It sounds like you are allowed to place material on the wetland concerned, as long as it doesn't compress, and as long as you can remove it afterwards.  (Is that the case?)  In connection with the restoration after removal, you should consult an ecologist and other bio-professionals.

I suggest that, unless you suspend your construction above the wetland, you place a geotextile before placing any other material.

Choices:

1) Use a lightweight material, such as blocks of expanded polystyrene to form most of the embankment.  This is very expensive.  Some load is transferred to the wetland, and inevitably, there will be SOME compression.  This compression will need to be estimated in case it is excessive for your works, and you will need to check the environmental acceptability also.

2) Install a piling system to carry load from a conventional embankment.  The embankment load could be transferred to the piles using a very heavy gauge geogrid, such as a para-web.  Some compression will still occur, as some of the load is still transferred to the underlying wetland, and this compression will have to be assessed as in 1) above.  Removal of the piles on demolition of the temporary works may present a problem.

In order to design the solution - even to select sensible solution options - you will need to have some geotechnical information.  At the very least, you should know the thickness of the soft wetland material, and the nature and properties of the underlying material - down to a sensible founding level.

Perhaps you should question the constraints placed on you:  Are you REALLY prohibited from compressing the wetland AT ALL?  (If you are, you will just have to suspend your temporary works from piles!)  Do the temporary piles have to be removed at the end, or can they be left in place?

I have probably raised more questions than I have answered, but I hope these thoughts are useful to you.

Sean,
Yes, the embankment will have to be removed once the permanent bridge/approach is finished. Currently, a layer of sand is placed, then the temporary embankment. When all of this is removed, it is found that the bog/wetland has settled and probably no longer functions as a wetland. Any wetland that is destroyed must be replaced. The desired condition is that the load imparted by the embankment be reduced so much compaction will be minimal. The embankment will of course have to support traffic.

How about lightweight foamed concrete?
Thanks,
roger

I suspect that you would need to get the density of foamed concrete less than the density of water to avoid excessive loads.  I am not up to date with foamed concrete technology, and I don't know what densities can be achieved.  However, even if the foamed concrete FLOATS on the water, remember the Principle of Archimedes.  The foamed concrete will have a finite mass, and will displace the same mass of water.  In doing so, it will sink into the wetland, inevitably compressing it.  Because of the width of a highway, the depth of compression under the temporary embankment will not be insignificant.

Beware that, even if you expect the temporary embankment to be in place for only a very short time, in reality you may not always have control over this period, and it may extend considerably.  Do not RELY on the embankment being in place only for a short time.

Here's an idea which may be considered a little "off the wall", but may work.

First, establish whether you would be allowed to excavate material in the path of the road to a given depth.  This depth would be determined by design.  The material would be stored in similar conditions, ready to be returned to its original site on removal of the temporary works.

Next consider using anchored pontoons to carry the temporary road.  The pontoons would have to be deep.  Install anchors (perhaps at considerable depth) and tension cables between the anchors and the pontoons, causing the pontoons to submerge further into the water, but without compressing what is left of the "soil" wetland material underneath.  The tensioned cables would minimise the rocking of the pontoons as vehicles pass over them.

I expect that the traffic would have to be limited and strictly controlled.  (If this can't be done, I guess there is not much choice but to go for the expensive option.)

On completion, de-stress and remove the cables, remove the pontoons, and replace the stored wetland material.

Even if this is too far off the wall, it might trigger some better ideas for you or other contributors.

Good luck.

Sean,
I have been doing a little searching and have found a possible solution.  The expanded shale people here in NC, (Carolina Stalite), claim that their product has been used in environmental and hydroponic settings. It is about half the density of crushed stone and very porous.  The application would be as a mat.  The advertised advantage would be that the portion that sinks into the muck could be left and would function as part of the wetland. It is a porous inert material.  
Thanks for the thoughts,
Roger