Progress report for FNE22-015
Project Information
We will address whether reduced fish protein feed can achieve similar growth, filet quality, and economics for brown trout as standard fish protein/ fish oil commercial feed. We will use an all-alternative feed experimental diet (a), diet with 50% fish meal reduced (b), diet with 25% fish meal reduced (c), commercial fish meal diet (d) to assess differences in feed conversion ratio, digestibility, weight gain, filet quality and Poly-unsaturated fatty acid content (PUFA), as well as the cost in labor and materials to make fish food. We are interested in whether the cost to benefit ratio, including sustainability, of land-based aquaculture (LBA) can be improved by producing feed in-house with light-duty pellet pressing and extrusion machinery.
|
Inputs |
Activities |
Participants |
Knowledge Gained |
Action |
Condition |
|
USDA NESARE Funding |
Conduct feed experiment |
Canopy Farms |
Feasibility of in-house feed making |
Took steps towards more sustainable fish operation |
LBA industry is a more sustainable protein source for human consumption |
|
USM Interns |
Fish husbandry techniques |
Acquired knowledge and skills for employment |
|||
|
USM Faculty Advisor |
Efficacy of alternative protein sources |
Can recommend alternative protein sources to small producers |
|||
|
LBA community |
Analysis of sustainable in-house feed |
Made informed choice between feeds |
The documentary Seaspiracy has created a furor in the public in 2021, taking on overfishing, plastic pollution, and the treatment of marine animals. The film spent little time on aquaculture, firing broadsides into farming techniques and waste management, but saved its most pointed criticism for the feed industry. The Aquaculture Stewardship Council felt it necessary to respond, posting to their website that “35% of fish meal comes from fish by-products,” and 88% of global fish harvest is “utilized for direct human consumption” (ASC 2021). Capture fisheries have flat-lined at 90 million tons per year (Richie and Roser 2019). Still, 10.8 million tons of fish are not going to direct human consumption; with potentially as much as 7 million tons going directly to fish meal. Indeed, 15 million metric tons of fish were rendered in 2016 (Kok et al. 2020). Fishery regulations have reduced the tonnage of fish going directly to meal, even as aquaculture grows globally, and thus has put a premium price on fish meal and fish oil. No matter how you grind it, this represents a sustainability problem for aquaculture.
Feed is the largest expense for any aquaculture operation. Even small operations spend significant resources on feed. Ethical and appropriate waste disposal is a logical secondary problem. Third is perception: does the product grown in an enclosed system have sufficient quality and integrity to meet consumer perceptions? For small operations like Canopy Farms, making our own feed is a way to meet the triple-bottom-line of increasing net farm income, reducing environmental risks, and protecting natural resources. Invertebrate protein and oil sources have a track record of making acceptable ingredients. They can also reduce the farm waste stream by recycling waste into invertebrate mass. Feeding freshwater fish bugs makes intuitive sense to many people since that is what they eat in the wild. We believe that relatively simple feed formulations can provide good growth for small operations, reduce costs, strengthen financial and environmental sustainability, and improve public perception.
Market forces have forced feed suppliers to look at other ingredients, including plant-based, animal by-products and other novel ingredients to keep the price of feed competitive (Kok et al. 2020). However, not all protein substitutes work for all species. Soybean meal, a popular, inexpensive, and readily available ingredient, contains anti-nutritional factors (including antigen proteins, enzyme inhibitors and other sugars) that has limited its use as a feed ingredient (Li et al. 2020). Indeed, farmers who tried early versions of soybean supplemented feed have sworn them off after carnivorous fish developed intestinal diseases. Black soldier fly larvae, an insect based protein substitute, has shown mixed yet promising results as a supplement. Around the 30% replacement for fish meal, poly-unsatruated fatty acid (PUFA) content begins to drop off and fish fillet appearance begins to change (Bolten et al. 2021). Many of these anti-nutritional factors can be alleviated to some degree by using additional supplements, enzymes or post-harvest treatments of feed ingredients. These formulations assume that fish eat the same meal every day, all the time, whereas in the wild fish eat a varied diet, both seasonally and daily. A mixed ingredient diet, consisting of a variety of animal, plant, and insect based protein sources, along with digestibility aids and probiotics, may be the key to producing the most sustainable aquaculture fish.
The remaining question is whether producing fish feed can be done economically on-site by small to medium sized farms. The cost of shipping can easily add 50% or more to a feed order. If a producer is using dozens of pounds of feed a day, making one’s own fish feed can garner long-term savings and more comfort in where those ingredients are sourced.
Advancements in feed could have significant impact on the growing aquaculture and aquaponics industries. Aquaculture in the US is expanding as we seek to raise more sustainable protein sources for a growing global population. Commercial aquaponics farms have become more prevalent over the past decade; however, high costs and high risk (including supply chain risk) limit new entries to the field. Currently valued at $630 million worldwide, the aquaponics market is expected to grow approximately 15% annually over the next 5 years (Research and Markets, 2020). Reductions in feed cost and risk would increase the attractiveness of this efficient growing technology for new farmers. With a robust existing aquaculture industry including leading aquaponics farms, alongside a history of small-scale agriculture, the Northeast is poised to become a leader in small-scale aquaponics and LBA.
Now more than ever with inflation driving prices higher, costs to run land based aquaculture and farms is increasing. To stay afloat, we need to be innovative to reduce costs wherever we can to compensate for increased fixed costs of prices, such as materials from local hardware stores and increasing electric and utility rates.
Canopy Farms is a 3500 sq ft aquaponics greenhouse attached to an Asian market and café. We raise Trout in the basement of the facility, the Asian market, café, and commercial kitchen is on the main floor, and on the roof is an enclosed greenhouse, utilizing the fish water from downstairs to fertilize our wide variety of crops such as greens, tomatoes, strawberries, bok choy, and chard. It might be the only building of it's kind, with plans occurring right now to install small scale shrimp aquaculture in the basement as well. We started construction in 2018, but weren't operational in sales until 2020-21. Since then, we have set up sales channels through our own CSA, the Portland Food Co-op, Bath Natural Market, and the Portland Farmer's Market in addition to providing the restaurants with produce and fish. We grow enough to supply those channels year round on 0.04 acres, a great benefit of aquaponic farming. We produce over 1000 pounds of produce each year.
We have a Pentair pump and air blower as our two key pieces of equipment at the farm, as well as a UV drum filter and Ozone generator for water filtration. Funded by SARE, we now own two chillers to raise Trout fingerlings to continue experimenting on for further research into feed viability and diets.
Cooperators
- - Technical Advisor (Educator and Researcher)
Research
Materials:
CF has two 2000 gallon aquaculture tanks as part of a RAS located in their Brunswick, ME facility. CF will also have access to three 100 gal, two 150 gal and one 300 gal RAS tanks, all of similar design at USM. USM also has a fully mechanical feed pellet press (5mm die) and a fully mechanical extruder used in previous experiments to make fish food. Both are inexpensive “made in China” units purchased on Amazon.com. Additional fabrication was required to make the units usable.
The commercial feed used will be Bio-Oregon, which has a warehouse facility in Westbrook, ME, less than 10 mi from the USM lab and 30 mi from the Brunswick facility. This will offer considerable cost savings in terms of shipping feed, a luxury not enjoyed by most producers. An average shipping price will be added to feed cost calculations to provide a more typical estimate of feed economics. We will follow the manufacturer’s recommended progression from crumble to full pellet.
Rainbow trout will be acquired locally. There are two fingerling suppliers in Maine. Fish will be acquired at 5 cm and grown to 10 cm on commercial fingerling feed. Once that size is reached 180 fish will be moved to the USM facility to begin the experiment. Feeding will start with a 10 day transition from commercial feed to the alternative feed.
|
Ingredient |
0% |
25% |
50% |
Control + BSFL |
100% control |
Source |
|
Fish meal |
0 |
12.5 |
25.0 |
50 |
50 |
market |
|
BSFL meal |
15 |
11.3 |
8.0 |
33a |
|
Enviroflight |
|
Soybean meal |
15 |
11.3 |
8.0 |
|
|
local |
|
E. fetida meal |
8.9 |
6.7 |
5.0 |
|
|
CF & USM |
|
Distiller Grain |
12.5 |
9.4 |
7.0 |
|
|
local |
|
Brewer’s yeast |
7.5 |
7.5 |
5.6 |
|
|
local |
|
Macro-algae |
5 |
5 |
5.0 |
|
|
local |
|
Salmon oil |
5 |
5.4 |
5.3 |
22 |
22 |
market |
|
Oat flour |
30 |
30 |
30 |
|
|
market |
|
Ascorbic Acid |
0.025 |
0.025 |
0.025 |
|
|
market |
|
Dicalcium phosphate |
1 |
1 |
1 |
|
|
market |
|
Vitamin mix |
0.1 |
0.1 |
0.1 |
|
|
market |
|
Other |
|
|
|
28 |
28 |
|
Table 1: Percentage of ingredients to make up fish feeds based on percent composition of fish protein in the feed mix. Proportions of ingredients in control not available at this time. a:This formulation will use commercial fish feed for 2/3 of the fed weight and whole commercial BSFL for the final third.
Water quality parameters monitored will include ammonia, nitrate, phosphate, temperature, dissolved oxygen, carbonate hardness, and conductivity. Fish will be maintained at 10C – 20C with 80% oxygen saturation for the duration of the experiment. Conductivity will be maintained near 1000 uS/cm. Ammonia levels will be maintained at less than 0.25 mg/L and nitrate at less than 150 mg/L. Water turnover (blow down) will be 10% per day. Feed composition may alter these parameters. When necessary additional water turnover will be used to control deviations or feeding may be suspended temporarily. These occurrences will be noted.
Methods:
At USM we will feed the five formulas, one formula to a tank. To accommodate the number of formulations we want to test and the resources available, this project will be a qualifier round to determine the two best formulas to move forward in a later NESARE grant proposal. In that future study, the best performing non-control feed will be fed to one of the CF 2000 gal tanks and the control will be fed to the other.
|
Tank |
2022 Feed Exp |
Future Feed Exp |
|
CF1 (2000 gal, 200 fish) |
Control |
Control |
|
CF2 (2000 gal, 200 fish) |
Control+BSFL |
2022 winner |
|
USM1 (120 gal, 28 fish ea) |
0% |
Control |
|
USM2 (120 gal, 28 fish ea) |
25% |
2022 winner |
|
USM3 (120 gal, 28 fish ea) |
Control |
Control |
|
USM4 (150 gal, 33 fish ea) |
50% |
2022 winner |
|
USM5 (150 gal, 33 fish ea) |
Control+BSFL |
2022 winner |
Table 2: Tank configurations and feed to be used, sorted by location and year.
Actions:
Fish will be grown out on commercial crumble for four weeks after receipt. At 10cm avg length 180 fish will be moved to USM. At CF 200 fish will be stocked into each of the 2000 gal production tanks. We will take 10 days to wean fish off the commercial crumble to a larger sized alternative or comparably sized commercial feed.
The 2022 experiment will occur in three eight week intervals. At the end of each interval six fish will be euthanized. Two fish from each treatment will be sent off for gut and liver histology, two will be sent off for biochemical analysis, including PUFA content, and one from each treatment will be sampled by the CF head chef for off-flavor or texture abnormalities. Fish to be sampled will go through a standard 3-day purge, often used in the LBA industry to remove off-flavor. Length and weight will be recorded from a representative subsample of fish at the end of an interval to estimate condition. We will also track generation of waste from fish feed by weighing drained-to-damp (24 hrs in 100µm mesh bag) solids removed from the systems, measuring % suspended solids with an Imhoff cone, and measuring water turbidity.
At the end of 24 weeks the remaining fish will be moved into a grow-out mode. As fish reach appropriate size for restaurant use they will be euthanized and moved to the CF kitchen. We will not recombine USM and CF fish to maintain biosecurity.
Measurements at the end of each interval will provide three measurements of a tank’s population through time. Those measurements will be used to look for significant change from interval to interval, i.e., through time, as well as between feed formulations and between each formulation and the control.
We will use ANOVA methods to ascertain which feed resulted in better growth and value over the course of the experiment. A repeated measures ANOVA design will be used to compensate for a single system used for each diet. If appropriate, we will use Tukey’s Post-hoc test to determine differences between treatments. Line graphs of continuous variables (e.g. growth) should allow visual separation of treatments based on fish performance. If necessary, differences in tank size can be accounted for using dummy variables in the ANOVA design or with separate tests of relative change regressed against tank size.
CF fish will be analyzed separate from USM fish with t-tests because of the vast difference in size of enclosures.
|
Diet Treatments |
Interval 1 |
Interval 2 |
Interval 3 |
Final |
|||||
|
Grams growth, solids, ammonia, histology score, % suspended solids, condition, pH |
Grams growth, solids, ammonia, histology score, % suspended solids, condition, pH |
Grams growth, solids, ammonia, histology score, % suspended solids, condition, pH |
Grams growth, solids, ammonia, histology score, PUFA score, % suspended solids, condition, pH, chef score |
The most significant change in our experiment was the decision to switch from rainbow trout to brown trout in August 2022. Because 2022 was the first year Canopy Farms was fully operational, an additional 3000+ gallons was cycling upstairs in the summer heat, which caused our downstairs tanks to increase in temperature to a level unsuitable for rainbow trout in the dead of summer. We also had a large die off of rainbow trout fingerlings due to a power outage in July that occurred at 2am when no one was at the farm and available to receive the outage alert to start the generator. Brown trout, being a slightly more hardy fish with the same market price as rainbow trout, became the more viable option, as well as the only one left so late in the season to purchase fingerlings.
Our increased sales channels have helped the farm's dependence on grant funding, and we hope with continued efficiency improvements like constructing vertical PVC pipes above our beds and repairing grow lights, we can continue to raise profit viability. In 2021, only half the greenhouse had working LED lights, limiting growth and harvestable produce. We hope to be at 100% dependable coverage with backup lights ready to be replaced if needed by early 2023.
Education & Outreach Activities and Participation Summary
Participation Summary:
This year, we gave over 25 tours of our facility and reached over 300 people with information about aquaponics, greenhouse growing, aquaculture, and local opportunities at our facility. In 2023, we already have school tours with surrounding middle and high school students scheduled, as well as a partnership between Mt Ararat High School students to assist in building an aquaponics system in their classroom. We also are a host site for the University of Southern Maine MEDF intern to develop the aquaculture/agricultural workforce by giving an opportunity for students to gain first hand experience working on a farm, aspects of running a business, and aquaculture practices. We hope to continue this work of education into 2023 and expand our programs by adding a summer intern and getting connected with more schools in our area.
Learning Outcomes
This year was the first year that the participating farmers at Canopy Farms raised trout fingerlings, so after this year we have increased our awareness of the water quality parameters, best practices in temperature management, and timeline management of when to receive fingerlings, when to transfer, and how much to feed/stock in density. This will inform our second phase of the experiment with a greater survival rate of fingerlings.
For education to the community, we have learned and would recommend (for Maine) running interpretive programs for indoor sustainable agriculture in the winter, where there are less activities for the public to attend, therefore boosting attendance at local "science cafe" nights we hosted at the farm. We found people interested to come to our event, and depending on funding would like to run this program again and invite other local farms in our area to discuss sustainable agriculture with the public.
Project Outcomes
We have found that brown trout fed a diet with mixed black soldier fly larvae meal (BSFLM) and fish meal perform nearly as well as fish fed commercial fish feed. Brown trout fed a homemade fish feed performed as well as fish fed a commercial feed. We have had zero mortality and good growth from our 100% menhaden homemade fish feed. Our homemade pellets are pressed, not extruded. Despite this we are seeing good growth from the fish and palatability of the feed. Our mixes with 50% BSFLM and 80% BSFLM have also performed well, showing good palatability of the food.
Our treatments with 100% BSFLM and commercial feed + black soldier dried larvae (COM+BSFL) have not performed as well. Fish tend to be unevenly sized in these tanks with high mortality. This is an indicator that some fish, as high as 50% in the COM+BSFL treatment, are not finding the feed palatable. We see large size differences and smaller fish becoming emaciated and more prone to disease (fin rot) and intraspecies aggression that leads to infection and death. In these treatments we have seen 30%-50% mortality by the end of the first month of the experiment. This is compared to 0%-8% mortality for the other treatments. Mortality in the COM+BSFL treatment was severe enough that after 30 days we opted to drop this treatment from the experiment.
Our experiment has another month to go before we reach the benchmark. At that point we will weigh and measure all the fish to collect data for growth and condition indices. We will also plot our water quality data. Observationally, the experimental feeds have held together well and not contributed to declines in water quality that the fish experience in our small recirculating aquaculture systems.
Although not complete, we see promise in the 50% feed formulation. This mix could result in a formulation that cuts fish content of RAS feed in half. This is particularly exciting because the replacement protein, BSFLM, comes from an organism that is very good at consuming human waste stream food. Our food also takes advantage of other soy and grain waste streams. Small operation RAS has a tenuous financial prospectus. If we can help farmers save on feed by running their own $1000 mill press once a quarter that could be a boon for the sector.
Thus far our experiment has addressed the issue we set out to study. We developed, collected the ingredients for and made four experimental fish feeds in house that adjust the amount of fish meal. Unfortunately, it does not look like completely removing fish from the feed for brown trout is an option. At least not with BSFLM as the primary ingredient.
Our experiment was plagued by challenges from the absurd to the sublime, some of which were avoidable. Our previous experience was with tilapia and rainbow trout in large systems. Brown trout are more aggressive, more inclined to jumping and escape attempts and more likely to overeat than either of those other species. In preparing, our primary concern was meeting the temperature requirements of the species so some of these other problems caught us off guard, like how short the season to order fingerling trout is. Collecting ingredients during a global pandemic is more difficult than one might think, etc. Some of these observations come only from experience with a new species and new process. We will certainly be better prepared for the acclimation and acquisition processes in the next iteration of the study.
The feed formulation ended up being more complicated than we originally thought. Researching the final feed formula pointed out a number of anti-nutritional concerns that we had not considered when first proposing the experiment. In this first iteration we went all out to find means in the literature to remove these anti-nutritionals. Acquiring and treating ingredients extended the prep time to starting the actual experiment. We're curious if these steps were really necessary. In other words, if we run pressed food with anti-nutritional elements treated against feed without anti-nutritionals treated is there a growth or survival difference? Extruders are the standard in making fish feed. Are the anti-nutritional treatments still necessary given the heat and pressure that are part of making extruded fish feed? Will fish grow better on the extruded feed than the mill pressed feed?