I may be misreading the graph.....what event caused the drop in woodland circa 1995?  It appears like 80% of the worlds forest went to other use.  Am I misreading this?  I know there has been rain forest depletion but the US is covering more acreage every year with forest.  Whats up?
The FAO data series for forest and "all other land" end in 1994.  So we don't know what happened after that.  (But it presumably wasn't too radical, or it would have had to show up in the agricultural land.
One problem with subtracting the Antarctic landmass from the global land area to get 13.4 billion ha. is that the resulting total includes a lot of mountaintops and deserts with insignificant productive value. Down below, Jason Bradford points out that ecological footprint analysis is another way of approaching this issue. The analysis at http://www.footprintnetwork.org includes a methodology paper saying:
Globally we identify 11.2 billion hectares of distinct bioproductive areas--cropland, forest, pasture, fisheries, and built-up land--that provide economically useful concentrations of renewable resources. These 11.2 billion hectares cover a little under one quarter of the planet and include 2.3 billion hectares of marine and inland fisheries and 8.8 billion hectares of land. The land area is comprised of 1.5 billion hectares of cropland, 3.5 billion hectares of grazing land, 3.6 billion hectares of forest, and an additional 0.2 billion hectares of built-up land assumed to occupy potential cropland  

They estimate that these 11.2 billion ha capture at least 80-90% of the total usable annual generation of biomass.

The site includes a couple of figures showing the decline over time in biological capacity, due mostly to resource depletion and ecosystem degradation, and in the area occupied for agricultural crops.  While data supporting the early part of the analyses are less reliable, the overall trends are probably more realistic than the ones above, which are based on incomplete data from the FAO.

Good point. Paving some of the nastier / less productive desert land in the foot hills of North Scottsdale is probably a lot less damaging that paving an equivalent acrage of prime agricultural land. Location, locations and location aren't everything ... there is also top soil, water and growing season length.