!"
"
Drove roads: keystone structures that promote ant diversity in !"
Mediterranean forest landscapes #"
$"
%"
Francisco M. Azcárate
a,*
, Javier Seoane
a
, Sara Castro
b
, Begoña Peco
a
&"
'"
a
TEG, Departamento de Ecología, Universidad Autónoma de Madrid. 28049, ("
Spain
)"
b
Unidad de Zoología, Departamento de Biología, Universidad Autónoma de *"
Madrid. 28049, Spain !+"
!!"
*
Corresponding author: Tel.: +34 91 497 3513; fax: +34 91 497 8001. E-mail !#"
address: fm.azcarate@uam.es !$"
!%"
#"
"
Abstract !&"
!'"
Drove roads are the traditional corridors used by pastoralists for seasonal !("
movements of livestock (transhumance). They cover a considerable land area !)"
in Mediterranean countries and, although they are an obvious source of !*"
landscape diversity, their influence on the diversity and composition of animal #+"
assemblages has not been documented. Ant communities were studied on four #!"
active drove roads, two in forests (submediterranean and conifer) and two in ##"
open environments (croplands and rangelands). They were compared with the #$"
respective matrix communities and their contribution to local species richness #%"
was evaluated. The effects were heavily dependent on the open or closed #&"
nature of the matrix. In forest environments, drove roads increased ant species #'"
richness at the local scale, acting as clear keystone structures. Their species #("
richness and functional diversity were highest on the fine scale, species #)"
composition was different, and a slight edge effect in the matrix was detected. #*"
In contrast, drove roads had little or even a negative effect in open environment $+"
locations. We conclude that drove roads have a high conservation value for $!"
ants in Mediterranean forest environments, in addition to their importance as $#"
reservoirs of plant biodiversity and generators of ecological goods and services. $$"
$%"
$&"
Keywords: Transhumance; Functional diversity; Spatial heterogeneity; $'"
Grasslands; Ant assemblages $("
$"
"
1. Introduction $)"
$*"
Drove roads, also known as stock routes, are one of the most characteristic %+"
components of traditional landscapes in Mediterranean countries (Ruiz and %!"
Ruiz, 1986; Mangas Navas, 1992; Merino García and Alier Gándaras, 2004). %#"
Active drove roads stand out in the landscape as well-defined strips up to 100 m %$"
in width with a savannah-like grassland appearance, in some cases running for %%"
several hundred kilometres. They are reserved for and ecologically modeled by %&"
transhumance, a traditional type of pastoralism consisting on the seasonal %'"
movement of livestock between summer and winter pastures. In Spain, drove %("
roads cover nearly 1% of the country's land area, with a total length of about %)"
125,000km (Mangas Navas, 1992; Merino García and Alier Gándaras, 2004). %*"
Due to their grid-like distribution, most of the country's land area is in contact &+"
with or near a drove road (Azcárate et al. 2012). Drove roads are at least &!"
several centuries old, and may have originally been based on the migratory &#"
routes of wild ungulates (Manzano and Casas, 2010). Because of their &$"
enormous area, long-term persistence, impact on the landscape structure and &%"
capacity to host herbivore migrations, drove roads can have played a major &&"
ecological role in the Mediterranean Basin. The current crisis in extensive &'"
grazing has led to the abandonment of transhumance and grazing uses of &("
drove roads (Ruiz and Ruiz, 1986; Ruiz, 2001), causing a loss of their &)"
differentiation from the surrounding ecological matrix. Their influence on &*"
populations and communities may thus disappear before they are identified and '+"
studied. '!"
'#"
%"
"
The few published ecological studies of drove roads have focused on their '$"
effects on plant communities. Drove roads have traditionally been regarded as a '%"
good example of ecological corridors for plant species (review in Bunce et al., '&"
2006), although no experimental evidence supported this view until Manzano ''"
and Malo (2006) detected epizoochorous seed dispersal over distances of up to '("
400 km. More recently, the effects of drove roads on landscape patterns, ')"
species composition and functional diversity of plant communities have been '*"
measured (Azcárate et al, 2012), showing that drove roads are a source of (+"
spatial heterogeneity and a reservoir for many plant species in non- or (!"
moderately-grazed habitats. (#"
($"
Drove roads could also have a noticeable effect on the diversity and (%"
composition of animal assemblages. Active drove roads maintain patches of (&"
open grassland in non-grazed environments such as forests, and hence ('"
increase spatial heterogeneity. Spatial heterogeneity and diversity of several (("
animal species groups are often (but not always) correlated (Duelli, 1997; Atauri ()"
and de Lucio, 2001; Szczepko et al., 2012). It has been argued that each (*"
animal species group depends on a specific structural aspect of the vegetation )+"
whose presence or quality can be detected at a certain spatial scale (Tews et )!"
al, 2004). At that scale, biodiversity is favoured by the occurrence of “keystone )#"
structures”, characterized by their ability to provide resources, shelter or nesting )$"
sites to that species group. Examples of keystone structures at different spatial )%"
scales are tree cavities in forests (for insects, birds and mammals), trees in )&"
African savannas (for arboreal rodents, ungulates, raptors and other species )'"
groups) and temporary wetlands in agricultural fields (for carabid beetles) (Tews )("
&"
"
et al., 2004; Remm and Lohmus, 2011). If a key structure affects several ))"
species groups, or groups with a strong influence on ecosystem functioning, )*"
then its conservation is of crucial importance. Drove roads might then function *+"
as keystone structures by favouring rich groups of terrestrial species with *!"
relevant roles on ecosystem functioning, such as ants. *#"
*$"
Ants are considered to be a focal group for the monitoring of terrestrial *%"
ecosystems (Underwood and Fisher, 2006; Crist, 2009). This is not surprising, *&"
given their ability to stockpile a considerable amount of primary and secondary *'"
production, interact with several organisms and act as ecosystem engineers *("
(Folgarait, 1998; MacMahon et al., 2000; Crist, 2009). Literature shows that *)"
ants respond strongly to land management (Bestelmeyer and Wiens, 1996; **"
Chen et al., 2011), and are sensitive to different levels of grazing (Read and !++"
Andersen, 2000; Boulton et al. 2005; Azcárate and Peco, 2011). Moreover, ants !+!"
are widespread, moderately diverse and easy to sample (Alonso and Agosti, !+#"
2000; Andersen et al., 2004). Their role as indicators has improved with the !+$"
development of the concept of functional groups to classify ants within species !+%"
assemblages, as first proposed in Australia (Andersen, 1995), and then !+&"
extended worldwide (Brown, 2000). More recently, the role of ants in ecosystem !+'"
functioning has been studied by measuring their functional diversity (Bihn et al., !+("
2010; Silva and Brandao, 2010) although this approach has still been little !+)"
addressed, in contrast to other taxa. !+*"
!!+"
The present study evaluates the role of drove roads as keystone structures. !!!"
Specifically, our work analyzes the effects of drove roads on ant assemblages !!#"
