Temporal Changes in the Spatial Structure of Coastal Dune Forest Canopies

Get Complete Project Material File(s) Now! »

Chapter 3 Assessing the Restoration of Coastal Dune Topography after Mining

Introduction

Natural forces shape beach sediments into coastal dunes, some of which are colonised by region-specific dune vegetation. Coastal dunes vary in height, aspect, and slope that result in microclimatic variability and ultimately spatially heterogeneous habitats. For example, variability in incident light and ambient temperature (Tateno & Takeda 2003; Bennie et al. 2008; da Silva et al. 2008), water retention (e.g. Pachepsky et al. 2001; Arbel et al. 2005), as well as nutrient and mineral accumulation in the soil (Chen et al. 1997; Oliviera-Filho et al. 1998; Tateno & Takeda 2003; da Silva et al. 2008). Spatial heterogeneity of these variables may contribute to species turnover and hence diversity (Larkin et al. 2006). Habitat variability related to topographic heterogeneity may therefore improve colonisation opportunities for species.
The restoration of topography should thus precede restoration efforts that aim at the recovery of biological diversity (Weiss & Murphy 1990; Lubke & Avis 1999; Palik et al. 2000; Larkin et al. 2006). However, I could find no example where the landscape engineering of the topographic profile of any ecosystem had been evaluated as a restoration goal. I posit that the topography of coastal dune ecosystems is core to their multi-functionality and therefore assess the restoration of the topographic profile after mining.
In this study area, mining destroys the coastal dune forest vegetation and the topographic profile of dunes (van Aarde et al. 1996c). Reshaping the sand tailings after mining forms part of the restoration programme and the resultant topographic profile had not been investigated until now. I used historical and recent remote sensing data to assess how closely the restored dunes match the topographic profile of the dunes before mining. Although the mining company does make efforts to rebuild the dunes, I expected the topographic profile (as characterized by the elevation, aspect and gradient of dune slopes) of post-mining dunes would be dissimilar from that of their pre-mining counterparts. Furthermore, fewer dunes would characterize the post-mining landscape and it was therefore expected that topographic heterogeneity would be reduced following mining and rehabilitation.

Methods

Study area

The study area consists of coastal sand dunes between 28°46’ and 28°34’ south. These parabolic-shaped dunes comprise porous, leached aeolian sand deposits left by a regressing Indian Ocean during the end of the last glacial maximum (Tinley 1985). These dunes run parallel to the shoreline and range in height between the Umlalazi River (80 m), southwest of the study area and the Umfolozi River (188 m) (Weisser & Marques 1979).
During mining, the dunes are collapsed ahead of the dredging pond where heavy minerals are extracted. After mining, sand tailings are stacked and bulldozed to resemble pre-mining topography, after which they are covered with a layer of topsoil salvaged from the mining face (van Aarde et al. 1996c).

Topographic data

Dune topography may be quantified or categorized at specific geographic locations according to the aspect, elevation, and gradient of slopes (Table 1). I used topographic layers from geographical surveys done during 1971 (pre-mining) and data products from a Light Detection and Ranging (LIDAR) mission conducted in September 2010 (post-mining). Using inverse-distance-weighting interpolation in ArcMap Desktop 9.3.1 (ESRI Inc., Redlands, California) (see Woolard and Colby 2002), I generated digital elevation models (DEMs) of matching extent and resolution (cell size) for the pre-mining (Figure 3-1a) and post-mining landscapes (Figure 3-1b). I used these to generate pre- and post-mining surface models of aspect and gradient using three-dimensional analyst tools in ArcGIS. I classified these models based on eight cardinal directions (aspect), seven elevation categories, and five gradient categories (see Table 3-1).

Statistical analyses

To calculate topographic heterogeneity I used the Zonal Statistics tool of ArcGIS to calculate the area covered by each elevation category and then used Shannon’s Diversity Index to estimate pre- and post-mining topographic heterogeneity for the study area (see Nichols et al. 1998). I also made this comparison for individual regenerating stands. I calculated diversity as the ∑ pilogpi for each elevation class, where pi is the proportion of the total area of the stand covered by elevation class i. A low diversity index indicates that a stand comprised little variability in elevation (low topographic heterogeneity), whilst a stand of similar size with more elevation classes will yield a high diversity index. A negative value for topographic heterogeneity (TH´) therefore suggests a reduction in topographic heterogeneity following mining.
To identify areas of change between pre- and post-mining dune morphology, I conducted an image differencing exercise, subtracting values of the post-mining DEM from that of the pre-mining (Figure 3-1c). I used GIS overlay procedures to sample dune morphological variables (aspect, elevation and gradient) at 161 geographically random locations (>100 m apart) from the pre- and post-mining datasets (Figure 3-1d). I calculated the frequency distributions of these random locations based on the eight aspect categories (cardinal directions), while the categories for elevation were widened to 25-m intervals (0-25, 26-50, 51-75, and >75 m.a.s.l.), and those for gradient to four categories (0-5, 6-10, 11-15, and > 15°) to avoid frequencies of less than five. I assessed differences in these pre- and post-mining frequency distributions using contingency table analyses for each feature of dune morphology. All statistical analyses were conducted in STATISTICA 10 (© 2011, Statsoft Inc., Tulsa, Oklahoma).

READ  INSTRUCTIONAL LEADERSHIP IN A TEACHING AND LEARNING CONTEXT

Results

Restored topographic heterogeneity was lower than pre-mining heterogeneity for the study site, but on an age-specific basis, five stands were less heterogeneous and five were more heterogeneous (Figure 3-2). However, cases where the topographic heterogeneity of stands was reduced after mining and rehabilitation were slightly more pronounced (max = -2.9) than those where heterogeneity had increased (max = 2.3).
The morphology of dunes after restoration differed from that before mining and, as indicated by the frequency distributions of random locations recorded within categories of aspect, elevation and gradient before and after mining (contingency tables: χ2 = 45.16, df = 7, p < 0.0001, χ2 = 84.12, df = 3, p < 0.0001, and χ2 = 24.69, df = 3, p < 0.0001, respectively) (Figure 3-3). After mining, more of the random locations were recorded on southeast- and north-facing slopes than before mining, though fewer fell on slopes facing all other aspects. There were also fewer random locations on elevations below 25 m, but more of the random locations were recorded on elevations of greater than 51 m after than before mining. Fewer random locations were recorded on slopes steeper than 15°, but there were more on gradients of between 0 and 10° after mining, than before mining.

Discussion

Topography may have a fundamental role to play in restoration given its well-documented influence on ecological processes and therefore community composition (e.g. Weiss & Murphy 1990; Tateno & Takeda 2003; Larkin et al. 2006; Laurance et al. 2010). The association of the indigenous forests of this area with coastal dunes suggests that the topographic profile played a role in the establishment of these forests by sheltering them from prevailing winds and fires, as has been documented for forests elsewhere (see Geldenhuys 1994). Restoring the topographic profile may be an important step in the rehabilitation programme, as it essentially erases the historical context of these dunes.
In the study and in line with the expectations, there was a reduction in topographic heterogeneity across the study site and restored dunes were in places taller and their slopes somewhat gentler than prior to mining. This suggests that there were fewer dunes in the same area than before mining began. There are several logistical constraints to building dunes with mine tailings. For example, building more dunes in the same area will require steeper slopes; however, this increases the risk of dune slumping, while sand movement and increased run-off will hinder plant growth. These constraints could explain the post-mining reduction in topographic heterogeneity.
The morphology of sand dunes is related to topographic heterogeneity, and is known to influence the spatial heterogeneity of habitat conditions as a result of the modulation of wind, water, light and soil conditions in much the same way as for other topographically variable ecosystems (Oosting & Billings 1942; Martinez et al. 2001). For example, the aspect of slopes in relation to the prevailing wind direction or the sun, can influence incident light and ambient temperatures (Tateno & Takeda 2003; Bennie et al. 2008; da Silva et al. 2008). Similarly, valleys may provide relatively moist micro-environments that are protected from the wind and sunlight relative to dune crests (Martinez et al. 2001). The topographic profile of restored dunes differed from that of the pre-mining landscape. Exceptions included large increases in the number of locations falling on north- and southeast-facing dunes and areas over 50 m.a.s.l., as well as decreases in relatively steep slopes (>15°) and low-lying areas (<25 m.a.s.l.). However, it is important to note that these comparisons were relatively strict as they are based on the same set of geographic locations that sampled dune morphological variables before and after mining and restoration. These changes therefore represent either a change in dune morphology, or the ‘shift’ of dunes across the landscape, or a combination of the two scenarios. The topographic profile has been shown elsewhere to influence biological diversity (Nichols et al. 1998; Martinez et al. 2001). Nevertheless, the implications of this increase in relatively warm and steep slopes requires further investigation by assessing the responses of biological communities, if any (see Chapter 5).

Chapter 1 General Introduction
Historical context
Landscape-level factors
Site-level factors
Coastal Dune Habitats of the Northeast Coast of South Africa
Study area
Rehabilitation program
Regeneration through ecological succession
Conclusion
Chapter 2 Temporal Changes in the Spatial Structure of Coastal Dune Forest Canopies
Introduction
Methods
Study area
Fragmentation
Age-related trends in the spatial structure of tree canopies
Results
Discussion
Chapter 3 Assessing the Restoration of Coastal Dune Topography after Mining
Introduction
Methods
Results
Discussion
Chapter 4 The Response of Millipede, Dung Beetle and Bird Communities to the Patch Substrate, Composition and Spatial Structure of Regenerating Coastal Dune Forest Canopies 
Introduction
Methods
Results
Discussion
Chapter 5 | Coastal Dune Topography as a determinant of Abiotic Conditions and Biological Community Restoration in Northern KwaZulu-Natal, South Africa
Introduction
Methods
Results
Discussion
Chapter 6 | Synthesis 
The historical context of landscape- and site-level factors
The response of biological communities to site- and landscape-level factors
The importance of age
Into the future: Reality check
References 
Appendix
GET THE COMPLETE PROJECT

Related Posts