Urban Sprawl Supports Rapid Invasions By Non-Native Plant Species
Rather than using a fixed time point to determine the current state of the invasion, as it is often done in similar studies, Gregorio attempted to map and follow a decadal time span of the invasion. Towards that, he obtained six LANDSAT scenes of the study area (1983, 1987, 1992, 1997, 2001, and 2006). Instead of mapping each scene separately, he merged them into a multi-temporal scene which he classified using Support Vector Machines, a highly sophisticated image classification technique. Each class represented a specific invasion period (e.g. 'expansion 1983-1987'), which resulted in a map that portrays the entire invasion process through time. Field data which included tree coring confirmed the accuracy of his work.
Gregorio found out that the rate of the species expansion was very fast, even comparable to the most effective invasive plant species known today. Starting from just 50 hectares in 1983, with most of them located in small stands located inside urban areas, Glossy privet now covers 2500 hectares, which are 20% of the forests in the study area. Invasion rates were slow in the 1980's (36.2 ha per year), but increased rapidly in later study period (152 ha per year). The expansion was not homogeneous in space and time. Early expansions occurred along a river valley and on forested slopes near the source populations. Later expansions occurred in other parts of the study area, as well as near the initial invaded forest stands.
To assess how invasion patterns are related to human settlements, Gregorio applied two techniques from distinct fields. The first, analysis of preference, is widely used in wildlife management, and tests whether the proportion of Glossy privet stands in certain buffer distances around urban areas is higher than its proportion in the entire study area. The second, neutral model simulations, is used in landscape ecology, and tests whether a certain factor structures the landscape. The methods is based on generation of many artificial (neutral) landscapes, which in this case yields an empirical distribution of the spatial association between Glossy privet and urban areas (i.e. how many times Glossy privet pixels appear adjacent to urban pixels) that is compared to the spatial association value obtained from the real landscape.
Both techniques revealed a clear and significat relation between urban areas and Glossy privet occurrence and expansion. The distribution of Glossy privet proportions varied with buffer distances, and between years. In 1983, a large proportion of Glossy privet (82%) was within 600 m of urban areas, whereas in 2006, this proportion dropped to 52%, but a substantial proportion of the species was found more than 600 m beyond the urban boundaries. Between 2001 and 2006, Glossy privet tended to expand further outwards of the urban areas.
Urban areas are a key factor in the distribution of Glossy privet, and Gregorio suggests that they facilitate the invasion in two ways: being a source of propagules for invading adjacent areas, and proving disturbed areas that are more easily invaded. Here, the original seed sources were located inside urban areas, which served as the starting points of the invasion. The invasion process was further facilitated by several other human activities, such as forest clearing and soil disturbance which are known to support invasions by supplying competitive advantage to invaders over the native species. Since Glossy privet is mainly dispersed by birds, the rapid spread of this species in the region is a result of the complex interactions between human activities, landscape pattern, and species behavior and distribution. These finding have important management implications, since urban growth is the study area is rampant and expected to continue, in addition to ongoing logging and other disturbances. Conservation efforts should therefore attempt to prevent urban development near forests of high conservation value. Regulations and environmental education can be used to limit the use of Glossy privet as an ornamental plant, especially in areas of new urban development.