Effects of large scale forest practices on wild rabbit

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

the case of Acacia spp in Quiaios National Forest, Portugal


Biodiversity monitoring actions are one of the most important tools for sustainable forest management. However, given the difficulty in implementing large-scale programs, the use of a restrict number of indicator species has been suggested. Wild rabbit seems to be a good choice to be used as an indicator species in the majority of Iberian forest ecosystems. The present study focuses on using wild rabbit data as a sustainability indicator of forestry management in Quiaios National Forest, a coastal forest included in the Natura 2000 network. Particular attention was given to the impact of forestry methods related with the control of invasive species (specially Acacia longifolia) and fuel reduction on the abundance and distribution of wild rabbits. Latrine count methods were used to monitor rabbit populations during six years (from summer 2002 until winter 2007/2008) in three scrubland zones. Two out of the three zones were strongly invaded with Acacia spp. trees, which were mechanically removed at the beginning of the study. The third zone was used as control area because it presented lower density of Acacia spp. and therefore no removal actions were applied. The obtained results confirmed that mechanical eradication of Acacia generates negative impacts upon wild rabbit abundances. However, it was observed that the different procedures used in Acacia control resulted in different levels of impact to rabbit populations. Methods used in 2005 were more aggressive, with strong declines in rabbit abundance caused by the destruction of feeding grounds, dens and shelter areas. Apart from the disturbance, which leads rabbits to move onto unmanaged areas, several cases of direct mortality by forest machinery were recorded. In 2006, direct mortality events were not recorded and a lower amount of rabbits moved away from the affected area. This study emphasises the potential of wild rabbit as an indicator of the sustainability of forest management practices. Furthermore, the abundance data obtained using a simple and cost-efficient methodology, constitute an important aid to the management of forest ecosystems.



Forestry management has been experiencing important changes in the last years. With communities becoming increasingly aware of the benefits of healthy forests, managers are being forced to adapt to a new paradigm of forestry management, mainly oriented to the integrity and sustainability of ecosystems (Bengtsson et al., 2000; Bodin and Wiman, 2007). In this way, the implementation of broad long-term monitoring programs to assess forests conservation status became one of the most important requirements of management strategies, because they allow the employment of an adaptative strategy in which management procedures are adjusted in real time and supported by scientifically valid information (Lindenmayer, 1999). These programs normally rely on biodiversity monitoring actions (CITAÇÂO). However, large-scale implementation of biodiversity monitoring studies is frequently difficult due to financial or logistic constraints (Lindenmayer, 1999; Pearman and Weber, 2007) thus, the use of indicator species has been suggested (Hanley, 1996; Noss, 1999; Kavanagh and Stanton, 2005).

According to Landres and collaborators (1988), indicator species are organisms whose characteristics are used as an index of attributes, when other variables are too difficult, inconvenient or expensive to measure in other species or at environmental conditions of interest. Therefore, the identification of the indicator species' requirements is a crucial step in the whole process otherwise this approach can simply result in the ignoring of other species needs or even their exposure to additional disturbance (Kremen, 1992). Thus, some of its requirements should be: being common, widespread, easy to monitor, sensitive to the management actions and having its biological and ecological features throroughly known (Hanley, 1996; Kavanagh and Stanton, 2005). Moreover, other requirements will depend on the monitoring goals, but normally it is assumed that indicator species should have ecological, economical or social relevance (Noss, 1999). Several taxonomic groups (ranging from Carabidae to ungulates) were used as indicator species in the past (Butterfield et al., 1995; Hanley, 1996; Mankin and Warner, 1999; Lundström-Gilliéron and Schlaepfer, 2003; Pearce and Venier, 2005; Drever et al., 2008).

In this study, wild rabbit (Oryctolagus cuniculus) is suggested as a potential indicator species to use in Iberian forest ecosystems. In fact, wild rabbit owns several features to play this role. Apart from being the main prey to a great number of vertebrate predators (Delibes and Hiraldo, 1979), wild rabbits are also known to have a great impact on other species populations both by decreasing the predation pressure to alternative preys or by influencing habitat structure and functionality (Delibes-Mateos et al., 2007). Wild rabbit is also a species of high economic and social value, especially as a game resource (Angulo, 2003). Despite its importance, Iberian populations of wild rabbit have suffered a drastic decline in the last few decades and currently is listed as "Almost threatened" CONFIRMAR (Cabral et al., 2005). Habitat loss and fragmentation (Moreno and Villafuerte, 1995; Carvalho and Gomes, 2003; Virgos et al., 2003), constant disease outbreaks (Calvete et al., 2002; Fouchet et al., 2006) (Fenner and Ross, 1994) and increasing hunting pressure (Angulo and Villafuerte, 2004; Calvete et al., 2005), along with high predation pressure upon diminished populations (Villafuerte, 1994; Villafuerte and Moreno, 1997), were the main reasons for Iberian rabbit decline.

The present study uses the variation in wild rabbit abundance to assess the impact of preventive forest practices related with the control of invasive species (Acacia longifolia). Information on rabbit abundance will allow adjusting forest practices, leading to the reduction of negative impacts thus ensuring better ecosystem management and minimizing its effects on biodiversity. The present study aims at answering the following questions: i) were there changes in rabbit abundance due to the mechanic removal of Acacia? ; ii) Can wild rabbit populations recover from the mechanical intervention and how fast is the recovery process? iii) what technical improvements can be made to the Acacia eradication forestry actions in order to decrease its negative impact on local fauna?

Material and methods

Study area

Quiaios National Forest is a 6,600 ha coastal forest situated in the Centre of Portugal (COORDENADAS). The area is included in the Natura 2000 network site "Dunas de Mira, Gândara e Gafanhas" (PTCON055). This site comprises an area of 22,511 ha located in the Mediterranean biogeographic region (Habitat Directive, 93/43/EEC). Also, the area includes 20 important habitats that support a great number of important species

including 20 important habitats that support a great number of important species (Vingada et al., 2000). The area is characterized by sandy soil and smoothly undulated surface (0-64 m a.s.l.). There are three different landscape units in Quiaios National Forest: dune, scrubland and pine forest. Dunes refer to inter-dunar fields limited by the frontal dune and by dense pine woods, interspersed by small pine and Acacia spp. patches and by a particular assemblage of vegetation composed by Salix arenaria, S. atrocinerea, Scirpus holoschoenus, Juncus spp., Carex spp., among others, as a consequence of temporal flooding during the wet season. Scrublands correspond to burned pinewood areas, which resulted from a major forest fire in 1993 and present a developing layer of P. pinaster and a very well developed shrub layer constituted by Corema album, Cytisus spp., Erica spp., Cistus spp. and Halimium halimifolium. The pine forest area is mainly composed by Pinus pinaster and P. pinea, presenting a patchy well-developed understorey of Myrica faya, Arbustus unedo and Acacia longifolia (Alves et al., 2006). In the last decade, the proliferation of some invasive species, especially Acacia longifolia, has become a critical problem, particularly in the areas affected by the 1993 forest fire. Natural ecosystems invasion by exotic species is widely recognized as a threat to native biodiversity (Samways et al., 1996; Vavra et al., 2007). In order to avoid the deterioration of the study area the forest authorities have been carrying out preventive activities against the spread of Acacia species since 2005

Control of invasive species

Mechanical interventions occurred in Autumn 2005 and 2006. The eradication program is based in the implementation of parallel bands of 2 and 4 meters, with different types of intervention. In the 2-meter bands the plant material is totally destroyed by mechanical rotary cutters pooled by tractors. Depending on the size of the material, trees can be previously cut down using chainsaws. In the 4-meter bands, all Acacia are removed manually (with chainsaws) and dragged to the 2-meter bands for mechanical mastication. Between the manual cut down and the mechanical mastication there was a delay of about one month that allows the plant material to dry, benefiting the mastication process. In the 2006 campaign, most tree logs were removed by local people before being mechanically ground up. As a result, there was smaller material left to grind and the number of passages of the rotary cutter necessary to produce and effective mastication were reduced (in the majority of the 2-meter bands, the cutter equipment only passed once) when compared with the 2005 campaign, where the rotary cutter passed 3 times over the same 2-meter band.

Sampling methods

Wild rabbit abundances

Rabbit abundance estimates were performed twice a year applying the latrine count method, from summer 2002 until winter 2008. Six linear transects (800-meter long and 5-meter wide). were defined in each biotope. Latrines were differentiated according to pellet number: (i) type A, until 50 pellets; (ii) type B, between 51 and 100 pellets; (iii) type C, with more than 100 pellets. The Abundance Index (AI) was estimated according to the following formula (for details see Sarmento and Cruz, 1998)

AI=(12,1C+5,1B+A)/TD (equation 1)

where, AI is the abundance index; A, B and C are the number of latrines corresponding to A, B and C types, respectively; TD is the transect distance (in meters).

Statistical analysis

The normality of data was tested through D'Agostino and Pearson tests. Because some data sets did not present a normal distribution all subsequent analysis were performed after log transformation (log (x+1)). Significant differences between biotopes with respect to the overall Abundance Index were analyzed using a unifactorial analysis of variance. The influence of season in wild rabbit abundance for each biotope was tested using a paired t-test. To understand differences in the seasonal use of each biotope during the entire study, a bifactorial variance analysis was performed, using biotopes and sampling periods has independent variables. Each analysis of variance was followed by a Tukey post-hoc test (in the case of one way ANOVA) and a Bonferroni post-hoc test (in the case of two-way ANOVA). Additionally, an agglomerative and ordinance analysis was used to analyze the data structure, using the Euclidean distance coefficient and the UPGMA method


Concerning the overall analysis of average rabbit abundances (Figure 1), no significant differences were found among zones. Although the highest abundance was found in the control zone, there were no significant differences between sampling zones before the intervention (F=0.984; p=0.37).

The maximum value of wild rabbit abundance (AI = 464.99) was recorded in the control area, during winter 2004, and the minimum (AI = 37.76) was reached in the 2005 mechanical control area during the winter of 2006 (Figure 2 and 3). The analysis of wild rabbit abundance variation revealed three patterns. In the first five sampling seasons, between summer 2002 and winter 2005, abundance values were similar in all sampling zones showing an ascending trend. However, a pronounced decrease was recorded in all sampling zones since winter 2005 until summer 2007, particularly in the mechanical intervention areas. Finally, a population recovery was observed in all sampled zones since summer 2007 until winter 2007/2008.

Two-way analysis of variance with factors "zone" and "season" (Table 1), only revealed significant differences in wild rabbit abundance among the different sampling seasons (F=5.493, p<0.001). However, significant differences between sampling zones were detected when compared rabbit abundances inside of each sampling season (Table 2). During winter 2006, the 2005 mechanical intervention area was significantly different from the control area (p<0.05) and also from the 2006 mechanical intervention area (p<0.05). In summer 2006, significant differences were only found between both mechanical intervention areas (p<0.001). Finally, in winter 2007 it was only possible to find significant differences between the 2006 mechanical intervention area and the control area (p<0.05). When comparing the evolution of abundance indices in each intervention zone (Table 3), significant differences were observed between the periods before and after the mechanical intervention (Bonferroni post hoc).

Agglomerative analysis (Figure 4 and 5) divided the different samples in two main groups. The first group includes abundance estimates obtained after mastication in both intervention zones, separating rabbit abundances referring to the 2005 from the 2006 interventions. In the second group, it was possible to observe a separation in 3 sub-groups: the first group aggregates samples of the last sampling seasons done in control area; the second sub-group joins samples of both intervention areas before mechanical interventions took place; and finally the third sub-group that includes samples performed in the control area during the initial stage of the study.


The present study confirmed the negative impacts of these large-scale mechanical interventions by showing a clear reduction in wild rabbit abundances. However, different levels of impact and recovery times were observed between interventions, probably related with the intervention methodology. The one-month period of waiting before grinding that occurred in the first campaign, allowed rabbits to use logs as shelter and resulted in a higher mortality. On the contrary, shelter possibility was highly reduced during the second campaign by removing a great part of cut material, thus minimizing the risk of being caught by the grinding equipment. Likewise, since the same intervention bands were grind more than once during the first campaign in order to destroy all the material, recolonization process was difficulted when compared to 2006.

Despite evidences suggesting the impact of mechanical intervention processes, it cannot be excluded that other biotic and environmental factors may have had also influenced wild rabbit populations during the study period. Indeed, it is also possible to find a slight decrease in the control area, relatively coincident with those found in both intervention areas. Likewise, the fact that recovery occurred simultaneously in all monitored areas may be indicative of a joint effect between more than one limiting factor. As a possible explanation, one can speculate that populations were being influenced by a severe drought that affected the study area between 2004 and 2007, with all the implications that it might carries in food availability, physiological conditions or foraging behaviour (Rueda et al., 2008). This possibility emphasises that, while the occurrence of a single limiting factor can be enough to induce the decline of a population, the conjugation of several favourable factors is needed for the recovery of species without restrictions.

When considering all the study period, the lack of significant differences between zones, seems to be due to the similar abundances registered before the interventions. However, when analysing each season individually, significant differences are found between control and intervention zones in those sampling seasons that followed mechanical interventions. These results highlight the importance of implementing regular and long-term monitoring programs since they allow managers to select the most appropriate time scale to assess the extent of problems as well as to identify its sources (Cabrera-Rodriguez, 2008; Delibes-Mateos et al., 2008).

Anyway, it seems clear that the methodology of fuel reduction and Acacia eradication used in Quiaios National Forest was the main driving force behind wild rabbit populations decline. It is, therefore, necessary to devise alternative methods with smaller impacts upon native species.

Practices monitored in this study are presently used in many state and private forests in Portugal. The use of mechanical mastication equipment towed by tractors is becoming a common forestry practice because it requires limited manpower and it is cost and time effective. It was also suggested to be a promising method for thinning overstocked forests with few negative impacts on soil compaction or water-polluting runoff (Hatchett et al, 2006). However, the risks of significantly affecting several animal populations must be considered, especially when referring to forests encompassing protected habitats and high biodiversity records.

Results showed that fewer log passages are likely to decrease the impact on rabbit populations. Therefore, the reduction of the area where mastication actually occurs would be a good procedure in large-scale mechanical interventions. This could be achieved by moving the major part of cut logs to a certain number of designated spots where grinding would takes place using wood clippers instead of towed grinding equipments. In this way, the major disturbance would be limited to smaller areas. Then towed grinding equipments could be used in a localized way. Some studies demonstrated that if forest mastication were performed in order to create a patchy environment, they could even enhance habitat quality for shrub and edge-associated species (Seavy et al, 2008). Although, interventions would have to be punctual.

Biological control would be another option to control Acacia species. It was successfully applied by Hoffman and collaborators (2002) through the introduction of a grazer insect (Trichilogaster acaciaelongifoliae) that affects both reproductive and vegetative parts of these plants. However, the use of any kind of biological control agents has to be extremely precautious. Previous assessments must be performed concerning the effects of the interaction between control agents and native biodiversity (Simberloff and Stiling, 1996).

When using any indicator species one must consider the variability inherent to each species, which may often overlap the remaining factors (Pearce and Venier, 2005). Concerning the use of wild rabbit, it is required to monitor a set of factors that are known to influence its populations, such as climate parameters (Palomares, 2003; Scanlan et al., 2006; Rueda et al., 2008), diseases (Calvete et al., 2002; Fouchet et al., 2006) or hunting pressure (Angulo and Villafuerte, 2004; Calvete et al., 2005). However, the results of this study underpin the feasibility of using this species as an indicator of the conservation status in Iberian ecosystems. The quick response showed by wild rabbit populations to this common forest management scheme is an indication of its usefulness in adaptive management approaches since it allows managers to react in real time to changes in ecosystems health. Its hability to recover from low numbers is another essential feature to be used in long-term programs. A trustworthy estimate of its populations were achieved using an indirect census method. Latrine count method revealed to be a simple monitoring process that can be easily applied in long-term and large-scale programs.

This work showed that long-term wild rabbit monitoring programs can provide reliable and comparable results, highly relevant at the ecosystem level, which may be a useful tool to forest managers, thus contributing to a more sustainable management of Mediterranean forest areas.