Monday, April 03, 2006

How to make durable cages for intertidal experiments

Stainless steel mesh cages have a long association with intertidal ecology and were a key part in some of the early studies that created the field of experimental ecology. The early work of Joe Connell (1961) is responsible for introducing most people to the use of mesh cages for manipulating the densities of predators and grazers. Connell used 22 gage wire mesh with 8 squares per inch, formed into 6"x6" square cages, to exclude predatory whelks from patches of barncles. All of your favorite intertidal ecologists have probably used mesh cages at some point in their careers, and there are a wide variety of cage styles and construction techniques out there. All of these different cages probably worked well, but I'm going to outline the single best method for making cages that do the job and are durable enough to someday hand down to your grandchildren.

Cages deployed in the field at Hopkins Marine Station

There is only one choice of material for making really durable cages: stainless steel. Plastic mesh works to a certain extent, but you'll have a hard time building a cage that will hold together under vigorous wave action out of only plastic. Regular steel obviously rusts, even when you buy the plastic-coated stuff. If you need to build three dimensional structures to mount on the rocks in wave-beaten areas, and you need them to last for more than just a few months, skip over plastic or other materials and step up to stainless steel.

Stainless steel mesh is available in a wide variety of sizes, as it has many uses in manufacturing. We'll base this instruction set on the assumption that you'll be ordering your stainless mesh from McMaster-Carr. McMaster-Carr specifies dimensions in inches rather than SI units, so this tutorial will abide by their conventions. You can search through the various mesh types by entering "wire mesh" into the search field at McMaster. Then choose either "woven wire cloth" or "welded wire cloth". The differences are explained below.

There are two alloys of stainless steel mesh usually offered, type 304 and type 316. Type 304 is preferrable for our uses, as it is nearly half the price of type 316. Type 316 is somewhat more resistant to corrosion by harsh chemicals, but unless you're doing your predator exclusions in some sort of sulfur-rich hot springs, type 316 will be overkill. Stick with type 304, which will not rust even after years in the field.

The size of the mesh that you use in your cages is obviously set by the minimum-sized organism you need to contain or exclude. If you're keeping out grazing limpets for example, you want to consider how small a shell size will fit through your mesh. Realistically, you'll have to balance the animal-filtering ability of your mesh versus the impact that the mesh will have on water flow and shading of the substrate under the cage. Keeping out the most voracious adult-sized grazers or predators will often be sufficient, and taking some care to open the cage to remove smaller interlopers every few weeks will likely be enough to remove the effect of that animal on the experimental plot. With this trade-off in mind, you can choose a mesh size.

Mesh is sold by a variety of dimensions. You can choose mesh based on the size of the openings in the mesh. An equally common method for shopping is to choose mesh by the number of squares-per-inch (i.e. 1x1, 2x2, 4x4). A 4x4 piece of mesh has 4 squares per 1" of linear dimension.
You will also need to specify the thickness of the wire that makes up the mesh. A good all-around wire diameter is 0.047". The 0.047" size holds up well for month or years, and is still thin enough to be workable when you're constructing the cages.

Mesh style
There are two categories of stainless steel mesh that we will consider: Woven wire cloth and welded wire cloth.

Welded wire cloth mesh on the left, woven wire cloth mesh on the right

Woven wire cloth is available in sizes from 2x2 through 16x16 and smaller. Welded wire cloth is only available in larger sizes, from 1x1 through 4x4. If you need to exclude very small organisms, you will end up using woven wire mesh in sizes from 4x4 through 8x8 or smaller. If you are going to build large cages with large holes to only keep out large organisms like bird and fish, you will need to use the welded mesh in sizes up to 1x1 (~1" opening).

Which style is better?
As mentioned above, for large-opening cages (>1/2" openings), you will be better off using welded wire mesh, and also using a larger, stiffer wire diameter (0.063" or 0.08"). The stiffness of a cage is tied to the mesh opening size and the mesh style (welded or woven). When you have relatively few strands of wire making up the cage (1x1, 2x2 cages), a woven mesh cage would not be strong enough to withstand flexing under breaking waves, as it would simply fold over. A welded wire cage will be inherently stiffer.

The stiffness of welded wire comes with some costs though. It is more difficult to make complicated shapes with welded wire, particularly circular features. Welded wire doesn't move and flex as well as woven wire when you're bending it, so the welded mesh will either start popping apart at the welds, or just get kinked. Welded mesh is best reserved for simple slab-sided square cage styles where only simple bends are involved, or perhaps hexagonal fences. Welded wire cages also fatigue after time due to their stiffness. A cage sitting in the field getting hit by a wave every 10-15 seconds during high tide for months at a time will undergo a huge number of flexing cycles, which will cause welds to start to fail and the integrity of the cage to fall apart. This is one reason for using larger wire diameters in cages made with larger mesh sizes, as it stiffens the cage and reduced the amount of flexing with each wave, reducing the amount of fatigue in the structure.

Woven wire mesh is a good choice for cages with smaller openings (3x3, 4x4, 8x8 etc). Woven mesh is easier to bend and form into complicated shapes such as circles. Woven mesh is also capable of being woven together at corners and seams, since you can leave longer pieces of wire hanging off your cuts that can then be woven. When you get into the very small mesh openings (6x6, 8x8 etc), welded wire mesh, if it was available, would be horrible to work with. At these smaller mesh sizes, the structural integrity of the cage is good because there is so much material in the structure. The completed cage can flex a little with each wave, but there are no welded joints to fatigue and fail over time. Woven wire mesh cages have no trouble standing up on their own in mesh sizes up to 4x4.

Cages, fences, cage controls and other structures come in a variety "standard" shapes and sizes. Some people prefer circular plots, others like simpler square or rectangular plots. All cages need some sort of flange to allow them to be bolted down to the substrate. This takes the form of a section of mesh bent out horizontal around the bottom edge of the cage. This flange should be >1.25" wide to give you some leeway in positioning the cage over holes in the rock. My cages usually have a 1.5" wide flange around the bottom. Stainless steel lag bolts get pushed through the mesh flange and screwed into plastic masonry anchors in the rock.

For all cage styles, try to make the whole structure out of one piece of mesh rather than trying to tie multiple pieces of mesh together into a structure. Structures bent up from one piece of mesh will be stronger than a pieced-together mess.

To cut the mesh, get a pair of tin snips. Tin snips come in straight, left, and right cutting styles. Any style of tin snip will be fine for this job. Straight-cutting tin snips are show on the right in the picture below.

Vise-grips on the left, tin snips on the right

If you have access to a jump shear in your shop, you can make large straight cuts quickly. This can be useful for cutting your big roll of mesh into individual pieces that eventually become the cages. You will still need the tin snips to make smaller cuts in the mesh for bending. Don't try cutting the mesh on a band saw, it will just get messy.

For making bends in the mesh, the best option is to find a sheetmetal brake in the shop. This will let you make most of the bends easily. If you don't have access to a brake, there are other ways to bend the mesh. The most burly method is clearly to bend it using your teeth. If you aren't up to that, I recommend finding a sturdy table with a top that sticks out a little ways. Using some c-clamps and a long piece of wood or metal, you can clamp one part of the mesh to the table, and then bend the other portion over the edge of the table. To get nice crisp bends you can take a hammer to the mesh to make sure it bends sharply at the edge of the table.

Bending mesh over the edge of a work bench

The round fence shown below was bent in this manner. The entire long strip of mesh was clamped to the edge of a workbench so that half of the mesh was sticking out. I then went along with a hammer and pounded the mesh down until it was bent at 90 degrees. After the entire piece was bent I took it out and put the circle bend in it. The cuts around the flange of the fence are spots where I had to cut the mesh to allow me to get the circular bend in the mesh. The fence is then closed with zip ties, or you can use any long pieces of wire sticking out the ends of the mesh to weave the two ends together.

A 20cm diameter fence made from one long piece of mesh

The square cage below was built in a slightly different fashion. The cage is 10x10cm, so it is fairly small. I made the bends in this cage by clamping it in a bench vise and bending the mesh over. The initial piece of mesh was cut to resemble a big + sign. Each arm of the + was bent down at 90 degrees relative to the center, and then each arm was bent out horizontal to form the flange. The four sides of the cage meet in the corners. I purposely left the wire mesh edges long when I cut the + shape. These long ends could then be weaved through the adjacent pieces of mesh to close the corners. Weaving the mesh together like this produces a very strong and stiff cage that is also fairly well sealed against animals getting in and out.

A 15x15cm full cage for excluding whelks

Excess wire on ends of mesh is folded over adjoining wire to tie the two sections together

The weaving process makes use of the other pair of pliers pictured with the tin snips above. These are small needle nose Vise-Grip pliers. I prefer these to normal pliers because they allow you to lock on to the wire while you're bending it, relieving you have having to squeeze as hard as you can the whole time. After weaving and bending a few hundred wires for a set of cages, your hands will thank you.

Leaving extra wire hanging off the ends of cut mesh

If you are using woven mesh, the ends of the mesh pieces will have loose ends sticking out. After time those loose ends can start to fray and pieces of wire can pull out. This often happens around the edges of flanges. The solution is to clean up the cut edges by bending the loose wire back over on itself, wrapping it around the wires running perpendicular to it. In the image below I have bent the loose wires around the corner over and under the wire that crosses them. This makes it impossible for the mesh to come apart, and has the added bonus of making the cage safer to handle, as you won't get poked by the cut ends of wire.

Loose ends of wire are folded over the crossing wire to tie the edges together

The other option is to take a material such as Z-spar epoxy putty and form it over the exposed edges of the wire mesh. This will help hold the wires together, and makes the edges smoother for handling. The downside to this method is that epoxy putties are brittle, so when you need to bend the cage around features in the rock you make end up cracking the putty.

Most of the above info pertains to woven mesh cages. When using welded wire mesh, the assembly process is somewhat simpler. You do not need to bend over the loose edges, as the welded wire will not unravel. You would still do well to leave extra wire sticking out at cage seams so that you can weave the seams together for a tight fit. This also helps strengthen the cage.

Mounting cages in the field
Cages should be mounted using the standard plastic masonry anchors and 1/4" stainless steel lag bolts.
Masonry anchors are available from McMaster-Carr, search for part number 95482A145. There are other lengths available besides the part number given here. A common lag bolt from McMaster-Carr is part number 92351A544, and again, there are multiple lengths available. A larger list of useful field equipment and pictures can be found here:

In the larger mesh sizes you can simply slip the bolt through the mesh openings and screw it into the rock. For smaller mesh sizes (6x6, 8x8), you will need to punch a hole through the mesh to get the screw through. My favorite technique is to just take the hammer drill that you use to drill holes for the blue anchors and drill through the mesh. I usually do this out in the field at the same time that I'm installing the blue anchors, so that I can easily line up the holes.

For all mesh sizes, you will need some sort of washer or plate to put on the lag screw to help spread the load on the mesh. The heads of the lag screws are fairly small and will pull through the mesh after a while. I use either large stainless steel washers, or pieces of plastic cut into 3"x1" strips with a >1/4" hole drilled through the middle. These help spread the load over more the mesh, and also help keep the mesh from folding up and allowing animals to sneak in.
For washers, a good example would be McMaster-Carr part number 90313A109. You can also simply search for "washer" at McMaster and choose a size that suits your budget.
If you're going to use plastic pieces instead of washers, drill the hole with at least a 17/64" drill bit. This will give you enough play to get the 1/4" lag bolt through.

One or two bolts for each side of a cage will be sufficient to hold it down to the rock. The image below shows a 4x4 mesh cage bolted to the rock using both stainless steel washers and plastic pieces.

A cage deployed in the field

That's more or less it for making cages. There are plenty of variations on this basic theme depending on what you're trying to accomplish.

Some other thoughts:
One common modification for cages that you need to open for censusing is to leave the top of the cage open (i.e. make a fence) and zip-tie a piece of plastic mesh over the top of the cage to seal it. The zip ties can be cut and replaced each time you need to access the cage. Alternatively you can unbolt the cage from the rock at census time. If you're doing point-contact counts in small plots, it will probably be easiest to simply unbolt the entire cage for counting.

For instances where you're dealing with small organisms that crawl along the rock surface, some people sandwich strips of neoprene rubber (wetsuit material) between the cage flange and the rock.

Cages will eventually get algae settling on them. This can be detrimental to your experiment, so you may occasionally need to clean the cages. Some people use stiff wire brushes to clean the cages. A more effective method is to kill the algae either by boiling or with direct flame. Chris Harley introduced me to the idea of bringing a Coleman stove out into the field and heating up a big wok full of water. Hot enough water will kill the algae and it will slough off during the next tide. A propane torch applied directly to the algae also works well.

Cage controls are a messy subject, and there are several styles out there. All of them have their downsides and introduce unwanted artifacts that differ from the full cage treatment. Cage controls all have some side or sides open so that the organisms of interest can move in and out. Some cage controls concentrate solely on mimicing the shading effect of the full cage, so they might have a piece of mesh held up on 4 posts. Other cage controls are concerned with also mimicing the effects of the cage on water flow, so they only have small sections of the sides removed, or perhaps two full sides removed.

A style of cage control with two open sides

The choice of cage control is up to you, but consider what factors you think are most likely to cause experimental artifacts when caging a plot, and design your cage control to *properly* mimic those effects.

Luke Miller
April 2006