The slowly changing variables on pastures

Before trying to measure resilience for a pasture ecosystem, a natural question is what indicators should be chosen.  The change of the indicators' value will lead to the flipping of the system's state, i.e. exceeding the boundaries of domains of attraction.

Slowly changing variables are not necessarily indicators that can be directly measured.  One reason, is that as its name indicates, they change slowly, in a short time period one can even consider them as fixed value.  Some slowly changing variables are too subtle, or technically hard to measure.

However, to determine the indicator one must address the slowly changing variables.  So, what are the slowly changing variables for pasture lands?  It is a hard question without a conceptual model at hand.  And a proper conceptual model is almost  impossible without being in the field and observe.  Anyway, I will try to come up with some ideas about the slowly changing variables while I am in my armchair.

In an Australian rangeland ecosystem, one of the major driver is the ability of plants to draw water from the soil.  Therefore the texture and moisture of soil will be a slow factor (Anderies et al., 2002).

The grass-shrub ratio in certain systems can also be an underlying slowly changing variable (Walker, 1993). 

What are the others?  Purely guestimate, nutrient in such a poor-nutrient system may play some intriguing role.  It will influence the composition of plants, and very likely the grass biomass.  Both are important slowly changing variables in Australian rangeland ecosystem.