Abstract
Rapid environmental change and human activity has dramatically
facilitated the spread of invasive species, expanding their impacts
beyond the original recipient communities. Predicting the potential
spread of invasive species and ways to stop it remain challenging, as
several abiotic, biotic, and management factors may alter outcomes.
Among the most problematic invasive species globally, Burmese pythons
(Python bivittatus) have established throughout much of south Florida
(USA) and pose substantial ecological, economic, and societal threats to
the region. To understand the invasion process, we use a new
spatiotemporal modeling framework, the spatial absorbing Markov chain
(SAMC), to model future spread of pythons while accounting for propagule
pressure and mortality risk from three hypothesized sources: (i) cold
exposure, (ii) vehicle strikes on roads, and (iii) removal management
programs. To parameterize this model, we integrated empirical and
model-derived data of python occurrence, movement, and behavior, and
physiology using coupled correlative-mechanistic models. In a simulated
invasion scenario, we found that removal management programs may have
the greatest potential for limiting future spread through long-term
mortality—accounting for 93.6% of all expected mortality and
exceeding both cold exposure (<0.1%) and road mortality
(6.3%). Furthermore, we demonstrate that circuit theory, a model which
does not account for demographic processes, likely overestimates spatial
patterns of connectivity. By examining invasions in a spatiotemporal
framework, SAMC can provide novel information—including spatial
patterns of survival, time-specific movements, and evaluation of
different types of removal management strategies—to guide the
management of invasive species.