Conserving plant biodiversity in a fragmented landscape
April 9, 2019
The Convention on Biological Diversity considers the creation of protected areas (PAs) as a key strategy in conserving biodiversity. Higher species diversity in these areas may be a result of favourable management practices, such as minimizing disturbance and removal of invasive species, or a PA may have been selected in a location with a relatively high level of existing biodiversity.
Unprotected areas such as private lands may also harbour a diversity of species however, private lands and PAs have rarely been compared to determine if one has greater diversity of species than the other. Since gaining a clearer understanding of possible links between biodiversity and land ownership could help prioritize conservation activities, McCune et al. (2017) set out to test for differences by comparing the diversity of plant species in forested private lands versus forested PAs. In their study, they considered three indicators relevant to biodiversity: native plant species diversity, relative abundance of exotic plants, and presence or absence of at least one species of conservation concern.
The researchers chose southern Ontario as their study region - the most populated area of Canada as well as one of the country’s biodiversity hotspots. As part of an earlier study, McCune (2016) recorded and estimated the abundance of every species of vascular plant growing in 156 one-hectare forest plots in 2014 and 2015. Data from those surveys, which included 110 privately owned and 46 PA plots, were used in the present study.
For each plot, McCune et al. (2017) gathered data on factors that can affect plant species diversity, these included:
- Total forested area within a 500 m radius of each plot (currently and historically in 1954)
- Forest age
- Forest patch area
- Distance to the nearest forest edge
- Topographic complexity
- Latitude of each plot
- Calendar day of each survey
The researchers incorporated most of these variables as well as land ownership type (private lands or PA) into three explanatory models, one for each of the indicators of biodiversity. Forest patch area, distance to the nearest forest edge, and current total forested area were not included in the models since they were highly correlated with other variables and would therefore introduce unnecessary, and potentially problematic, redundancies.
Contrary to expectations, McCune et al. (2017) found that, with all other factors being equal, privately owned lands had higher diversity of native species and were more likely to contain a species of conservation concern compared to PAs. One possible explanation is that relatively high recreational use of PAs in southern Ontario could serve as a significant disturbance leading to lower biodiversity.
They also found that regardless of land ownership, plots with more forest on the surrounding landscape showed greater diversity of native species, lower abundance of exotic species, and a higher chance of housing a species of conservation concern. Higher forest connectivity or greater distance from disturbances such as agriculture and communities could be driving these responses. Finally, plots consisting of young forests were shown to have a greater abundance of exotic species and fewer species of conservation concern than plots of old forests. These young forests were regenerating on former agricultural lands which can have a negative impact on biodiversity.
This study highlights the importance of privately-owned lands, as well as PAs, in conserving plant biodiversity in the highly populated and fragmented region of southern Ontario. Furthermore, McCune et al. (2017) suggest that evaluating the degree to which PAs are effective in housing biodiversity must take landscape context and forest age into account.
Read the full article (journal subscription required):
McCune JL, Van Natto A, MacDougall AS. 2017. The efficacy of protected areas and private land for plant conservation in a fragmented landscape. Landscape Ecology. 32:871-882. doi 10.1007/s10980-017-0491-1
McCune. 2016. Species distribution models predict rare species occurrences despite significant effects of landscape context. Journal of Applied Ecology. doi:10.1111/1365-2664.12702