Identifying alternative stable states in a tropical landscape using remote sensing

Although natural regeneration is often an efficient process for recovering degraded areas, the emergence of invasive species can complicate this process and lead the ecosystem to an alternative stable state (ASS) that disrupts ecosystem functioning. Identifying this condition is a key step in the design of restoration strategies. In this sense, we used a spatially explicit approach using the landscape metric area and the analysis of land-use and land-cover change to identify the dynamics of areas dominated by P. aquilinum , an invasive species well known for producing ASS. In our study site, a Mexican semi-evergreen tropical forest and seasonal lowland flooded forest, P. aquilinum has been in ASS for decades. We analyzed P. aquilinum dynamics in a period of 28 years (1989–2017) using a supervised classification of Landsat imagery. We found that, at the landscape scale, agriculture and livestock increased 224%, while semi-evergreen tropical forest and P. aquilinum cover showed losses of 13% and 72%, respectively. At the patch scale, we also found a statistically significant decrease in P. aquilinum patch area between 1989 and 1997 and 2017, but not between 1997 and 2017. Our results show that the areas dominated by P. aquilinum only change in time mainly through management toward productive activities, while patches not managed retain their area. Our study highlights the importance of spatial scale analysis for the understanding of landscape ecosystem dynamics for guiding decision-making and further research.