Is there a middle way between permanent plots and chronosequences?
Although permanent plots have proven critical to studies of vegetation dynamics, their logistic limitations have led to the wide use of chronosequences as an alternative. Here, we test whether or not an approach combining permanent plots and chronosequences could be used successfully to accurately predict the vegetation changes that one would see in permanent plots in the same area. We used plot data from five pastures in the Luquillo Mountains of Puerto Rico, USA, and found that (i) for species composition and abundance, the five pastures did not form a single trajectory or show any convergence, (ii) for successional rate, there was a general decrease with time since abandonment for most pastures, but that decrease was not monotonic and one pasture of the five had no decrease at all, and (iii) total species richness and total plant cover showed pastures that lined up well by age, forming almost a single trajectory with little variation. We conclude that the utility of using chronosequences either alone or with permanent plots depends largely on the parameter under study with broad structural parameters, such as total species richness and total plant cover, performing best.