Research and educational support for organic dairy farming in the South

Research sites:

Center for Environmental Farming Systems Pasture-Based Dairy Unit (CEFS), Goldsboro, NC: CEFS is a seasonal-calving (autumn-calving), pasture-based dairy that manages half of its cattle to an organic health standard and the other half conventionally. All cows in both management groups receive conventional concentrate feed as well as pasture in amounts sufficient to meet the U.S. organic pasture requirement. However, cows in the organic group graze in pastures that can be certified organic and receive organic treatments and disease prevention strategies. The herd had approximately 130 milking cows each year. The herd is composed of Holsteins, Jerseys, and crosses between those breeds.

Commercial organic and conventional dairies in North Carolina: In 2010, all 6 certified organic dairies, 6 conventional dairies of similar size and geographic location, and the CEFS herd were surveyed to determine herd health practices related to milk quality as well as to compare milk quality. The farms had between 18 and 157 milking cows.

Arkansas lost all of their organic dairy producers when the processors and cooperatives dropped them, and so data from organic dairy farms in Arkansas were not available as planned for this research project.  However projects were conducted related to fly repellents on commercial farms and at two research sites in Arkansas.
Research from North Carolina State University:

Study A: Research on alfalfa-grass mixtures under organic and conventional management

Project Design: Two varieties of alfalfa and two varieties of prairie grass were tested at CEFS in replicated grazing plots (Field 1: 0.8 ha or 2 acres each; Field 2: 0.5 ha or 1.25 acres each) and in smaller 12m x 18m (40’ x 59’) hay plots in a split-block design over 2 years. These plots had a sandy loam soil type, and  had historically been planted with sorghum-Sudangrass hybrid for summer grazing and annual ryegrass in winter (Field 1) or was in hybrid Bermudagrass sod for several years (Field 2).

Fields were prepared for planting following application of lime and fertilizer according to soil test recommendations and following organic regulations for the organic plots, except that certified organic seed was not used. Conventional plots without prairie grass had the herbicide Eptam® (EPTC, s-ethyl dipropylthiocarbamat) incorporated at 4.1 L/ha (7 oz/acre) before planting alfalfa. Field 1 was planted in October 2009, Field 2 was planted in October of 2010 after a smother crop of sorghum-Sudangrass (planted May 2010) was disked in, and the hay plots were planted in October 2009 following two diskings in August to disrupt establishment of cool-season grasses and another disk in early September after lime application.

The two varieties of alfalfa used in this study were Arriba alfalfa with a fungicide seed coat (America’s Alfalfa®) planted at 22.5 kg/ha (20.1 lb/acre) and Haygrazer (Cimarron USA®) inoculated with Rhizobium meliloti (Nitragrin AB®) planted at 29.0 kg/ha (25.9 lb/acre). The two varieties of prairie grass that were used in the hay plots were Matua prairie grass (Barenbrug®) and Lakota prairie grass, planted either in fall (broadcast at 16.8 kg/ha (15 lb/acre)) or spring (broadcast at 16.8 kg/ha (15 lb/acre)). Conventional plots of alfalfa without prairie grass were sprayed on December 9, 2009 with sethoxydim (Poast Plus®) at 1.75 L/ha (3 oz/acre) for ryegrass control.

In the hay small-plot trial, both varieties of alfalfa were evaluated with both varieties of prairie grass in a replicated split-block design with embedded subplots of prairie grass. In the grazing trial, only Matua prairie grass was evaluated with both varieties of alfalfa to simplify the design of the trial. In the grazing trial, the design was a replicated split-block design (management x alfalfa variety), with Matua drilled into organic and conventional plots in February of 2010 (16.8 kg/ha (15 lb/acre) in Field 1), late August of 2010 (22.5 kg/ha (20.1 lb/acre) in Field 1) and February of 2011 (22.5 kg/ha (20.1 lb/acre) in Field 2). Establishment of prairie grass was more successful during the fall planting.

Harvesting, either for the hay trial or grazing for the grazing trial, occurred in the organic plots when the alfalfa weevil presence reached the economic threshold of 3 alfalfa weevil larvae per stem. If this threshold was reached in the conventional plots, those plots were sprayed with Lorsban-4E® (chlorpyrifos). Alfalfa scouting occurred from February through April each year. Harvest occurred in both organic and conventional plots when alfalfa maturity reached 1/10th bloom (late bud to early bloom). Hay plots were harvested to a 10cm (4”) stubble height and grazing plots were harvested to a stubble height of 10-15cm (4-6”). Each harvest was examined for forage composition including volunteer grasses and weeds.

Hay trial harvest dates: For organic plots, 2010 harvests occurred in early April, mid-June, late July, and late August. For conventional plots in 2010, harvest occurred in mid-May, mid-June, late July, and late August. Organic plots in 2011 were harvested in late April and mid-May; conventional plots in that same year were harvested only in mid-May. Grazing trial harvest dates: In year 2010, organic pastures were grazed in early April, early June, late July, and mid-August 2010. Conventional pastures were grazed in late April, early June, late July, and mid-August 2010. In year 2011, the organic and conventional grazing plots were grazed a single time in early May 2011.

Statistical analyses were performed using SAS 9.1 (SAS Institute, Cary, NC). For the hay trial, the MIXED procedure was used with management (organic or conventional) and prairie grass treatment (alfalfa interseeded with prairie grass versus alfalfa not interseeded with prairie grass) as fixed effects, replication as a random effect, and using repeated measures to estimate cumulative production. Degrees of freedom for contrasts were corrected using the Kenward and Rogers method. Statistical analyses for the grazing trial were performed in the same way as for the grazing trial, except the prairie grass treatment was not included in the model.

Study B: Comparisons of milk quality on North Carolina organic and conventional dairies

The goal of this project was to compare milk quality, measured by somatic cell count (SCC) and milk microbiology, between organic and conventional dairies in North Carolina during the warm months of the year (May-October). This study was performed in 2010 on all 6 certified organic dairies, 6 conventional dairies of similar size and geographic location, and the CEFS herd. Milk samples were collected during a single milking for each dairy, and samples were collected from all functional quarters of every cow contributing milk to the bulk tank. One set of milk samples was taken from each quarter for microbiological analysis and another set was pooled from all quarters for SCC analysis. Somatic cell count is a commonly used metric for milk quality, as it is a good proxy for mastitis infection and directly affects the shelf life of dairy products. Milk microbiology is a direct measure of the prevalence of mastitis in the herd, and provides information on the type of mastitis (contagious or environmental) present.

Farmers were also surveyed about their milking protocols, dry cow treatment, and treatment of mastitis.

Statistical analysis was performed using SAS (SAS Institute, Cary, NC), with consideration of management type as a fixed effect and farm within management type as a random effect. Logistic regression was used to determine differences between management groups for the cow-level prevalence of different mastitis-causing organisms.

Study C: Evaluating herbal alternatives to antibiotics for use as dry cow therapy

The objective of this study was to evaluate two commercially available (through a veterinarian) herbal products for use as dry cow therapy.  Most dry cow therapy products are antibiotics for intramammary infusion, and one product (Orbeseal®) is an internal teat sealant with no antibiotic activity on its own. Because organic dairies in the United States are prohibited from using antibiotics, this study investigates two potential alternatives to antibiotics for dry cow therapy.

The first trial evaluated Phyto-Mast® and Cinnatube™ as dry cow therapy, with 5 different treatments: Phyto-Mast alone (n = 255), Cinnatube alone (n = 230), Phyto-Mast and Cinnatube (n = 214), no treatment (n = 241), and conventional antibiotic + internal teat sealant treatment (Quartermaster® and Orbeseal®; n = 104). This trial was evaluated at CEFS as well as on two commercial organic farms and two commercial conventional farms in North Carolina. The conventional treatment was only used in the CEFS herd on the conventional cattle, as the commercial dairy producers were not interested in testing it on their farms. Cows were sampled between August of 2010 and March of 2012. Another aspect of this trial was examining the effect of the treatments on milk production; this portion was studied in the CEFS herd only, as full lactation records were available for that herd from monthly testing.

The second trial evaluated Phyto-Mast alone (n = 72), Orbeseal alone (n = 71), Phyto-Mast and Orbeseal (n = 68), and no treatment (n = 68). This trial was conducted at CEFS from August 2012 - March 2013.

For both trials, milk samples were taken directly before dry off and about a week after freshening from every udder quarter of every cow in the study - 1 set of samples for microbiological analysis, and 1 set for SCC analysis from each quarter. Milk production (Trial 1), milk microbiology, and SCC were compared. Generalized linear mixed models were used to perform statistical analysis, modeled in SAS version 9.2 (SAS Institute, Cary, NC). Farm and cow were included in the models as random effects and all other measured variables were considered for inclusion in the models. Somatic cell counts were converted to somatic cell scores (also called “linear scores”) using the formula ln(SCC/100,000) + 3 for the analysis. All results are presented as least squares means ± standard errors.

Study D: An in vitro  assessment of the antibacterial effects of plant-derived oils

In this study, the antibacterial activity of the plant oil ingredients of Phyto-Mast was assessed against three mastitis-causing pathogens in vitro. The plant oils include essential oil of thyme (Thymus vulgaris) and canola oil-extracted Angelica dahuricae, Angelica sinensis, Gaultheria procumbens, and Glycyrrhiza uralensis. The mastitis pathogens were Staphylococcus aureus, Streptococcus uberis, and Staphylococcus chromogenes, isolated from mastitis cases in North Carolina. The Clinical and Laboratory Standards Institute standard for determination of antibacterial activity in broth was used, but ultrapasteurized homogenized whole milk was used instead of broth to more closely approximate the environment of the udder.

Essential oil of thyme was tested by itself  at 1%, 2%, and 3% concentrations (vol/vol) because it has known antibacterial activity. The other plant-derived oils were tested alone and in every possible combination for a total of 15 different treatments, each tested at 0.5%, 1%, 2%, and 4% concentrations to simulate concentrations achievable in the cow’s udder provided that her milk production was about 20 lbs/day at the time of dry off. The combination treatments had equal volumes of each plant oil to a total volume of 5 mL - for example, a treatment with three different oils would have 1.67 mL of each oil combined to make up that treatment prior to testing on the bacteria.

Controls included milk only to ensure that pasteurization was successful, milk + bacteria as the reference control, and milk + bacteria + penicillin-streptomycin at 1% and 5% concentrations to ensure that the bacteria used were sensitive to antibiotics. All treatments and controls were run three times, randomized by date, bacteria, and technician to reduce variability. Canola oil (1% and 70% concentration vol/vol) and Phyto-Mast (1, 2, 3, 4% concentration vol/vol) were also tested to determine if canola oil has any antibacterial effects on its own and to see if Phyto-Mast has antibacterial activity  in vitro.

Bacteria were grown on agar plates at the beginning of the experiment, then grown just prior to the beginning of each replication in Mueller-Hinton broth. Vials were prepared for the study by adding a calculated volume of milk and the treatment to be tested to a total volume of 1 mL. Vials were vortexed to mix the contents, a constant volume of bacteria were added, and the vials were vortexed again to mix the bacteria, treatments, and milk before placing in the incubator. Vials were incubated 24 hours. Vials were vortexed again to evenly distribute bacteria, then serial ten-fold dilutions were used to determine bacterial counts. The dilutions were plated on agar plates and incubated for 24 hours, then the colony forming units were counted and recorded. Results are reported as the growth of the bacteria in the treatment sample relative to the milk + bacteria control.

Study E: A comparison of production, reproduction and animal health between two groups of dairy cattle managed organically or conventionally in a pasture-based system

This study examines data on 503 cow records from 4 calving seasons at CEFS. The herd had approximately 130 cows each season, with pure Holsteins, pure Jerseys, and Holstein x Jersey crossbred cows. The crossbreeding system involves inseminating purebred cows using semen from the same breed one year and from the other breed the next year. This same breeding plan is used on the F1 females (HJ and JH), so that cows will have calves that are 75% Holstein (HX) or 75% Jersey (JX) the next year. Cows that were >50% one breed were inseminated with semen from the other breed so that 75% of a single breed was the highest percentage of any one breed in a crossbred cow. Production data (milk. protein, fat), as well as dairy manager-reported reproduction and herd health events were examined to compare the organic and conventional groups as well as the different breed groups. All data were taken from PCDART, the herd management software used at the farm. Most health events were recorded as presence or absence of the event during each lactation, but calving difficulty was recorded as the score from 1 to 5, with 1 representing a calving that required no assistance, 2 required some assistance, 3 required moderate assistance, 4 required considerable force, and 5 was extreme difficulty in calving. Calving, insemination, and herd check dates were used to determine fertility measures including pregnant to first insemination, total pregnancy rate during the breeding season (January-March, or 90 days post-calving), the interval between calving and first insemination, the gestation length, and the calving interval.

Statistical analysis was performed using the GLIMMIX procedure in SAS (SAS Institute, Cary, NC) to compare production, reproduction, and health events between organic and conventional management and among breed groups. Calving season was included in all models as a random effect. Results are presented as least squares means ± standard errors unless otherwise noted.

Research from the University of Arkansas:

From 2010 to 2013, fly repellents that could potentially be used on organic dairy farms to repel horn flies were evaluated.

2010: A horn fly repellent evaluation was begun in May of 2010 in Washington County, Arkansas, on three farms: two conventional dairy farms (treatments; Farm A&B) and one pasture-based beef farm (Control).  Monitoring of horn fly populations started in May (Farm A = 41 avg. horn flies/cow; Farm B = 23; Control = 24).  Horn fly populations were evaluated by visually estimating the number of horn flies on ten randomly selected animals on each farm.  High resolution pictures were also taken of those same cows and number of horn flies per cow determined by manually counting the flies in each picture. The threshold was reached on June 1st at one of the dairies (Farm A = 105 avg. horn flies/cow), so the repellent dust was applied the following week using one dust bag (dispensed .07oz of dust per animal per day) placed in the alleyway of the milk barn exit in order to insure that the cattle would be treated each time they exited the milk barn (twice daily).  The fly numbers were not declining after one month, so a second dust bag was hung adjacent to the first to double the application rate (Figure 1). After 10 days using two dust bags, the rain guard on the bottom of each dust bag was removed to further increase the application rate.  The threshold for Farm B was not reached until July 13th (137 avg. horn flies/cow), so the dust bags were hung at Farm B on July 20th.

Figure 1. Dust bags hanging in exit alley at a dairy barn.

Monitoring of horn fly populations continued at each farm, even after it became apparent that the repellent, C8910, was not effectively reducing horn fly numbers when compared to the control farm (Aug 3rd at Farm A = 1181, Aug 10th at Farm B = 195).  The horn fly population at Farm B was much smaller than that at Farm A; however, the farm owner asked us to remove the bags.  The repellent dust at Farm A was replaced with an insecticidal dust which contained zeta-cypermethrin and piperonyl butoxide (Python®); the repellent dust at Farm B was removed and insecticidal ear tags (beta-cyfluthrin - Cylence Ear tags) used for the remainder of the study.

Because the repellent dust, C8910, was not effective, it was proposed by the manufacturer that a formulation with lower volatility compound or a product containing a different percent of active ingredient might be more effective; therefore we initiated a preliminary evaluation.  The control farm for this trial was selected for evaluating these alternative formulations because insecticides had not been used on that farm and the horn fly population was still very high.  Three groups were randomly selected to receive one of three treatments, 15% active low volatile compound (15% lv), standard compound 1% active, standard compound 5% active. Dust was applied to each cow individually using a head gate to restrain the animal and a hand-operated dust blower to apply the treatments (~1oz per animal). Horn fly numbers were visually estimated and then pictures taken as above to be manually counted in the lab for each cow at three days after treatment and one week after treatment.

2011: An experiment to evaluate geraniol was conducted in Savoy, AR at the University of Arkansas Savoy Research Station in 2011. The product tested was a 1% water-based geraniol solution (after mixing 1:5 Concentrate:Water) manufactured by FASST Products, LLC., Brooklyn, NY. In spring of 2011, heifers were used as the study animals.  The herd was divided into three groups: untreated control (n=18), one group treated once per week (n=19), another group treated twice per week (n=19).  In the fall, the study was repeated using ~600lb steers which were separated into two replicates. The experimental design was the same as the spring trial except the fall trial had a second concurrent replicate.  The fall trial consisted of 8 steers per group.  Replicates were physically separated, but maintained on the same farm.  Treatments were applied using the 3-D Quik Hand™ Cattle Sprayer (3-D Cattle Equipment, LLC. Pine Ridge, AR).  All treatment groups received the same application rate (approximately 130mL/animal) of a 1% geraniol solution during each treatment day, however, one treatment group in each replication was treated once per week (each Tuesday for the duration of the study); the other was treated twice per week (each Tuesday and Friday for the duration of the study).  Weekly visual estimates (10 heifers sampled per group in spring; each steer sampled per group in fall) were made by estimating the number of horn flies on one side.  These estimates were used to calculate the average number of horn flies per animal per group.

The number of horn flies per steer was decreasing initially after being treated with a water-based geraniol mixture, but would begin increasing within one day.  Therefore, it was important to determine the residual repellency duration.  The residual evaluation for the water-based geraniol was conducted on days that both treatment groups would receive the geraniol spray for the weekly fly study.  The initial trial indicated that the geraniol was no longer active at four hours post-treatment, therefore, subsequent trials focused on the zero to four hours after treatment timeframe. Visual fly population estimates were determined prior to treatment and then at one, two, and four hours post-treatment.  Data was collected on three separate days and all treated animals received the same application rate on each day; application rates on each day that residual activity was monitored were 142.2 mL/animal, 137.8 mL/animal, and 123.1 mL/animal.  Animals were divided into two replications each consisting of a control and two treatment groups; each group was made up of 8 steers.

In 2012, a product containing 0.8% geraniol (No-Fly) was evaluated. Because geraniol in the 2011 evaluation had a short residual time, we began exploring options for increasing the amount of the time geraniol would remain active.  One option which was encountered is the addition of vanillin (Khan et al. 1975).  Vanillin was shown to at least double the duration of residual activity in mosquito repellents and in the case of DEET, increased residual activity much greater with higher concentrations of vanillin.  Vanillin alone was minimally repellent or non-repellent.

A commercially available product sold under the trade name, No-Fly, and manufactured by Crystal Creek, Inc. contains geraniol and is marketed as an organic alternative to conventional fly control.  No-Fly contains a similar percentage of geraniol (0.08%) as used in the 2011 study, but also contains vanillin as well as other essential oils and plant extracts.  No-Fly repellent is available as either a water-based or oil-based solution.  The oil-based solution contains vanillin, while the water-based solution does not.  For this study, we used the oil-based solution containing vanillin.

Evaluations of residual activity were conducted at a University of Arkansas Research Station in Fayetteville, AR to evaluate the residual activity of No-Fly.  The methods used for the residual activity experiment in 2011 were followed in 2012.  Each experiment evaluated a different concentration of No-Fly repellent and was repeated on three separate days.  The first concentration tested was the recommended 1:1 ratio of mineral oil to No-Fly and was conducted in May 2012. Heifers were divided into two groups: untreated control (n=11) and a treatment group (n=10).   Treatments were applied using the 3-D Quik Hand™ Cattle Sprayer (3-D Cattle Equipment, LLC. Pine Ridge, AR).  Visual fly population estimates were determined prior to treatment and then at one, two, and four hours post-treatment.  Data was collected on three separate days and all treated animals received the same application rate (approximately 130mL/animal) on each day and the data analyzed using analysis of variance with means separated using least significant difference where applicable.

After seeing no increased activity over geraniol alone or the addition of vanillin using the 1:1 solution, an evaluation was conducted using full strength, undiluted No-Fly repellent.   The same groups, application methods, and data analysis were used in the full strength evaluation. One week elapsed between studies without any treatments to allow remaining residual activity, if any, to cease.

2013: In 2010, a dust formulation of C8910 was evaluated (first study reported in this document), which was not effective in reducing fly populations.  In 2013, an oil formulation became available for evaluation.

Experiments were conducted at the University of Arkansas Research and Extension Center in Fayetteville, AR to evaluate the residual activity of the oil formulation of C8910.  Heifers were divided into two groups: untreated control (n=9) and a treatment group (n=9).   Treatments were applied using self-treatment mop oiler (Pest Doom Oiler, 45” x 7”, 4 gallon capacity) associated with the animals’ water source; each animal had to pass under the mop oiler each time they went to and from water. Being a self-treatment device, the application rates varied based upon use.  The average application was 24.8mL per animal per day.

The number of horn flies per animal was visually estimated weekly and analyzed using analysis of variance with means separated by least significant difference where applicable.