Rachel Yost-Dubrow

While the processes that support our global reliance on fossil fuels dominate the public psyche, sand mining, or the large-scale extraction of sand from natural habitats, largely goes unnoticed. Despite its low profile, sand mining accounts for 85% of global mineral extractions, serving primarily as a raw material in concrete production or as a source for rare metals (Pearce, 2019).

Unsurprisingly, sand mining has a dramatic impact on surrounding ecosystems. Riverbeds are the most common target for extraction, where sand plays an essential role in determining the course and speed of the water as it flows downstream and supports the riverbanks on either side. These are all essential elements of river habitats and even minor changes may negatively impact native flora and fauna (Ojos Negros Research Group).

“(Penn State Special Collections, 1950)” “(WBEZ, 2013)”

Sand miners and the landscape they leave behind. Left image: Penn State Special Collections, 1950 / right image: WBEZ, 2013.

While river habitats aren’t a primary concern on Stradbroke Island, a sandy refuge about 30 kilometers off the coast of Brisbane, the consequences of sand mining are just as severe. Nearly 70 years of dredging has reshaped the island’s coastline and devastated native flora at the mining sites (Patel, 2020). Explore the interactive map below for more detailed geographic information about the island.

The impacts of sand mining are potentially mitigated through subsequent rehabilitation efforts wherein residual sand is returned to the site and covered with topsoil from the island. Seeds from native trees and shrubs are then sewn on top (Audet et al., 2013). However, local stakeholders disagree about how effective this process has been. The mining company argues that rehabilitated sites fully recover in about 10 years while many scientists are not convinced (Sweett, 2010).

“(Penn State Special Collections, 1950)” “(WBEZ, 2013)”

A view from Stradbroke island and some of its smallest residents. Left image: Wassell, 2011 / right image: Pink, 2009.

Given the uncertainty, researchers at the University of Queensland collected data on Stradbroke annually from 2015-2019 to better assess the effectiveness of the rehabilitation process. The abundance and diversity of ant populations on Stradbroke is used as a proxy for how effective the rehabilitation process has been given the established relationship between ant populations and broader biodiversity trends in a particular ecosystem (Zina et al., 2021).

Each year three sites were randomly selected at four former mines, rehabilitated in 1979, 1980, 1996 and 1997 respectively. Control sites that were never mined were also selected for comparison. Ten pit traps were assembled at each site and the abundance and diversity of captured ants was measured the following day. The analysis below offers a brief summary of the findings.

Analysis

There is a significant positive correlation between the number of ant species found in the pit traps and the total number of ants collected in each trap. This relationship is shown in the graph below and further supported by the p value for the ANOVA of log of ant abundance by ant diversity as shown in the table.

Analysis of Variance Table
  Df Sum Sq Mean Sq F value Pr(>F)
Diversity 1 96.06 96.06 83.97 4.25e-19
Residuals 789 902.5 1.144 NA NA

The relationship between ant abundance and diversity is unsurprising; when there are more total ants observed there is a greater chance for additional species to be observed. More interesting is the relationship between ant abundance and the year mined sites were rehabilitated.

Analysis of Variance Table
  Df Sum Sq Mean Sq F value Pr(>F)
Treatment 1 1.46 1.46 1.331 0.2489
Study Year 2 27.23 13.61 12.41 4.928e-06
Rehabilitation Year 1 8.593 8.593 7.836 0.005248
Treatment * Study Year 2 22.35 11.18 10.19 4.274e-05
Treatment * Rehabilitation Year 1 77.99 77.99 71.12 1.62e-16
Study Year * Rehabilitation Year 2 4.462 2.231 2.034 0.1315
Residuals 781 856.5 1.097 NA NA

The graph above shows that the rehabilitation process in 1997 significantly improved current ant populations at mined sites as compared to control sites. This was not the case for rehabilitation efforts undertaken in 1979 or 1980, and is unclear for 1996. The nuances of this trend are reflected in the p values of the ANOVA table which do not find a significant effect of mining and subsequent rehabilitation overall, but show a strong interaction between treatment (control versus mined) and rehabilitation year. The year the study was conducted also has a significant impact on ant abundance.

Further experimentation could improve the analysis done here by providing data from other rehabilitation years or mined sites that have not undergone rehabilitation. As this study is ongoing these suggestions are highly relevant and may inform future experimentation.

References