other systems?
6.4.1
Conservation programs after agriculture abandonment
According to the general theory of community resilience in relation to levels of invasion/disturbance, the relative stability of a species communitysensuSutherland(1974) is characterized by its resistance (i.e. the degree to which it resists to changes due to disturbance) and its resilience (i.e. the degree to which the initial community characteristics are restored after displacement). Habitat modification, as a consequence of disturbance, is an important factor in the invasion process. Disturbance in fact promotes invasion (Crawley, 1987; Hobbs, 1989; Mack and D’Antonio, 1998; Godfree et al., 2004). For instance, the major disturbance on Banks Peninsula, as in other areas of New Zealand (McEwen, 1987; Rose et al.,
1998;Williams and West,2000;Duncanet al.,2001;McWethyet al.,2009;Wilson and Meurk,2011), can be found in the frequent fires with forest areas that were converted to grasslands for grazing by sheep and cattle, where there is a high intensity of oversowing of alien pasture species especiallyAgrostis capillaris,Rytidospermaspp.,
Anthoxanthum odoratum,Lolium perenneandPoa pratensis. AsParkeret al.(2010) highlighted, areas converted from an agricultural to a forested state contain high numbers of alien species due to soil compaction. In reverted areas, the high levels of land-use activities (e.g. oversowing, fertiliser addition and livestock grazing) result in a high chance of dealing with high levels of invasion (Wilson and Meurk,2011). Land managers will need to bear this in mind when considering future conservation or restoration programs on Banks Peninsula but also elsewhere.
6.4.2
Environmental drivers in heterogeneous landscape and
applicability to other landscapes
The heterogeneous landscape and the natural history of the Banks Peninsula have presented me with a unique site for the study of plant invasions and for disentangling
the gradients of human-related disturbance and climate-topography. Many of the explanatory variables that have been used in my study are similar to previous studies where climate, environmental (e.g. topography) and potential human-related disturbance (e.g. land-use and propagule pressure) explained the distribution of alien species [e.g. Rouget and Richardson (2003); Pinoet al.(2005)]and native species (Deutschewitz et al., 2003; Dark, 2004) in less heterogeneous landscape (Boughtonet al.,2011) or at larger plot size (Mariniet al.,2009) or at smaller extent [e.g. <12 km2; Parkeret al.(2010);Souza et al.(2011)]. Studies of native and
alien plant species composition patterns in other modified landscapes have shown that the native and alien species communities differ in their spatial and ecological distributions together with the environmental factors with which they are correlated (Wilsonet al.,1989;Wilson,1989;Oneal and Rotenberry,2008;Brown and Boutin,
2009;Otýpkováet al.,2011). Two studies (Wilson,1989;Wilsonet al.,1989) found that in other semi-natural to high managed pasture areas of the South Island in New Zealand, native and alien species communities differed spatially and ecologically and that these were differently related to environmental factors (e.g. elevation, soil fertility and water).
In brief and where the information is available, we need to take into account human-related factors or proxy measures (e.g. land-use change, fire frequency) and environmental gradients when studying drivers of plant invasions at local and regional landscape-scale, especially where the landscape is highly disturbed.
6.4.3
NARR scale dependence and factors related
Following my findings, the scale dependence of NARR is undeniable. Even when the plot size is small (<100 m2), as in the case of Banks Peninsula, and where the
environments can be assumed to be relatively homogenous, habitat heterogeneity [i.e. diversity of habitats;sensuFranklinet al.(2002)] may still be encountered and it may still vary to the extent that NARR on Banks Peninsula (or studies worldwide that used the same plot size) significantly shifts from positive to negative and vice versa. Therefore, this high variability and the spatial dependence of NARR makes it difficult to disentangle the main drivers of NARR, not just on Banks Peninsula, but especially where the plot size used is large enough that the influence of other factors (e.g. climate and geology but also disturbance rates and propagule pressure) may be significant. This interplay of different factors that are involved, depending on which spatial grain/plot size is used, poses a considerable ecological challenge. Regarding
this, a question can be raised: is there an appropriate spatial grain/plot size to study an ecosystem for biological conservation purpose and ecological restoration? The scale dependence of ecological patterns and processes makes observations at one plot size often not applicable to other plot sizes (Wiens, 1989; Levin, 1992). In particular, findings from studies conducted at small spatial grain/plot size may not be extrapolated to large spatial grain/plot size (Rastetteret al.,1992). As discussed in Chapter 3, Chapter 4 and Chapter 5, the NARR changes sign and magnitude according to different spatial grain/plot size and species communities. A negative NARR at small spatial grain/plot size and in native dominated communities is, in fact, opposed to a positive NARR at large spatial grain/plot size and in alien dominated communities. In the conservation biology context, this scale dependence poses a considerable challenge, given that there is a considerable number of detailed studies that involve small plot size (Gaston and Blackburn,2000). Before embarking any decision, policy and decision makers need to bear in mind this scale dependence of NARR. Moreover, only few studies have included land-use and -management gradients (Peltzer and MacLeod,2013). In my case study, I have shown that land-use and -management gradients are a significant factor in explaining NARR and therefore including a direct measure of human-related factors (e.g. land-use changes) may be important in arriving at a more precise measure of invasibility "to prioritize areas for invasive species control and prevention, for ecological restoration efforts that seek to create invasion resistant communities, and for understanding the processes that govern community assembly across a variety of ecosystems" (Fridley,2011).