Power density (in watts per square meter [area]) is a product of electric field and magnetic field (both in units per meter [linear]).

This breaks down if you are not in the far field.

Watch out for the difference between 'directivity' and 'gain'.  The difference is loss.

Another answer might be to use NEC software.

Personally, I dislike the idea of "capture area".  Although it makes sense for reflectors, it doesn't seem to have any conceptual basis for high gain antennas with small frontal area (such as a long Yagi).

In addition to VE1BLL, you will need to account for the ground effects (mutltipath reflections, terrain obstructions) as well as fade margin.   Most manufacturers have a recommended fade margins for their receivers.

Look at:
http://www.proxim.com/support/all/stratum/technotes/tn2001-08-13a.html

http://rf.rfglobalnet.com/software_modeling/categories/14.htm

Analyzing the antenna pattern of a low frequency antenna is extremely difficult in a real life setup (i.e. where you place antennas outdoors next to other objects that you can't control). You may just want to calculate peak gains of the two antennas in the chain, find the loss and assume a -15 dB drop out from that maximum.
Ham radio guys (and women) can help you get a feel of transmission problems if you are working down low in frequency between 1 and 100 Mhz. They have alot of great experience.
kch