-2 d-1 to 20 g m-2 d-1 as a reference for a marine N-fixing Anabaena sp., it could be reasonable to expect a 50% to 200% increase over our current baseline production. It is possible that increasing CO2 supplementation and reducing pH further to at least 9.0 could lead to cost effective increases in biomass and total N.

In 2015, we identified light and CO2 as the most limiting factors to growth in the field. CO2 supplementation is ubiquitous in microalgae culture. Results from 2015 field experiments confirmed that supplemental CO2 is required to achieve gains in biomass and total N under outdoor production conditions, but we may need to bubble CO2 at a higher rate to achieve a lower pH (9.0) and enhance biomass and N concentration economically.
Summer 2016 field experiments utilized CO2 to maintain a sufficiently low pH to improve dissolved inorganic carbon (DIC) availability in raceway cultures throughout the day, especially when photosynthesis rates are high between 10 a.m. and 2 p.m.

During the summer 2016 field season, we conducted experiments to better manage light availability to Anabaena sp. cultures. We compared two inoculation densities, both higher than historical inoculation rates, in an effort to decrease photo-inhibition and photo-damage. We hypothesized that this would improve outdoor culture survival rates and increase daily productivity (biomass gains) and final 14-day batch biomass and N concentrations. In addition, we performed a phytohormone assay to determine if certain phytohormones and phytohormone concentrations are consistently produced during a 14-day batch period. Cyanobacterial-based phytohormone extracts have been shown to improve growth, yields and drought tolerance in a variety of crops.
Increased inoculation rates combined with decreased culture depth and CO2 supplementation achieved 14d batch biomass and N concentrations of 542 mg L-1 and 60 mg L-1, respectively. This is a 100% increase in 14-day batch biomass and N concentrations.

Six phytohormone related compounds were detected – abscisic acid (ABA), salicylic acid (SA), cytokinin (CK), trans zeatin riboside, and three auxin compounds: indole 3-acetic acid (IAA), indole 3-acetamide, an IAA precursor, and Indole 3-carboxylic acid. Concentrations of auxins, CK, and ABA were consistent across experiments, while SA showed more variability.

A cyanotoxin assay conducted using gas chromatography-mass spectrometry did not find detectable concentrations of microcystins, anatoxin-a, or cylindrospermopsin.

Task 1.2 – Evaluate batch vs. semi-continuous growth of cyanobacteria.
Two operational modes were tested for biomass and N yield in a four-week study: semi-continuous and batch. After six days of growth, 25% of the semi-continuous treatment was harvested and filled to volume with fresh medium every other day. After 14 days of growth, 85% of the batch treatment was harvested, and the remaining 15% was used as seed to begin a second 2-wk batch. Biomass yield was estimated by optical density (OD) and chlorophyll content, and N-fixation was estimated by Total Kjeldahl Nitrogen (TKN). At the end of four weeks, biomass yield and total N fixed was calculated for the batch and semi-continuous treatments.
There was no treatment difference in the total (four-week) biomass (µmol chlorophyll d-1 and OD d-1) and N yield (TKN d-1). However, during the first 2-wks, the batch system was significantly higher in biomass and N yield than the semi-continuous system. On the other hand, during weeks 3 and 4, the semi-continuous production system was significantly higher in biomass, specifically chlorophyll (µmol chlorophyll d-1) and N yield compared to the batch system. The biomass and N yield within the batch treatment declined from set 1 (t1=0 to t2=14) to set 2 (t1=15 to t2=30). The semi-continuous treatment increased in biomass and N yield from set 1 to set 2.

One explanation for similar total (four-week) biomass and N yield between treatments is that the cell density of semi-continuous was maintained outside of the optimal cell density range during the harvest regime. The optimum cell density is the range in cell density that yields the highest productivity due to the most efficient use of resources such as light and nutrients. In semi-continuous production systems, the harvest regime should be based on maintaining the identified upper and lower biomass concentrations that correspond to the optimum cell density. Harvest should occur every time the culture hits the maximum optimum cell density and should ideally be harvested down to the minimum optimum cell density. A harvest regime based on the minimal and maximum optimum cell density maintains the cells in the exponential growth phase. Knowledge of the upper and lower optimum cell concentration will aid in the maintenance of an approximate steady state for an outdoor raceway, allowing for cells to remain in the exponential growth phase and maximizing growth and N-fixation.

Task 1.3 – Explore methods to optimize light absorption.
We compared batch culture concentrations under full outdoor light exposure to that under high tunnels. High tunnels may provide protection when photosynthetically active radiation is highest and cultures are dilute, but may also limit light availability later in the 14-day batch period when cultures become increasingly dense. No significant differences were seen in 14-day biomass or N concentrations between high tunnel and outdoor production. There was also no significant difference in 14-day biomass and N concentrations between two inoculation rates.

However, results from the 2016 field experiments showed that we can increase 14-day batch biomass and N concentrations by increasing inoculation densities and decreasing culture depth in combination with CO2 bubbling (reported under Task 1.1 above). Increasing inoculation densities can protect cyanobacteria from photo-damage, and decreasing culture depth reduces self-shading and increases light availability.
No net gains in biomass were recorded on cloudy days during the second week of production, suggesting that light can be limited in on-farm production. Production parameters designed to allow for increased light availability (e.g. shorter batch periods, shallow raceways, increased mixing, etc.) could increase growing season yields. In fact, as reported above, when inoculation rates were increased in conjunction with decreased culture depth and CO2 supplementation, the 14-day batch biomass and N concentrations increased by 100%.

Objective 2) Evaluate the utilization of cyano-fertilizer in irrigated fruit (peaches) and vegetable (lettuce, sweet corn) systems

Task 2.1—Compare cyano-fertilizer with commonly-used organic fertilizers (compost, fish emulsion, feather meal, and blood meal) in terms of impact on plant growth, yield, quality, and N recovery.
We assessed the effects of organic N fertilizer application and N rates on nutritional value, water use efficiency, and N dynamics of sweet corn and lettuce in a two-year field study conducted at the Colorado State University Horticulture Research Center, Fort Collins, CO. The fertilizers used in this study were blood meal, feather meal, fish emulsion, and cyano-fertilizer. Both fish emulsion and cyano-fertilizer were supplied in four split applications over the growing season through drip irrigation, while the blood meal and feather meal were sub-surface banded prior to planting. Lettuce and sweet corn were used as indicators to evaluate the effects of organic N fertilizers on nutritional value, water use efficiency, and N dynamics. The aims of this study were to evaluate the effect of different types of organic N fertilizer on nutritional value (β-carotene, Fe, and Zn), marketable yield, water use efficiency (WUE), residual soil NO3--N, N content, and N use efficiency (NUE) of horticultural crops, particularly lettuce and sweet corn.

All fertilizer treatments in year 1 increased β-carotene concentration in leaf tissue compared to control, while only fish emulsion had a higher β-carotene concentration compared to other treatments in year 2. The high indole-3-acetic acid (IAA) applied in the fish emulsion treatment could have increased β-carotene concentration in lettuce in both years. Amount of IAA applied in the fish emulsion treatment was positively correlated with β-carotene concentration in both years. However, a significant negative correlation was found between marketable yield and β-carotene concentration in leaf tissue.

In lettuce, the blood meal treatment had lower leaf Fe and Zn concentrations than other fertilizer treatments applied at 112 kg N ha-1. The cyano-fertilizer treatment had a higher leaf Fe concentration than other fertilizer treatments applied at 56 kg N ha-1. Leaf N concentration was positively correlated with Leaf Fe and Zn concentrations. Amount of NO3--N applied in organic N fertilizers was negatively correlated with leaf Fe concentration. The cyano-fertilizer, fish emulsion, and blood meal treatments increased Fe concentration in sweet corn compared to feather meal. Amounts of NO3--N, Fe, and Zn applied in organic N fertilizers were positively correlated with kernel Fe concentration, while amount of NH4+-N applied was negatively correlated with kernel Fe concentration. There was no N rate or treatment effect on leaf and kernel N concentrations in sweet corn.

The phytohormones applied in organic N fertilizers may have affected field water use efficiency (fWUE), instantaneous water use efficiency (iWUE), kernel number, and leaf gas exchange components of sweet corn. Cyano-fertilizer apparently had a higher WUE, likely due to the high amount of salicylic acid (SA) in the cyano-fertilizer. A positive relationship was observed between the amount of SA applied with iWUE and fWUE. The amount of NH4+-N and Ca applied in the feather meal treatments were negatively correlated with both iWUE and fWUE.

An N rate effect was observed in marketable yield and NUE of lettuce in both years. Blood meal and feather meal fertilizers, with higher percentage of N applied as NO3--N compared to other fertilizer treatments, had a higher residual soil NO3--N concentration. Greater residual soil NO3--N was observed in the 0-30 cm depth compared to the 30-60 cm depth. Organic growers could achieve higher marketable yield and NUE when applying fertilizers at rates between 28 kg N ha-1 and 56 kg N ha-1 compared with 112 kg N ha-1. In sweet corn, the feather meal and fish emulsion treatments had a higher residual soil NO3--N compared with other treatments. The fish emulsion, cyano-fertilizer, and blood meal had higher leaf and kernel N contents and NUE compared with feather meal at 56 kg N ha-1. The cyano-fertilizer treatment had a higher marketable ear yield and NUE compared with other treatments at 112 kg N ha-1.

Overall, in both lettuce and sweet corn, cyano-fertilizer usually yielded at least as high as, or higher than, other organic fertilizers applied at the same N rate.

Task 2.2—Assess the effectiveness of cyano-fertilizer in contrast to farmers’ current practice on two working peach orchards.
Mature peach orchards were selected for use at Osito Orchard (Farm A) and Ela Family Farms (Farm B). Ela Family Farms has been certified organic since 2003, and Osito Orchard was transitional and became certified in 2016. Both are certified by the Colorado Department of Agriculture. On both farms, the peach cultivar used was Sun Crest grown on a Lovell rootstock. The trees at Farms A and B were planted in 2008 and 1999, respectively.
Experimental plots consisted of five adjacent trees in the same row, with the entire plot receiving the treatment, but measurements were only taken from the three central trees. At farm A, two treatments were applied. The first treatment, “High Manure,” was 100 lb N/acre as the dried chicken manure fertilizer True Organic 12-3-0 (Spreckels, CA) which was the grower’s typical N fertilizer, made from a combination of feather meal, meat, and bone meal. A second treatment, “High Manure+Cyano,” was 100 lb N/acre True Organic 12-3-0 with 10 lb N/acre from cyano-fertilizer. During summer of 2014, the trees at Farm A were heavily infested with green peach aphid (Myzus persicae), which damaged or killed many first year branches and severely lowered fruit counts and yields in some trees. Then, on April 3, 2015, a freezing event killed most of the blossoms on the trees, and it was decided to discontinue the experiment at Farm A.

At farm B, three treatments were applied: 1) “High Manure” was 100 lb N/acre from Richlawn 5-3-2 (Platteville, CO) a dried poultry manure, 2) “High Manure+Cyano” was 100 lb N/acre from Richlawn 5-3-2 with 4.5 lb N/acre from cyano-fertilizer, and 3) “Low Manure+Cyano” was 75 lb N/acre from Richlawn 5-3-2 with 4.5 lb N/acre from cyano-fertilizer. At farm B in 2015, a second group of treatments was added, comprised of the same treatment applications, but on new plots, within the same orchard. In 2015 the plots from the Farm B 2014 treatments continued to be evaluated for residual effects, but they were only given their respective amounts of Richlawn 5-3-2, excluding the cyano-fertilizer. On April 3, 2015 a freezing event killed most of the blossoms on the trees in the orchard, resulting in roughly 65% yield and fruit count reductions, since temperatures dropped to approximately -3.8 degrees C for a period of at least 3 hours.

Tree trunk cross sectional area (TCSA) was measured in the spring prior to fertilizer applications, and in the autumn after the growing season had ended. Circumferences were measured at a height of 21.3 cm from the orchard floor. Circumferences were then converted into TCSA by the formula: TCSA= (Trunk circumference/2π)2 x π. TCSA change was calculated by the formula: TCSA change= (End of season TCSA- Beginning of season TCSA).

Chlorosis was monitored because moderate to severe chlorosis was evident in 2015. To quantify chlorosis, leaf chlorophyll was estimated using a SPAD 502-PLUS meter (Konica Minolta, Osaka Japan) on first year, fully expanded leaves from the middle to the growing tip on all trees at Farm B on August 29, 2015. Each tree was divided into 4 quadrants based on compass directions. The mean of 5 measurements was recorded for each quadrant of each tree.

On Farm B in 2014, peaches were harvested by picking crews, in two rounds on August 20, 2014 and August 27, 2014, with all fruit being harvested by the end of the second picking. In 2015 on Farm B, all peaches were harvested August 17, 2015. Peaches were harvested into boxes, and total fruit count and fruit weight were measured for each plot.

The results at Farm B in 2014 showed increases in yield and decreases in TCSA growth in Cyano-Manure treatments compared to the No-Cyano manure treatments. It is likely that the lower TCSA growth followed the higher fruit count, because of the nutrient allocation being directed to the fruit (and away from the trunk) to a greater degree. The No-Cyano treatment had lower yields, but greater TCSA growth, indicating that lower fruit set may contribute to greater vegetative growth. In 2015, there were no significant differences found in fruit yield, average fruit count, or TCSA among treatments, possibly due to the impact of the freezing event during bloom.

However, in 2015, SPAD chlorophyll readings were significantly higher (P=0.0350) in Cyano-Manure treatments compared to the No-Cyano treatment. SPAD was significantly and positively correlated to distal leaf Fe concentration (R=0.571, P=0.0136). The effect of leaf Fe concentrations on chlorophyll content is well established. Siderophores are sometimes released by cyanobacteria in response to Fe limiting environments, and may have been applied with the cyano-fertilizer, causing increased Fe uptake and reduced chlorosis.
In summary, in the on-farm research carried out on organic peach orchards, application of cyano-fertilizer in addition to compost increased peach yield and reduced the growth of the tree trunks. In addition, cyano-fertilizer increased the SPAD chlorophyll readings of the leaves, and the SPAD readings were positively correlated to the distal leaf iron concentration.

Task 2.3--Evaluate the barriers to integration of cyano-fertilizer into current organic farming systems.
We have identified three primary barriers to integration of cyano-fertilizer production on-farm into current organic farming systems: land requirement, management needs, and investment cost.

Using current improved methods of production, in order to provide 50 lbs N/acre for 1 acre of vegetables, cyano-fertilizer production ponds would require 0.07 acres (3136 sq ft) of land. The ponds could be placed on low-productivity land; however, this is still a substantial investment of land for fertilizer production that cannot be used to grow something else (thus, the opportunity cost is high).

In addition, the management of cyano-fertilizer production is not straightforward and will require training, monitoring, and making adjustments throughout the process. This investment of time and expertise may also limit the dissemination of cyano-fertilizer into organic systems. The cyanobacteria are living organisms that respond to their environment, and thus, they are not always predictable or reliable and have to be managed precisely.

The cost to build the production raceway ponds is by far the largest production cost for cyano-fertilizer. This is an up-front, start-up cost that is likely to require several years of fertilizer production to pay for itself. At this time, cyano-fertilizer is cheaper than the most expensive organic fertilizers, such as alfalfa meal, but is not yet competitive with fish emulsion or other liquid fertilizer products.

Objective 3) Quantify the direct costs and benefits of on-farm production and utilization of cyano-fertilizer to optimize economic returns for farmers

Task 3.1—Quantify the costs of cyano-fertilizer production as compared to commonly-used organic fertilizers.
The cost of materials dominated the total cost of raceway construction; materials were ~83% of the total cost ($1163 for a 6 ft x 36 ft raceway), while less than 20% of the cost was for labor. The 1st-year fertilizer cost savings in our experiments was not adequate to cover the raceway construction costs. Therefore, we are now aiming to reduce construction costs and increase raceway productivity in order to increase the value proposition for on-farm production of cyano-fertilizer. Based on work by other experts in micro-algae production, as we scale up we should be able to achieve raceway construction costs of $5-10/m2, costs much lower than our current raceways ($26/m2).

Operating costs for a 7-month production period (fifteen 2-wk cycles) are estimated to be $353 ($196 for electricity and $157 for nutrient solution). This does not include the opportunity cost of about 3 hours/week (90 hrs over the entire production period) of farmer management of the cyano-fertilizer production process. Our calculations also do not include the labor and fuel savings achieved by applying cyano-fertilizer (or fish emulsion) through an irrigation system, as opposed to soil application and incorporation.

We developed scenarios to evaluate the impact of variable labor costs on total raceway construction costs. Although the amount of labor needed for raceway construction is fixed, the rate of pay is variable. Changing labor costs from $10/hr up to $30/hr, although a 3-fold increase, only results in + 8.7% in construction cost. Therefore, the minimization of material costs (fixed costs) is most important in achieving shorter breakeven periods.

Task 3.2—Appraise the economic benefits of cyano-fertilizer use.
Alternative fertilizer costs are important to the profitability of on-farm cyano-fertilizer production. The higher the costs of the alternative fertilizers, the easier it is to achieve profitability of the cyano-fertilizer. Variability in the price of different organic fertilizers is high; organic fertilizers vary almost 5-fold from the cheapest to the most expensive. Therefore, although profitability can already be shown in comparison to the most expensive organic fertilizers (e.g., alfalfa meal), we are focusing on reducing construction costs and increasing cyanobacterial growth and N fixation in order to be able to compete with the cheaper organic fertilizers. The shipping costs range from 33.3-60.0% of the total costs of organic fertilizers, respectively. In general, the larger the amount purchased, the lower the shipping cost per pound of N.

In addition to direct price comparisons per lb of N, on-site fertilizer production also provides intangible value and addresses farmer needs and preferences. In our market survey, farmers professed needs to reduce the bulkiness and handling costs of low nutrient materials, minimize odor, and allay sustainability concerns. Although, at this time, we cannot assign a hard dollar value to these advantages of cyano-fertilizer, they do play a role in farmer decision-making.

The current cost-benefit ratios for organic fertilizers range from 0.6 – 2.7. Therefore, the current production system already presents a cost savings when compared to alfalfa meal fertilizer. However, to be competitive with a wide range of organic fertilizer products, increasing productivity by 2-3 times, or reducing costs by ½ to 1/3, or a combination of the two will be required.

Based on our on-farm experience applying cyano-fertilizer directly through drip and micro-sprinkler irrigation lines without any clogging, it is clear that applying the cyano-fertilizer through fertigation is the most profitable scenario on farms with irrigation systems (nearly all vegetable farms in the western USA are dependent on irrigation).

Task 3.3 --Evaluate the economic feasibility of on-farm production and use of cyano-fertilizer.
The economic evaluation of the on-farm cyano-fertilizer production system found that cyano-fertilizer is already competitive with the most expensive organic N fertilizers. However, to be competitive with fish emulsions and blood and feather meals, the cost per lb of N must be reduced to about half of current costs. Based on the experiments described above under objective 1, we have successfully doubled N fixation rates while increasing costs by only about 20%. However, we had to reduce the depth of the ponds to 55% of the original depth to achieve these higher N fixation rates, and this change in depth effectively increases the land requirement by 1.8 times. This is a major detriment to economic feasibility of the current system.

Objective 4) Determine the carbon footprint of cyano-fertilizer compared to other methods of fertilization

Task 4.1—Monitor the N2O and CO2 emissions from cyano-fertilizer applied to land as compared to commonly-used organic fertilizers (fish emulsion, feather meal, and blood meal).
We evaluated the effects of four organic fertilizers (feather meal, blood meal, fish emulsion, and cyano-fertilizer) applied at different rates (0, 28, 56 and 112 kg N ha-1) on N2O and CO2 emissions from a lettuce field (Lactuca sativa). The study was conducted for two years and compared pre-plant applied solid fertilizers (feather meal and blood meal) and multiple applications of liquid fertilizers (fish emulsion and cyano-fertilizer). Three days a week, N2O and CO2 emissions were measured using a closed-static chamber, and gas samples were analyzed by gas chromatography. Pre-plant applied solid fertilizers significantly increased cumulative N2O emissions as compared to control, but multiple applications of liquid fertilizers did not. Emission factors for N2O ranged from 0-0.1% for multiple applications of liquid fertilizers and 0.6-11% for pre-plant applied solid fertilizers, which could be overestimated due to chamber placement over fertilizer bands. In general, solid fertilizers with higher C/N ratios (3.3-3.5) resulted in higher CO2 emissions than liquid fertilizers with lower C/N ratios (C/N of 0.9-1.5). Therefore, organic farmers should consider the use of multiple applications of liquid fertilizers (such as cyano-fertilizer and fish emulsion) as a means to reduce soil greenhouse gas emissions while maintaining high yields.

Task 4.2—Modify the Daycent model to include cyano-fertilizer and other organic fertilizers for use in quantifying and comparing the carbon footprint of those fertilizers.
Carbon footprint is commonly defined as the total amount of greenhouse gases produced to directly and indirectly support human activities, usually expressed in equivalent tons of carbon dioxide (CO2). Nitrous oxide (N2O) is a greenhouse gas with a global warming potential of 265–298 times that of CO2. Organic fertilizer application can affect N2O and CO2 emissions from soil by influencing nitrification, denitrification and microbial decomposition. DAYCENT, a widely used biogeochemical model, has been extensively tested for major commodity crops but not for specialty crops. The objectives of this study were to compare simulated and measured N2O and CO2 emissions from irrigated lettuce plots from different organic fertilizer treatments. The N2O and CO2 emissions were measured from a lettuce field (Lactuca sativa L.) with nine treatments: four organic fertilizers applied at different nitrogen (N) rates (28 and 56 kg N ha-1) and an unfertilized control. The measured data from the low N rate treatments were used for calibration, and the data from the high N rate treatments were used for validation. Comparison of daily N2O and CO2 emissions simulated by DAYCENT and measured from the field yielded coefficients of determination (r2) of 0.0004 to 0.48 and 0.002 to 0.65, respectively. DAYCENT simulated the effect of blood meal and feather meal (single application) on both N2O and CO2 emissions better than for fish emulsion and cyano-fertilizer (multiple applications). Cumulative emissions from DAYCENT were overestimated except for cumulative N2O emissions from feather meal and blood meal treatments. The effect of single organic fertilizer applications on N2O and CO2 emissions were simulated well by DAYCENT, and daily N and C mineralization rates could further improve the performance of the single application simulations.