Pond Boss Forums STOCKING A NEW POND Creating the food chain Food Chain in a 100% Clay Pond

I dug some more test pits at our property this weekend.

One of the good low locations had nice sandy clay that was getting more clay-rich with depth. Certainly enough clay to seal a pond.

I then moved up to some higher ground where the "upper end" of a 1-acre pond would be located. After cutting about 10" of top soil, the location was 100% clay. I got down to max. depth of 8.5' for my mini-excavator, and was still digging in clay. The clay sides of my pit were smooth and polished and packed very hard.

Question for the experts:

If I finished a pond in that clay soil, I think it would be a very long time before any aquatic plants began to thrive in the pond. The pond will be sourced with pumped water from an adjacent groundwater pond, so there will essentially be zero surface water inputs of silt and nutrients.

Will this pond only grow FA, or be exceedingly low productivity?

Do I need to build some nice flats around the pond shores at a depth of 3' and re-cover those areas with 6-12" of top soil to get aquatic plants growing in the pond to aid all of the organisms at the very bottom of the food chain?

Any experience from managing a similar pond or just any advice would be greatly appreciated!


Thanks, FishinRod.

P.S. I do have tons of cedars to clear from my fields. I also have lots of hedge apple trees that I can prune for crib material. I could build some wood cribs with the hedge and just keep throwing in more cedar trees as the previous cedars decay.

Is it possible to run a decent "natural" pond with just lots of surface area for periphyton to sustain the base of the food chain?

I have seen trout ponds and CC ponds that were essentially "sterile" clay basins. That worked in those situations because the fish were fed and intensively managed. I can't pull that off on our farm. (However, I would like to do a small pond adjacent to my proposed larger pond that is just a CC pond with feeder and a few RES.)

I am nowhere near a professional on the subject but I don't think a clay base will be any problem for growing plant life, I have my dam built out of crappy clay coal rocky mix, the top of it from the deepest pit in the pond, 20 plus ft, and stripped all the topsoil out of shallow areas that I could in the coves, the plant life is as vibrant there as in any place around the pond.
Clay is a misunderstood product, it will actually grow pretty decent crops as far as productivity goes, the molecules are just so fine and tight that it will not absorb or hold moisture that crops need to do the best in. Most of your pond weeds and plants are mostly hydroponic anyways, very seldom do they have a big root system, other then cattails and maybe some lilies, the rest can pretty much be raked out with minimal effort, mostly just growing off the water that is available to them with just enough roots to keep them from floating around. jmo

Thanks, that is encouraging!

It is hard to grow terrestrial plants on hard, packed clay. I transferred that expectation to aquatic plants. Sometimes it is good to be wrong!

Rod, no experience with heavy clay but I do like the idea of shelves. Anybody have experience with heavy clay soils? I’m talking productivity, not sealing

Get the soil tested for nutrients/pH, etc., etc. Just like a farmer would test their crop field soil. You might be surprised at what the test results are.

Your hard packed clay on land, - how much water does it have to grow plants?

FR,

Knowing you are in tall grass prairie country, you may be surprised at just how much organic matter and nutrients are in the topsoil. Sounds like you may have a hard pan underlying. To me its all good. However be careful of have much topsoil remains in the pond. Organic percent greater than 3% is borderline hyper-eutrophic and any shallow areas inundated that have that level or greater organic matter content will grow vegetation prolifically.

I have many places on the farm where there is 1-3' of top soil with excellent organic matter. That top soil then grades into sandy loam or clay loam subsoils. I then hit almost 100% clay at depths of 6-8'.

This area was 8-12" of good top soil with prairie grass roots and lots of other organic activity. The soil profile then went almost straight into hard-packed clay with very little transition zone. I only had a 6100# mini-excavator, but I could not even dig this stuff effectively. I am used to digging wet clay where you get a bucket-full and then push the break-out limits of the equipment. This was the opposite - I would scrape and scrape and barely get the bucket teeth to bite. It got a little better at depth when there was a tiny bit of residual moisture in the clay. I could scrape partial buckets in that stuff.

There is essentially zero penetration of my deep-root species of prairie grass into this clay. I don't know if the problem is poor soil quality of the type that might show up in a soil testing analysis as you suggested above, or if this clay is just too dense for the roots to penetrate? The bluestem and switch grass roots go to 6-8' depth in the other areas of the farm with loose, sandy soils.

The NRCS data says our average precipitation is 25-33" per year. The grass in this area is just barely observably thinner than the grass in the areas of deeper soils. Our blend of native prairie grasses is just mind-boggling (to me at least) at how efficient it is in capturing and utilizing the limited water it receives. I really couldn't believe how well the grass was thriving on this ground on such a thin layer of "usable" soil.

That is my expanded analysis of how my terrestrial plants are doing. I wouldn't even call the material that will form the bottom of a pond in this location a "soil". I would call it a clay horizon above bedrock. (No idea of the depth to bed rock. I have never hit bedrock in any of my test pits.)

I just have zero experience if that type of material would grow aquatic plants, so I have zero confidence on the plant cycle development in that type of pond!

Any more advice from you (or others) would still be greatly appreciated.

I was thinking on exactly those same lines - but wondering if a few hyper-eutrophic areas that prolifically grow vegetation would actually be a benefit to the pond?

I could have a few flat "mesas" in the pond that would be 3-4' below normal pool level. I could cut almost a vertical slope around most of the mesa and go straight to 8' deep water. (I think DD1 was thinking along the same lines.) I would deliberately finish the top of the mesa with good, organic-rich topsoil.

Would plant growth there help "steal" much of the nutrients from my pond water and mitigate the growth of filamentous algae? It would also provide a micro-environment for the organisms at the base of the food chain plus some shelter for the forage fish species.

The groundwater pond that will be my water source is going to be closely tied to the creeks in the area that exist in the same sand layer as the groundwater aquifer. These creeks certainly good a large dose of run-off from adjacent agricultural land. I suspect my water supply will have some residual fertilizer in it during certain times of the year. I may need some area of abundant aquatic plants to help soak up that input.

I think a properly designed pond could prevent any such prolific plant areas from spreading too much by going straight to deep water? Or is my likelihood of getting an undesirable plant that spreads through seeds or floating fragments so high that my entire shoreline from 0-3' will quickly grow enough plants that my pond will not lack for the plant portion of a healthy pond ecology?


(I read almost all of the threads on Pond Boss about vegetation and algae in ponds. This is one of the most common problems from new members. I am trying to solve some of these potential problems during the "design phase", so I apologize for giving everyone such vague data to work with! However, I am pretty confident that if I could give everyone perfect data, this crew could easily create an optimal design. )

I am not even close to being an expert but I do have a clay based ground water pond. After the 12" of top soil I have pretty clay soil with some sand veins (clay percent ranging from 19% to 38%). Then when I get to 8'-12' deep it turns to 100% clay. My water is what would be considered way too clear but is a pretty color. Anyways the pond is only 2 years old and it started to have plants growing in all of the soil types. Even at crazy depths I did not think stuff would grow at. I know I was worried about plants growing when I dug my pond. Even thought about planting some in there to get it started but life got too busy. I do have some woods about 30' from the pond which drop leaves in there which probably help with nutrients. I hope this helps.

Keep in mind the controlling factor of aquatic growth success is not the soil type, it's the available nutrients in the water that are used-otherwise it would be classified as terrestrial.
Decomposing organic material releases nutrients that can cause increased success or faster growth with aquatic vegetation but it has to be converted to be used.

Thanks, Snipe. That leads me to an additional question.

I believe terrestrial plants can only exchange CO2 and O2 through their stoma. Some terrestrial plants can handle submerged roots for shorter periods and some for years depending on the exchange mechanisms in the roots. (And that is what makes them "terrestrial" plants.)

Obviously, aquatic plants must be able to do the exchange through either leaves on the surface of the water, or leaves that can utilize dissolved CO2 in the water.

Terrestrial plants (generally) get their nitrogen, phosphorus, and sulfur from the soil.

This then leads to the specific question I should have asked in the OP:

Do aquatic plants have the ability to pull nutrients (nitrogen, phosphorus, sulfur, etc.) directly from the water, or do they utilize their roots to pull those nutrients from the bottom of the pond as those nutrients build up in the soil and muck?

Or possibly, do some classes of aquatic plants take nutrients directly from the water, and some must take up nutrients from the soil?

My belief is that they have the ability to pull the majority of what they need directly from the water, and that the biggest "job" of the roots is to anchor the plant in place. Remember, that there are some underwater plants that have no root structure, like coontail.

Take Eurasian Water Milfoil for instance. A broken part of the plant will float along the surface of the pond, grow in length and put out roots to anchor itself once it stays in one place near the pond bottom long enough. How can it grow in length and put out roots if it isn't taking what it needs directly from the water column itself?

Thanks for the additional info, esshup.

My main concern was that a hard-packed clay pond would primarily grow FA and other undesirable crap like milfoil.

I think I need to add planting some desirable pond edge species to take up some of the nutrients into the final pond design.

FishingRod,

The thread and questions have taken a lot of twists and turns but with regard to the food chain I think the mix of vegetables is of critical importance and the single most limiting factor in terms of the carrying capacity for fish. Varying populations of plants are the main contributor to variability in standing weights for waters of otherwise equal nutrient composition. Supported standing weights of fish are indeed highly variable and can be manipulated by controlling what veggies are growing.

Most (though not all) plants are beneficial when their proportion of the veggie biome is ideal. It's finding the right balance that is difficult part.

A couple of years ago I wanted to understand nutrient recycling (both native nutrients and introduced ... eg feed). What I learned from my study is that 3% organic composition of the pond soil is limiting. This is HYPER eutrophic nutrient loading. Organic composition of some prairie soils exceed this organic content and so any inundated native soil which is shallow enough to get sunlight will produce abundant plant matter. 3% is limiting because organics are mobile in pond soils. They tend to accumulate to this concentration but no higher. Rather what happens is they migrate deeper expanding the region of the 3% organic limit. This has been demonstrated in fish production ponds which are many decades old being fed at rates measured in tons per acre. So for a recreational pond where one wants good water quality ... creating an initial basin with the desired nutrient reserve ... I think ... is a very important consideration.

In the image below, I was attempting to model trophic nutrient cycling where understanding native nutrient reserves could provide a reasonable estimate of supported fish standing weights. The first tier from mobilization onward are average expectation of nutrient mobilization (of the standing organic pool) and the % organics of the soil is a variable of the spread sheet. The maintenance% is assumed where the approximate standing weight of LMB & BG is limited to 600 lbs per acre for a 3% pond soil. Though probably not exact ... it is not that far off when the lion's share of nutrients are being utilized by phytoplankton (as opposed to naiad, coontail, cattails, and such). Now I don't mean that you don't want plants other than phytoplankton ... what I mean is that you want these only in moderation at densities that support fish standing weight (as habitat/cover).



Now one of the things I will mention is that "having no nutrients" can be dealt with by feeding or fertilization. The 40% assumption of dry protein % is reasonable for dried phytoplankton but also for a high quality feed. So imagine the feeding as a mobilization of nutrients. Now two things.

1. Feeding bypasses the food chain and so could feed LMB (or BG) directly but we are also interested in its contribution to the food chain.

2. The second part is maintenance and growth (of the standing weight). If the standing weight is maintained annually with no harvest of fish then all of that is assimilated is recycled (in other words none of the feed nutrients are sequestered in fish). What is assimilated is determined by the digestibility of the feed. If 100% fish meal ... the digestibility will be between 70% and 80%. Digestibility of most feeds are much less than that (it really depends on how much high quality animal protein is in the feed). Most all of what binds the feed or spikes the protein % (think spikers like soybean meal) are not going to be assimilated by BG or LMB. So these kinds of additives are going to be residue that will decay as a declining percent each year. In the first year, probably around 50% of this residue nitrogen/protein can be recycled as mobilized nitrogen. This declines to around13% annually by the 5th year. After about 5 years about 22% of the original waste remains and is decaying more slowly with each year.

Now this is important. Of the amount assimilated ... all that is used as maintenance will be recycled as ammonia and this is readily available to plants. Also natural morts are volatile in terms of their nutrients and quickly recycled. So it depends on the quality of the feed just how much of the feed will be recycled to grow plants when the feed is supporting the maintenance of fish (higher quality feeds have higher potential for mobility because they have higher digestibility). When a fish grows, the nutrients are sequestered until those periods where the nutrients are mobilized by the fish for metabolism (think periods where prey consumption is insufficient to cover metabolism). When the fish mobilizes nutrients in its flesh for metabolism then the nutrients are mobilized for plants in the pond as waste products ... also all natural morts return nutrients to the pond for plant use. For nutrients sequestered in the growth of fish, the only way to prevent their eventual return for use by plants is to harvest fish. This increases the amount feed going into growth and reduces the amount of feed going into maintenance. Harvest also improves the production of fish flesh and the growth rates of remaining fish.

With fish meal, it takes almost exactly the same amount to grow a weight of fish as it does to maintain the same weight of fish for a year. When feed only maintains that as a limiting standing weight there is no conversion and no gain in the standing weight. All of the digested nutrients return to water via maintenance and natural mortality. The greater the percentage of standing weight harvested (at end of years growth) ... the better the conversion. Weight loss of the standing weight for maintenance and natural mortality will reduce the conversion and remobilize the nutrients.

Modest rates of feeding don't contribute a whole lot to mobilized nutrients. IOWs, the level of mobilized nutrients will be more driven by other sources of nutrients. A ponds mobilized nutrients primarily come from pond soils, leaching of nutrients from the watershed, and deposition of organics (think topsoil erosion, cattle and birds, and atmospheric deposition ... in that order). 3 to 10 pounds/acre of nitrogen are deposited by rain and snow every year. Feeding a 40% protein feed at the rate of 100 lbs/year will contribute ~ 6.5 lbs of Nitrogen and if fully mobilized for use by phytoplankton this would only support 26 lbs of BG and LMB. So you can imagine where 100 lbs of feed will maintain a 100 lb standing weight of fish (without a food chain) that feeding 100 lbs per acre year would support less than 126 lbs of BG & LMB(with recycled nutrients supporting a food chain). So how much one should feed really depends on non-feed mobilization of nutrients. In a normal year, the deposition of nitrogen from the atmosphere could be comparable to 100 lbs per acre of a 40% feed. All ponds accumulate nutrients and so have the potential to grow carrying capacity over time. Eventually, they have too much. There are a couple of ways to approach this to maximize the life of the pond. You could design the nutrient reservoir to support the standing weight desired (for me a mesotropic trophic level) and allow natural accumulation to eutrophy the pond OR you could start with less than the nutrients needed and supplement nutrients until the pond supports that standing weight (for me a mesotropic trophic level) on its own. At this stage controlling the mix of plants to maximize fish carrying capacity.

To reduce the accumulation of nutrients I think the following is important.

1. A good buffer zone around a pond denying direct access by cattle.

2. A sediment pond if possible to catch eroded organic laden soils before they enter the pond. Maintain it also on a reasonable schedule ensuring that the deposits cannot leach back into the pond (IOWS. take the deposits out of the watershed)

P.S.

There was a paper I read regarding standing weights supported at differing trophic levels (in Florida lakes). A number of variables were analyzed (eg secchi depth, nitrogen, phosphorous) and the results definitely showed correlation of fish standing weights with increased nutrient levels ... BUT ... there was a whole lot of variability. For example, for a secchi of 5 ft the standing weight supported could be as low as 32 lbs per acre or as high as 253 lbs/acre (averaging 80 lbs/acre). One of the things I was interested in understanding is what gives some waters higher standing weights over others. One is the mix of fish where trash fish significantly increase standing weight. The other trend I noted was the average depth relative to secchi depth. There were definitely higher standing weights where lake average depths exceeded the secchi depth considerably. Where the average depth was less than secchi depth, these were consistently the least productive water. Having depth in a pond gives greater space, more potential for oxygen storage, and inhibits macrophytes over a large portion of the water body. I think this is important for designing BOWs to have higher natural carrying capacity.

Hey FishingRod,

Since this seemed to have interested you, I will share some other findings. The model above suggests that ~4.165 lbs of LMB & BG can be supported for every 1 lb of Nitrogen mobilized. This should be thought of as a maximum number and that other factors could reduce the weight carried. So one of the questions I wanted to answer was how can the full effect of feeding be modeled where feed wastes are utilized as fertilizer for the pond food chain. This paper is study of the production of BG/LMB where the feed formulation (Auburn #2) can be found in in this publication. Essentially the feed is 32% protein catfish feed. In the first reference paper, after feeding for 4 years there was 2307 lbs of fish produced, of which 1129 lbs were standing at draining and 1178 lbs were harvested. So I worked up a simulation to model this realized scenario which is depicted in the image below.



You will notice that by year 4 the mobilization of nutrients for a constant feed rate is relatively stable and only grows marginally thereafter. By year 4 nitrogen mobilization from the feed input is supporting a food chain that can support 423 lbs of fish. But the carrying capacity is 1129 lbs and so feed is supporting 706 lbs of fish on its first pass. So another way of stating the same thing is that food chain effects from the utilization of feed wastes increased the carrying capacity by roughly 60% as a secondary effect. An additional way of stating the same thing is that feeding increases carrying capacity by 166% over the food chain effect. So in a very lean pond (one with essentially no nutrients) it is possible to custom design the food chain nutrients and carry more fish than its mobilized nutrients could otherwise carry. Another way of saying the same thing is that one could design a meso-eutrophic trophic status that supports a carrying capacity that would otherwise require a eutrophic or possible hyper-eutrophic trophic status if it were solely dependent on the food chain. The carrying capacity is maximized by a basin that starts essentially devoid of nutrients allowing one feed more without exceeding the desired nutrient mobilization.

So whether one wants native nutrients or introduced nutrients to support his carrying capacity depends partly on budget and partly on how far one wants to push the trophic status of his water. The combination of native nutrient mobilization and introduced nutrient mobilization should be below this manager imposed limit. Hyper-eutrophy will occur and water quality will suffer when nitrogen mobilization exceeds 144 lbs of N per acre-year. Carrying the greatest weight of fish at the best water quality is facilitated by having a nutrient poor basin, preventing deposition of non feed nutrients, and feeding to fill out the trophic goal. In this sense, water quality drives the usage of feed where carrying capacity is maximized in the most nutrient poor basins which allow greater feeding rates for a particular goaled nutrient mobilization.

The simulation in the image above ties with the earlier simulation on native nutrients and also ties precisely with the results in the paper on feeding LMB-BG ponds with the Auburn #2 feed. I will mention that the simulation is caused to tie with the paper by adjusting the protein sequestration rate in the spreadsheet. Protein sequestration is the amount of protein sequestered within the gain of fish biomass. From this we can calculate the Gross FCR of the Auburn #2 feed. Sequestering 13.44% of the protein of a 32% protein feed produced 706 wet pounds of fish when the fish is 80% water and 66.7% of its dry weight is protein. This yields an FCR of 3.1. Not great ... which is what should be expected of BG when fed a feed formulated for catfish. We can further tie this to energetics of fish meal conversion. I earlier posted a thread on the conversion of GAM to LMB and in the reference paper it was found that LMB can convert 50% of the energy in GAM to its own energy content and that it can digest 80% of the energy in GAM. If we assume that plant proteins are not digestible by BG and LMB and that BG and LMB can similarly convert fish meal (a good assumption in that both LMB and BG have equivalent wet weight energy density) then we can predict the amount of fish meal protein (or equivalent) that would be required in the feed to produce the observed results. So the proportionate weight of protein sequestered is (.32 * .1344 = 4.3 % the weight of feed). At a conversion of 50% the required amount of protein in the fish meal is (4.3% / .50 = 8.6% the weight of the feed). Since the fish meal they used was 60% protein the proportion weight of fish meal required is (8.6% / .6 = 14.33% the weight of feed). As it turns out, the #2 feed is 12% by weight fish meal but it is also 5% by weight blood meal. So animal proteins only exceed the 14.33% required by 2.67%. Its pretty dang close ... don't you think? I suspect that most of difference lies in the lower energy content of the proteins in blood meal and modestly less than the 50% conversion efficiency I earlier assumed.

Now I want to share one other thing. Let's remove the harvest of 35.5% of the standing weight. This is shown in the image below. Notice how it hardly affects the year 4 carrying capacity? IOWs, abstaining from the 403 lb annual harvest would only add 33 lbs of fish. There is a very good reason for this. The harvest reduces the quantity of feed/forage used for maintenance allowing that feed/forage to convert as gain. When there is no harvest, the feed/forage goes to maintenance unless natural mortality occurs. The reference's experiment demonstrated that 35.5% of the carrying capacity can be harvested and the weight regrown each year. Swingle had such great intuition in postulating that fish will grow into maintenance ... something that is now corroborated by the findings of energetics. Abstaining from harvest CANNOT help one grow a significantly greater weight of fish. Abstaining from harvest can at best increase the number of fish (which of course means they must be smaller than they would otherwise be without harvest).

That is a very interesting concept that you are describing.

If my pond is initially too nutrient lean, then I can fairly readily turn the balance to my favor with a feeding program, and the secondary effect of the nutrient fertilization will also help the rest of the food chain.

By having an artificial "control knob" it might even be easier to run closer to maximum carrying capacity and then pull the fish back from the precipice by turning off the feeder compared to NOT being able to cut back the food chain in a naturally nutrient-rich pond. (I am not planning to run near maximum carrying capacity, but mistakes do happen.)

P.S. Thanks for taking the time to type up your long evaluations.

You're welcome Rod.

How "hot" the treatment pond in the study was depends on how much native fertility there was. The feed itself would create a eutrophic water condition (in a 1 acre pond) but definitely not unacceptable if the native fertility were very low. The feeding rate being ~10 lbs daily of a 32% protein feed. On the other hand, couple the feed nutrients with a native carrying capacity of 170 lbs of fish/acre and then the nutrient mobilization would be hyper-eutrophic. One should keep in mind that the standing weight of fish greatly exceeds what it would be if it were completely dependent on the mobilized nutrients ... roughly double ... and so there is twice as much competition by fish for space and oxygen. In other words, they are under greater risk of tragedy than under the same nutrient mobilization conditions but no feeding (which tragedy is common in hyper-eutrophic water especially without aeration).

There is a lot of freedom for customization ... ranging from very modest where clear water is desired to the very edge of reason for those who cannot get enough or who are never satisfied.

Nutrient mobilization can be reduced by increasing the protein sequestration rate. Essentially this means feeding a feed with a greater proportion fish meal and/or animal proteins. The catfish feed only converts around 13.5 % of its protein to BG flesh. By my reckoning, 40% feeds formulated for BG convert at around 18% of its protein. So this means less protein and thus less nitrogen needs to be added to support the same carrying capacity. The improvement in conversion efficiency results in about a 26% reduction in nutrient mobilization. Equivalent results of the example above would require 7.36 lbs daily or 1325 lbs annually of a 40% feed converting 18% of its protein. At $1.50 per pound .... approximately $2000 per year. At a harvest rate of 365 lbs/acre year the cost per pound of harvested fish $5.47 per pound. That's not bad given that it more than triples the carrying capacity of fish the nutrient mobilization would otherwise carry ... again assuming the basin is of very low fertility. The example of the study should probably be limit that anyone should attempt and only then when aeration is in place. For me, I like something modestly under the middle.

I am a fisheries biologist for a private pond and lake stocking company in MO. In highly productive mid-west soils we rarely have a problem with not having enough nutrient load to sustain high levels of growth. We typically have the opposite problem with hyper-eutrophic systems where dissolved oxygen will likely end up being the limiting factor of productivity. Aquatic systems are usually phosphorus limited and occasionally nitrogen limited, never seen a case of potassium limitation in a pond/lake.

In a commercial production setting we do fertilize our ponds with 16:32:0 to jump start/super charge our phytoplankton and zooplankton production. Most species we raise readily accept a commercial diet and we supplement their diets with that. Fertilization isn't necessary in a natural setting and can be detrimental to water quality, we regularly run commercial aeration and flush ponds with fresh well water during summer months. A feeder can be a great addition to a farm pond/lake if the landowner is trying to push the growth of their fish.

The most important factor for growing your fish in a farm pond I have found is to have correct habitat for spawning of forage fish and by stocking the forage fish (minnows and panfish) prior to any predatory gamefish. Many people think deeper water is better because it us supplying more water/area for the fish to grow, but actually most fish spawn in shallow flats and shallow water will be the most productive for zooplankton, phytoplankton, and invertebrates. Deep water is great for over-wintering fish and providing additional water volume to reduce oxygen depletion. So having a variety of depths and habitats is important as different species will have different requirements, and each species needs will change with the seasons.

Your clay is a blessing, make sure you get your dam cored and sealed well and will help minimize water losses through dry summer months (I am from central NE originally and know how dry the summers can be). As jpsdad has pointed out, your soil will have ample organic nutrient load for you plankton bloom.

I think you’re right. Lusk told me a long time ago to concentrate on the forage and the predators will be just fine..

I saw a little video on this where they installed tractor rims in the clay and then filled the rims with topsoil to grow some submerged vegetation in. Below the rim was clay for the seal on the pond and the rim also prevented some of the side creeping of the plant (at least initially). I'd guess this could be done with anything from landscape edging to concrete blocks or rocks.

Where there's a will.....there's a way.