First, I have never worked with T. cacao specifically, so my response is a generalization.

If you have growth chambers capable of measuring net carbon exchange rates/photosynthetic rates, then you could determine the light curve(s) for each line/variety you are developing.  The compensation point and shape of the light curve will go a long way in telling you which lines are more/less shade tolerant.  Light curves are a good tool for rapid phenotype testing, but ultimately you need the proper equipment and it can be expensive.  Further, there may be differences between the juvenile plants (i.e. that fit in a chamber) and the mature plants in the field.

There are other methods that use spot photosynthetic and gas exchange measurements (e.g., LiCor photosynthesis system) that you could deploy in a field trial with different levels of shading applied (using neutral density filters).  Agronomic measurements (leaf area, dry weight, internode extension, etc) will be able to shed a fair amount of light (pun intended!) on the shade tolerance as well.

I know there are not a lot of specifics here, but I hope there is something useful.  Good luck in your testing!

Thomas

 Find a flat uniform field plot then develop shade cloth canopies that represent  0, 10, 20, 40 and 80% shade. Measure the growth rates at the different shade treatment rates by measuring height number of leaves and trunk diameter. Physiological measurement will also be useful. If you can bring your plants to fruit it would give a lot of information and the shade intensity variation would allow you to get a response curve. Make sure you have sufficient replication for your analysis of variance such a study might need a split plot design and you may want to consult with a statistician to help design the field arrange and determine the replicate size and number. Paul 

Under controlled conditions, we can study the response of different accessions by changing the length of photoperiod as an expression of varying intensity of shades...