Habitat Restoration for Endemic Lizards in an Oilfield in Payunia, Argentina

Habitat Restoration for Endemic Lizards in an Oilfield in Payunia, Argentina

Leandro M. Alvarez (corresponding author: Instituto Argentino de Nivología, Glaciología y Ciencias Ambientales [IANIGLA], Universidad Nacional de Cuyo, Mendoza, Argentina, and Asociación para la Conservación Biológica Argentina [BIOTA], Bosques Telteca 2237 (5509) Luján de Cuyo, Mendoza, Argentina, lalvarez@mendoza-conicet. gob.ar) and Bárbara Guida-Johnson (Instituto Argentino de Investigaciones de las Zonas Áridas [IADIZA], Universidad Nacional de Cuyo, Gobierno de Mendoza, CONICET, Mendoza, Argentina, and Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Cuyo, Mendoza, Argentina).

Fossil fuel extraction is an anthropogenic activity that completely modifies habitats at the local scale. Indirect impacts include human expansion into wild areas, introduction of invasive species and pathogens, soil erosion, water pollution, and illegal hunting (Butt et al. 2013). These environmental hazards become more problematic for biodiversity when exploitation takes place in areas with high conservation value. Oil and gas extraction sometimes occur within protected areas in the Neotropics (Suárez et al. 2009), and this is the case for two-thirds of the protected areas under provincial jurisdiction in Neuquén, Argentina (Fiori and Zalba 2003). One example of competition for land uses is the Auca Mahuida Nature Reserve, located in the Payunia region, Neuquén. In Payunia, the landscape, glaciation cycles, and pronounced climate changes resulted in the occurrence of a hot spot for reptile diversification (Martinez et al. 2011). The Auca Mahuida Nature Reserve is characterized by a high richness of lizards and the existence of several endemic taxa (Corbalán and Debandi 2008, Avila et al. 2015) with different degrees of habitat specialization and conservation concern (Abdala et al. 2012).

Habitat restoration for reptiles has been evaluated in different types of degraded ecosystems, and these studies highlighted the importance of microhabitat characteristics (Kanowski et al. 2006, Bateman et al. 2008, Perry et al. 2009) as well as connectivity with source populations (Kanowski et al. 2006) to ensure colonization success. When restored patches occur within a matrix where the reference ecosystem prevails, as in the case of oilfields, the resulting matrix is permeable, and it contains the source populations necessary for colonization. Therefore, microhabitat characteristics are the primary determinant of reptile assemblages in restoration sites, and landscape attributes only have a minimal effect (Triska et al. 2016). There are examples of restoration activities in the Payunia region, which are focused on the recovery of native vegetation and topography (Pérez et al. 2013, Martínez Carretero and Dalmasso 2015). However, to our knowledge, there are no previous studies related to habitat restoration that target any lizard species or rock environments. Our objective was to assess the effectiveness of constructed rock shelters to serve as refuges for the endemic lizard assemblage in an oilfield in Payunia.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Lizards observed at restoration sites: A) adult individual of L. crandalli, B) adult individual of P. sitesi, C) adult female of P. roigorum and D) adult male of L. sitesi.

Payunia is a large volcanic field located in north-eastern Neuquén, Argentina, between 33°30' and 38° S. This study was conducted in the Auca Mahuida Nature Reserve, a protected area of 77,020 ha located on the southern border of the Payunia region (Llambías et al. 2010). Auca Mahuida landscape is composed of rocky outcrops and hills, shrublands, and grasslands with distinctive elements from the Payunia phytogeographic province (Martínez Carretero 2004). Lizards taxa endemic in the Payunia region are Liolaemus austromendocinus, Diplolaemus leopardinus (Victoriano et al. 2010) and Phymaturus roigorum (Figure 1) (Lobo and Abdala 2007). The endemic taxa in the Auca Mahuida range are Liolaemus cyaneinotatus, which is mostly recorded on bushes and rock edges (Martinez et al. 2011), Liolaemus sitesi, associated with grasslands and found underneath stones, Liolaemus crandalli (Avila et al. 2015) and Phymaturus sitesi (Avila et al. 2011) (Figure 1). The rock-dwelling lizard assemblage includes L. austromendocinus, L. crandalli, P. roigorum, P. sitesi, and D. leopardinus. P. sitesi and P. roigorum are large-bodied viviparous and herbivorous species (Avila et al. 2011) that occupy outcrops of an appropriate size and thin crevices generated by freeze-thaw activity in boulder rocks (Alvarez and Gizzi 2016). These lizards have a prolonged reproductive cycle characterized by low fecundity, reduced reproductive frequency, and limited annual reproductive output (Ibargüengoytía 2004, Boretto et al. 2007, Boretto and Ibargüengoytía 2009). Diplolaemus leopardinus is a large predator with a restricted distribution of fewer than 20,000 km² (Victoriano et al. 2010). All these taxa have a vulnerable conservation status in Argentina because of their specific micro-habitat requirements, reproductive cycles, and the habitat loss and degradation caused by oil extraction activities (Abdala et al. 2012).

We selected eight sites for the experimental design, to which we added a total of 77 rock shelters. We report data from 130 surveys over three years (2015–2017). In 2015 we selected five impacted sites (characterized by vegetation removal and topsoil loss) associated with abandoned oil extraction platforms and roads. We visited each site four consecutive days, and we spent two hours in the degraded patches to corroborate lizard absence. In late 2015, we manually built rock shelters, stacking disperse fragments of basalt and andesite from the surroundings, selecting only residual blocks. Each shelter was made from 20 to 25 rocks (Figure 2). The number of shelters per site varied according to its surface (e.g., 2, 3, 5, 10, 15 and 20). In late 2016, we added three sites following the same criteria. Each site was characterized according to its size (m²), perimeter (m) and its border exposed to natural habitats (m) (i.e., the length of site borders adjacent to undisturbed environments) (Figure 2). In January and February of 2016 and 2017 (first and second season, respectively), surveys were conducted weekly. We recorded all individuals seen at each of the shelters, and we identified the species. All observations were made during the optimal period for lizard detection (i.e., 11:00 to 13:00), only on sunny days and avoiding sampling under windy conditions, adding up to 20 days in total, by season.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

General view of A) rock shelters within a site and B) a rock shelter.

To assess the effectiveness of the rock shelters, we estimated the average occupancy of all lizard species at each site and each shelter per season, measured as the average percentage of rock shelters used by lizards on each survey. We also estimated species richness and calculated the relative abundance per site, according to Sutherland (2006) to evaluate the frequency of each species that occupied the refuges. To answer whether the number of constructed rock shelters affects lizard abundance, we ran Generalized Linear Mixed Models (GLMM) in R software, using the lme4 statistical package to account for repeated measures (Bates et al. 2015). Season and survey were treated as random factors (with survey nested in season), and the number of shelters, as well as the physical characteristics of each site, were treated as fixed factors. Then we performed a stepwise backward analysis (starting from the full model and eliminating non-significant variables), to identify the best model to explain data variability of lizard abundance. For model selection, we considered the lowest value of residual deviance and Akaike’s information criterion (AIC). Our results suggest that building rock shelters at disturbed sites increases lizard presence and abundance, probably by improving habitat conditions. All restored sites were used by seven lizard species in total, including Liolaemidae and Leiosauiridae. The rock shelter average occupancy was 31.35 % (± 11.41) and the species richness varied between sites (ranging from one to five species per site; Table 1). L. crandalli was the most frequently recorded species. This species exhibits plasticity in its habitat preferences, and its diet is mainly composed of insects (Avila et al. 2015). Phymaturus sitesi was observed at all sites except for one, whereas P. roigorum was seen at two sites, characterized by smaller sizes of shelters, and shorter distances between the shelters and the natural vegetation on the edges. Consequently, to ensure availability of resources for individuals of the Phymaturus genus, other measures must be included, such as revegetation of preferred shrubs like Mulinum spinosum. Diplolaemus leopardinus was recorded in two surveys at different sites, and isolated individuals of L. cyaneinotatus and L. sitesi were observed hiding in the lower rocks of shelters (Tables 2 and 3).

The selected model, after rescaling the variables, had the lowest values of residual deviance (549.8, null model = 731.1) and of AIC (559.8, null model = 737.1). Lizards abundance were mostly explained by the number of rock shelters per site (p < 0.0001) and the border exposed to natural habitats (p = 0.009) (Table 4). Whereas a higher number of built shelters produce greater heterogeneity and resource availability, longer edges adjacent to undisturbed environments could be associated with a higher probability of occupancy.

The construction of rock shelters can be implemented in combination with other measures that focus on physical conditions and vegetation availability to restore abandoned infrastructure in oilfields in and around nature reserves. The use of natural materials available in the surroundings of sites represents an additional advantage since it minimizes restoration costs. This factor, added to the simplicity of the technique and its potential impact on endemic species with high conservation value, may promote the implementation of this measure by oil companies in the region.