Combining local data and scientific models to prioritize conservation for European ground squirrel and safeguard grassland habitats

doi:10.21203/rs.3.rs-4822522/v1

Download PDF

Research Article

Combining local data and scientific models to prioritize conservation for European ground squirrel and safeguard grassland habitats

https://doi.org/10.21203/rs.3.rs-4822522/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Context

Promoting habitat networks of natural grasslands within agricultural landscapes is crucial to supporting biodiversity and protecting endangered grassland species. Understanding the degree of fragmentation of these habitat networks assists in better elucidating their value to the grassland network. However, natural grassland characteristics in intensively used landscapes often need to be more adequately documented, which hinders effective grassland biodiversity conservation.

Objectives

We combined local data and modeling to identify conservation priorities for natural grasslands through assessing population and habitat patch characteristics for European Ground Squirrel (Spermophilus citellus, EGS), a keystone grassland specialist, in agricultural settings.

Methods

We used available information with presence/absence data and two spatially explicit models (LARCH and Circuitscape) to assess the potential of the current landscape in northern Serbia to protect the EGS. We applied the LARCH model to indicate potential habitat networks for the EGS and Circuitscape to assess connectivity of areas within and between these networks and identify areas of interventions that will serve as corridors between networks after restoration work. Together with the presence/absence data, this is used to set priorities for conservation actions for each network.

Results

We identified the presence of 15 habitat networks. The networks differ in connectivity, size, capacity, and sustainability to support local EGS populations.

Conclusions

The results revealed areas on which spatial adaptation measures and actions should be deployed to accommodate the long-term survival of EGS. In addition, the findings help the conservation of (semi)natural grassland and future land planning in terms of sustainable land use in an agricultural setting.

EGS

grasslands

connectivity

LARCH

conservation

monitoring data

Circuitscape

Increased food production in agricultural areas (especially during the second half of the 20th century) often results in habitat degradation, fragmentation and loss, requiring conservation actions to protect species inhabiting the remaining (semi)natural habitats, e.g., grasslands. Often the focus of these conservation actions is to use the remnant fragments and isolated sites as a foundation for developing regional green networks within which measures and actions may be deployed, albeit with limited resources (Green et al. 2018). Establishing a network of well-connected sites allows individuals of species to access food and water, find mates, conquer new potential habitats and exchange individuals with other populations (Hilty et al. 2020; Unnithan Kumar and Cushman 2022). In the long term, a stable habitat network also offers species the chance to shift their ranges as an attempt to adapt to climate change (Heller and Zavaleta 2009; Hannah 2011; Littlefield et al. 2019). However, network-level habitat connectivity rarely guides plans of conservation actions and measures (Jalkanen et al. 2020). Recent research emphasizes the imperative to augment wildlife sites in terms of size, quality, connectivity, and quantity (Lawton et al. 2010; Hodgson et al. 2011). Since stochastic events are more damaging for small and isolated populations, increasing population size by enlarging habitat patches (“bigger”), increasing their quality (“better”), connecting them with other patches using corridors (“joined”) and even creating new habitats through restoration actions (“more”) should work in favor of increasing the resilience of the habitat network and boosting metapopulation dynamics, crucial for fighting global changes (Isaac et al. 2018). Therefore, it is evident that habitat patch quality, size, and density, as well as the matrix in between the habitat patches must be considered when designing conservation strategies of species (Hanski and Ovaskainen 2000; Fahrig 2003; Opdam et al. 2003; Fahrig 2011; Didham et al. 2012; Rybicki and Hanski 2013; Synes et al. 2020).

Scattered within a heavily used agricultural landscape of central and south-eastern Europe, lives the endangered European ground squirrel (Spermophilus citellus L. 1766, EGS). Due to agricultural use, land abandonment, urbanization, and increasing road networks, the distribution of this species has declined in Serbia and across its range. As a result, it currently holds the status of strictly protected in the Serbian legislature (Official Gazette R. Serbia 2010), is listed in Annex II of the Habitats Directive and Annex IV of the Species Directive, and Appendix II of the Bern Convention, and has been categorized as endangered (EN) on the IUCN red list of species (Council Directive 92/43/EEC; Council of Europe 1979; Hegyeli 2020). The iconic grassland inhabitant requires precise management actions. As it is a keystone species of (semi)natural grasslands, its disappearance from these areas negatively impacts broader grassland communities (Lindtner et al. 2018, 2020). EGS is a small herbivore mammal (Rodentia: Sciuridae) with a functional role in improving soil structure, nutrient cycling, and maintaining plant community composition (Lindtner et al 2020). Its presence influences the abundance of other animal populations as it is an essential link between different trophic levels; and since individuals can alter habitats and regulate resources, EGS is considered a terrestrial ecosystem engineer (Lindtner et al. 2018; 2019; 2020). Therefore, we consider it a reliable indicator and model organism for conserving natural grassland ecosystems.

EGS remains in habitat patches in northern Serbia within a primarily agricultural landscape. Assessing habitat and landscape-level conditions and patterns can be significant, given the long-term downward trend in the distribution of grasslands and the adverse effects of this loss on the broad natural grassland community within the agricultural landscapes (Marini et al. 2019). For example, grassland habitat capacity and surrounding landscape heterogeneity showed environment-induced changes in small mammals’ local population distribution or demography (Nikolić et al., 2019; 2020). Available evidence indicates that the agricultural matrix may impede or facilitate the movement of individuals (Fahrig et al. 2011; Vasudev et al. 2015, Keeley et al. 2021), suggesting that structural and functional connectivity of grassland patches affects population viability and the long-term survival of targeted populations. Intensive population monitoring and comprehensive analysis of EGS habitat distribution could help prioritize conservation initiatives of natural grasslands within the agricultural landscape (Nikolić et al. 2019). For a ground-dwelling species like EGS, knowledge of landscape permeability and habitat connectivity is essential. Usually, this is acquired by including exact or estimated population data in a multi-scale, multi-level hierarchical modeling framework (McGarigal et al. 2016; Zeller et al. 2016; 2017). In these cases, spatially explicit models and keystone species data are relevant for choosing the most effective planning strategy for conservation, improving habitat quality, increasing quantity, and/or restoring connectivity (Hodgson et al. 2010; Marjakangas et al. 2023).

This study aimed to assess the viability of current populations of EGS in northern Serbia and determine the connectivity of natural grassland habitats to prioritize conservation measures and actions. The study's findings and recommendations could help inform conservation efforts and management strategies to support the viability of EGS populations and enhance the overall conservation of natural grassland habitats in northern Serbia. Thus, the investigation aimed to (1) establish the locations of connected grassland habitat clusters (habitat networks), (2) assess dispersal connectivity within and between these networks and identify areas of “interventions” which will serves as corridors after restoration work (3) use the presence/absence data to determine the differences among these networks and prioritize conservation measures. To achieve these objectives, we applied a framework (Fig. 1) that combined local knowledge of EGS habitat preferences and dispersal capacity (Ćosić et al. 2013; Nikolić et al. 2019), two scientific models, and monitoring data (Nikolić et al. 2019). The models were used to assess the potential viability of the populations within the ecological networks and the connectivity within and between these networks. The monitoring data was further used to prioritize conservation measures such as improving quality or connectivity within or between the networks.

Study area

The study was conducted in Vojvodina, an autonomous province in the northernmost part of Serbia with a total area of 21 506 km². Only about 6% of the area is under some form of protection (Puzović et al. 2015) and vast areas are designated for intensive agricultural production. EGS occupies 2.3% of the study area and its distribution has rapidly declined in the last few decades (Nikolić et al. 2019). As a result of the large-scale conversion of natural grasslands into arable land in Vojvodina during the second half of the 20th century, most local EGS populations are restricted to small remnant grassland patches inside a predominantly agricultural matrix (Nikolić et al. 2019). The number of occupied and abandoned patches (Fig. 1) varies across different spatial scales, reflecting the species’ response to the declining landscape heterogeneity (Fahrig et al. 2011).

Overview of modelling framework

We used a three-step framework to identify the conservation priorities in grassland habitat distribution for EGS (Fig. 2). In the first step, we identified the potential ecological networks for EGS. The ecological networks were determined based on the dispersal capacity of EGS and the potential viability of these networks was based on the habitat quality of the patches, the configuration of these networks and the presence of so-called potential key patches (Verboom et al. 2001). For this step, the model LARCH was used (Opdam et al. 2003, Verboom and Pouwels 2004). As the LARCH model indicates potential habitat networks based on species' dispersal capacity, it overestimates the connectivity between patches for ground-dwelling species such as EGS. Therefore, the potential viability of these networks is met when connectivity between patches is not a problem. In the second step, we developed resistance maps for EGS and identified areas with weak connectivity within the ecological networks. We also determined the potential connectivity between the networks to identify locations for corridors. For this step, the model Circuitscape was used (McRea et al. 2008; 2009). This provided information on which network lacks connectivity and where areas are located that could connect different networks. Finally, in step three, we combined the information from the models with monitoring data to distinguish between different categories of ecological networks and prioritize conservation measures for EGS in northern Serbia.

Ecological network of European ground squirrel habitats

For the present study, the map of 195 identified potential habitat patches provided by Nikolić et al. (2019) was adopted along with researchers’ field data and experience. We used the land use type to determine the main habitat suitability and four additional criteria as correction factors (see details in Supplementary file section S1.1). The total score for habitat quality in the patch was based on all five criteria (Table S1). The full list of scores and locations of the mapped habitat patches is available in Table S2.

The habitat map was input for LARCH to assess the habitat patch capacity and determine the habitat networks of EGS. A dispersion capacity of 5 km for ESG was adopted from Nikolić et al. (2019). Sometimes EGS can disperse further, which is important for genetic exchange. However, these rare events should be neglected for metapopulation dynamics, and dispersal distance in LARCH is set at a distance that includes 90% of all dispersal events (Opdam et al. 2003). LARCH estimates the potential number of reproductive units (RUs) in every patch based on the habitat quality and size of the patches (Verboom and Pouwels 2004). For small mammals, a patch of one ha with the highest quality index is expected to potentially accommodate at least 5 RUs. The obtained results can be used to identify key patches (KPs). A key patch is defined as a patch large enough to contain a population with an extinction chance of less than 5% in 100 years, given an immigration rate of 1 individual per generation (Verboom et al. 2001). These patches act as sources within ecological networks and are often occupied when the species is present in that specific ecological network (Verboom et al. 2001). The threshold for short-lived mammals of 100 RUs was used in this study to identify key habitat patches (Verboom et al. 2001, Verboom and Pouwels 2004). As individuals live in small colonies with a female biased sex ratio that are more sensitive to local extinction due to disturbances compared to species that reproduce as pair, we used 500 RUs as a threshold for a viable network for EGS instead of the standard of 200 RUs that is used for small rodents like voles (Verboom and Pouwels 2004).

Assessing connectivity within and between ecological networks

We used different methods for assessing the connectivity within- and between-ecological networks dispersal. We differentiated between the connectivity assessments to 1) identify patches within the networks that could be situated beyond the range of key patches, thereby posing potential risks for occupation and 2) to pinpoint potential corridors between ecological networks. For the connectivity within ecological networks, we assessed the connectivity between key patches to other patches in the ecological network as key patches act as sources and an ecological network is more stable when patches are well connected to key patches (Vos et al. 2007). We followed the variation in landscape patterns and their impact on EGS habitat cohesion at previously tested scales found by Nikolić et al. (2019). We assumed that the movement of individuals within the network is constrained by the quality of habitat and its surroundings. The characteristic of the habitat is defined in table S1 and for the quality of the surroundings we used information from Nikolić et al. (2019; 2020). For the assessment of the connectivity between ecological networks we assessed the connectivity between all patches as potential gene flow between networks is determined by all patches in the landscape. We assumed that dispersal of individuals is mainly determined by the type of land use, elevation and water courses between networks and not by detailed information within the networks (Mateo-Sánchez et al. 2015).

We used Circuitscape (v 4.0; McRea et al. 2009) to identify the area of the highest landscape permeability (between networks) and the potential movement trajectories of individuals within the habitat networks (within networks). Circuitscape uses circuit theory and resistance (or conductance) surfaces to predict connectivity between nodes (source), whereas high current intensity between them identifies areas and paths potentially crucial for patterns of animal movement (McRae et al. 2008; 2009). Thus, we developed species-specific baseline maps of landscape permeability and habitat connectivity for EGS in the lowland area of Vojvodina. We employed the resistance-by-distance method between mapped EGS patches (source–network nodes) to develop between and within network connectivity models. All 195 patches from the dataset Nikolić et al. (2019) were used as input for assessing connectivity between networks and key patches generated by the LARCH model for assessing connectivity within networks.

We generated a “current density” surface within the study area to assess between networks connectivity with the developed resistance raster and mapped habitat patches (see details in Supplementary file section S2.1). This model estimates connectivity across every possible movement trajectory among every pair of locations (mapped patches) in the so-called pairwise mode. To generate potential movement maps of individuals within habitat networks, we used the resistance surface, key populations as source nodes, and all other mapped patches as ground nodes - the locations individuals dispersed into (see details in Supplementary file section S2.2). This way, the “current” surface is estimated based on a 1:1 iteration between source and ground nodes, where we set source nodes to have a current value of 1 and ground nodes to have a current value of 0.

Prioritizing conservation measures for each network

We combined information from the LARCH analyses and the Circuitscape analyses with monitoring data to choose which main conservation strategy (Hodgson et al. 2010) or combination of strategies might be needed to improve the viability of population networks of EGS in northern Serbia. We distinguished between; current measures are sufficient, improve habitat quality of patches, restore more patches within network, improve connectivity within network and connect to other (viable) networks. For currently unoccupied networks it could be considered not to invest in further conservation efforts and use resources for improving still occupied networks that are not viable. Ecological networks with weak connectivity and potential areas for connecting ecological networks were used to indicate those areas that should be improved to enhance the dispersal of EGS within and between networks.

Viability of habitat networks of EGS

All habitat patches cover a combined area of 2586 ha in Vojvodina. Within the Banat region, 12.8% of the patches are of excellent or sound quality, while only 1% of the patches in Bačka and Srem are in this category (Supplementary file section S1.2; Table S2, Fig. S1). LARCH defined 15 potential habitat networks. Six of these networks have habitat patches big enough to sustain populations with more than 100 reproductive individuals (i.e., key populations; Fig. 3a). The number of key patches within those six networks vary (Table 1). Five networks, with ID_2, 4, 5, 8, and 13, are considered potentially viable (Table 1). One network, with ID_9, contains only one key patch, while the total network is large enough for approximately 350 RU.

Table 1

LARCH modeling results for the fifteen habitat networks. Names are based on the location within the regions of the potential populations within the habitat networks (see also Fig. 3a, b)
Network ID	Name	# Patches	# Key patches	Average quality	Sum RU	Viability
1	Small South Banat	5	0	0.45	84	no
2	Fruška gora	27	2	0.74	1012	yes
3	Farkaždin	1	0	0.00	0	no
4	Lok	5	2	0.65	607	yes
5	Greater South Banat	96	22	0.72	9187	yes (strongly)
6	Begejci	1	0	0.75	14	no
7	Gakovo	3	0	0.33	2	no
8	Central Banat	32	19	0.73	6885	yes (strongly)
9	Tomislavci	12	1	0.33	364	no
10	Bačko Dušanovo	1	0	0.50	4	no
11	Aleksa Šantić	1	0	1.00	23	no
12	Aradac	3	0	0.50	35	no
13	Trešnjevac	3	1	0.83	805	yes
14	Bikovo	3	0	0.67	98	no
15	Srpski Krstur	2	0	0.38	41	no

Connectivity within and between ecological networks

The ecological networks with the highest viability are the best-connected ones (Fig. 3b, Table 1). Of all the networks with more than ten patches, network 8 shows the highest connectivity (Fig. 4a, b). Network 5 is also well connected for most of the patches. Of all the viable networks, network 2 shows the lowest connectivity and many patches are not well connected with the more stable key patches.

Prioritization of conservation measures

Comparing the LARCH results and the monitoring data showed that potentially viable population networks and key patches provide a good base for protecting EGS. The results show that 93% of all key patches and 54% of others are occupied. Also, 66% of all patches are occupied in viable networks, and in non-viable networks, only 38%. The only patch occupied in network 9 was, in fact, the key patch. When the current network only has a few patches and these are all abandoned, they may be considered as a lost cause. Alternatively, they may be given a low priority as large efforts are probably needed; networks with ID 1, 3, 6, 7, 10, and 15. When resources are scarce, priority should be given to currently occupied patches at risk if they are not viable or have connectivity gaps. Based on the analyses, we conclude that two networks (5 and 8) need no further conservation measures. However, connecting them with the surrounding non-viable networks will improve overall networks sustainability. The analyses showed that the third viable network (2) lacks connectivity. It can be restored by improving the permeability within the network or by restoring more patches (Table 2). This will improve the gene flow in these networks’ total population of EGS. The analyses also show that in the northern part of Vojvodina, several small networks contain occupied key patches that are not viable; networks with IDs 9, and 13 and occupied networks without key patches; networks with IDs 11, 12 and 13. These networks are at risk of becoming abandoned as they are isolated. Connecting these networks with other one another or with network with ID 8 will improve the stability of the total EGS population in the northern part of Vojvodina.

Table 2

Overview of networks and preferred conservation measures based on the LARCH and Circuitscape results, and monitoring data. Connectivity within networks is assessed as high when large parts of the network have a high connectivity, moderate when some parts of the network have a high connectivity, low when none of the network has a high connectivity (Fig. 3a) and it is not assessed when the network consist of one habitat patch. Connectivity between networks is assessed as high when a potential corridor, regardless the distance, to another network is all high, it is assessed moderate when a potential corridor is partly high and low when it shows no clear potential corridor (Fig. 3b).
Netw. ID	Name	Average quality	Patches	% Occupied	Key patch	Viability	Connectivity within networks	Connectivity between networks	Cons. measures*	Conservation measures (Hodgson et al. 2010)
1	Small South Banat	0.45	5	0	no	no	low	high	1 and 4	improve quality and connect to 5; or no further conservation efforts
2	Fruška gora	0.74	27	78	yes	yes	low	moderate	2 and 3	improve connectivity within the network and restore more patches within the network
3	Farkaždin	0.00	1	0	no	no	-	high	1 and 4	improve quality and connect to 5; or no further conservation efforts
4	Lok	0.65	5	100	yes	yes	moderate	moderate	2 and 3	improve connectivity within the network and restore more patches within the network
5	Greater South Banat	0.72	96	61	yes	yes (strongly)	moderate	high	0	preserve current status of the network
6	Begejci	0.75	1	0	no	no	-	low	4	connect to 8; or no further conservation efforts
7	Gakovo	0.33	3	0	no	no	low	low	1 and 4	improve quality and connect to one large network with ID's 9-149; or no further conservation efforts
8	Central Banat	0.73	32	69	yes	yes (strongly)	high	high	0	preserve current status of the network
9	Tomislavci	0.33	12	8	yes	no	low	low	1, 2 and 4	improve quality, restore more patches and make one large network with ID's 9–14
10	Bačko Dušanovo	0.50	1	0	no	no	-	low	1 and 4	improve quality and make one large network with ID's 9–14; or no further conservation efforts
11	Aleksa Šantić	1.00	1	100	no	no	-	low	2 and 4	make one large network with ID's 9–14
12	Aradac	0.50	3	100	no	no	low	high	1 and 4	improve quality and make one large network with ID's 9–14 and/or connect to 8
13	Trešnjevac	0.83	3	67	yes	yes	moderate	high	4	make one large network with ID's 9–14
14	Bikovo	0.67	3	100	no	no	low	low	4	make one large network with ID's 9–14
15	Srpski Krstur	0.38	2	0	no	no	low	moderate	1 and 4	connect to one large network with ID's 9–14; or no further conservation efforts
* 0 = current measures are sufficient, 1 = improve quality of patches, 2 = restore more patches within network, 3 = improve connectivity within network and 4 = connect to other networks (with ID's)

Combining the results from LARCH and Circuitscape with the monitoring data from Nikolić et al. (2019) provides a good overview of potential conservation measures for each network. Although Circuitscape analyses show that the connectivity of some viable networks might need improvements (Table 2) certain discrepancies should be noted, as connectivity between network areas identified by the “circuit” method is significantly larger than expected based on within-network connectivity (Fig. 3a, b; Table 2). The connectivity results provide insight into the potential improvement zones for restoration measures to improve natural grassland cover and enhance the likelihood of long-term survival of EGS and grassland species in the study area. Thus, spatial plans should include an increase in habitat surface area, habitat density, and habitat quality (Verboom and Pouwels 2004; Bierwagen 2007; Kalarus and Novicki 2015; Van Teeffelen et al. 2015; Albert et al. 2017; Benedek and Sîrbu 2018; Benedek et al. 2021; Barão et al. 2022). The study confirmed that between-network and within-network connectivity is poor in the majority of identified habitat networks located north within our study landscape (i.e., all situated north of the ID_8 network). In this area, we should perform landscape-level conservation planning to increase the percentage distribution and density of natural grassland habitats and habitat-level measures to ensure adequate habitat management and improve the presence of transitional habitats. This approach would enable viable networks at the regional level in the most efficient way (Jackson and Fahrig 2012; Chen et al. 2023). The results yielded by this study, along with the produced maps, provide an example of where good spatial governance could support EGS and other natural grassland species and ecosystems.

Even though criticized, the species-specific network approach in highly modified agricultural areas is suitable since the complexity of its application is neutralized with biodiversity localized in the remaining semi-natural parts of the landscape (Jalkanen et al. 2020). For example, the identified KP in this study needs connections to all local populations by improving surrounding grassland habitat network links (e.g., to secure connections between the identified KP and other populations in the different parts of the network ID_2). Similarly, connecting the KP with other mapped but abandoned habitat patches by increasing the amount of habitat along with improving the quality by establishing regular habitat management (mowing or grazing) could potentially ensure the stability of such network. Sometimes translocation of individuals to those abandoned habitats might be necessary as some networks are isolated (e.g., in the network ID_9). Management of the sustainable network links is crucial when considering network cases such as the viability in the ID_2 network, which depends much more on the environment than on stochastic demographic processes (Ćosić 2015), highlighting the importance of spatial factors in the preservation efforts aimed at this part of the studied landscape. Furthermore, connecting isolated habitats characterized by medium-size capacity and populations with sufficiently large densities embodied in moderately permeable landscapes (e.g., ID_4) by steppingstone grassland corridors will increase the viability of populations within the network (Howell et al. 2018; Mims et al. 2023; Mohammadpour et al. 2023; Kim et al. 2024). Finally, improving grassland habitat density within the sustainable network in its impermeable parts (e.g., KP and other habitats in network ID_5) would positively affect abandoned habitats within the adjusted unsustainable ones (habitats in network ID_1, whose capacity needs improving). This comprehensive strategy would also enhance the local as well as the regional population’s resilience to the predicted increase in the frequency of extreme weather events because more extensive and more stable populations have a better chance of survival (Coetzee 2017; Frankham et al. 2017; Ashrafzadeh et al. 2020).

Transitional habitats that individuals use during dispersal differ significantly from those suitable for life and reproduction (Pulliam 2000; Cushman et al. 2013). In the present study, areas with only one or two inhabited or abandoned habitats (ID_3, ID_6, ID_7, ID_10, and ID_15) are essential for connectivity between networks and the connection of regional populations in the landscape. Furthermore, in improving the connections, we should simultaneously improve habitat and landscape characteristics (Howell et al. 2018; Fahrig 2019). Promoting connectivity between networks is relevant since even a few immigrants can establish gene flow between populations. This assertion supports the findings reported by Ćosić et al. (2013), indicating no genetic bottlenecks for EGS in Vojvodina in the recent past. As shown by available evidence, changes in land use can potentially prompt EGS to leave unsuitable areas. For example, Nikolić et al. (2019) have established that, compared to the historical prevalence of EGS populations in Vojvodina, they have recently moved east and south, where they currently thrive in the most significant numbers.

Researchers used the LARCH model in several studies to estimate the viability of populations on several dispersion scales (Van der Sluis et al. 2003; 2005; 2009; Pazúrová et al. 2018). The model relies on ecologically evaluated landscape indices (habitat suitability and capacity, dispersion, and population size) from the perspective of an analyzed species or group of species. The previous practice has shown that habitat capacity is a sensitive model parameter (Verboom et al. 2001; Verboom and Pouwels 2004; Regolin et al. 2021). For this reason, we conducted additional field research to evaluate the habitat capacity values yielded by the LARCH model. Moreover, even smaller areas can support more extensive and stable populations in these habitats, as shown by genetic analyses (Ćosić et al. 2013), indicating that combining model output with local knowledge improves the robustness of the results. For our research, the evaluation of habitat quality might even be further enhanced via quantitative methods such as analysis of satellite images and vegetation indices.

In the present study, for EGS - a grassland habitat specialist, we evaluated spatial connectivity within and between the habitat network scale to provide an overview of all connectivity links and potential corridors (McRae et al. 2008; Zeller et al. 2012). The scale of this spatial variation has already provided insights for natural grassland restoration and the proposal for designation of some regions of Vojvodina as designated ecological zones for protecting grassland ecosystems (Nikolić et al. 2019). Information related to the permeability of certain landscape areas is helpful to identify areas in which one should direct investments to promote grassland connectivity. For example, in their study, Ćosić et al. (2013) demonstrated that historically, the Danube is a barrier between populations. Still, the Tisza River is not. This assertion is confirmed by establishing the Vojvodina landscape matrix permeability. In addition, our analysis aids in identifying parts of functionally unlinked areas within and between networks, representing areas at which to focus revitalization measures of the grassland cover to support EGS viability and grassland biodiversity.

However, our findings are insufficient for determining how common EGS movements are, as the assessment of habitat connectivity within the heterogeneous matrix depends not only on individual traits but also on the available empirical data on the movement of individuals (Zeller et al. 2012; 2014). Therefore, the main limitation of the present study stems from the need for more information. Future research should focus on telemetry studies, landscape genetics analyses, and obtaining improved habitat maps (e.g., EUNIS level IV). Findings yielded by such investigations would significantly improve the current knowledge of the movement of EGS individuals through the landscape matrix and the response of individuals and populations to changes in land use. This information might help to improve parameters for dispersal capacity and permeability values of the landscape. This knowledge might be helpful in prioritizing the conservation measures needed in the northern part of the region where networks need to be connected to protect currently occupied networks, like networks ID_9 and 11–14.

In conclusion, conservation measures at the regional level could yield results quickly, establishing sustainable habitat networks capable of buffering climate change in the long term (Beier et al. 2008; Albert et al. 2017; Keeley et al. 2021). The implementation of active measures related to land use designation for agricultural activities and restoration of natural grassland habitats need to consider the ownership structure of parcels or the inclusion of the private sector into conservation initiatives (Waldron et al. 2020). Moreover, when planning designated areas, a comparative analysis of people's societal and economic needs that depend on the targeted landscape is mandatory. In this context, the spatial approach can be precious, as it facilitates collaboration among different sectors and interest groups on strategic planning (Keeley et al. 2019; Hilty et al. 2020). Finally, even though this study only focused on the European ground squirrel as a model organism, the conceptual and methodological approach we used and the results we obtained might be applied for other species and ecosystems to prioritize between conservation measures to improve habitat quality, increase habitat quantity or improve connectivity (i.e. Hodgson et al. 2010).

Acknowledgements We would like to thank the European ground squirrel community, Bird Protection and Study Society of Serbia and local community for friendly advice during fieldwork campaign.

Funding: This work has been supported by The Ministry of Science, Technological Development and innovation, Republic of Serbia, under Grant 2024: 451-03-66/2024-03/ 200358, The Rufford Foundation grant “Building a better future for European ground squirrel in Serbia” and H2020 project ANTARES (SGA-CSA. No. 739570).

Competing Interests: The authors have no relevant financial or non-financial interests to disclose.

Author Contributions: TNL, RP and W GW W, contributed to the study conception and design. Data collection and analysis were performed by TNL, MA, DR, DĆ and NĆ. The first draft of the manuscript was written by TNL, RP and MA and all authors commented on previous versions. All authors read and approved the final manuscript.

Data Availability: The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Albert CH, Rayfield B, Dumitru M, Gonzalez A (2017) Applying network theory to prioritize multispecies habitat networks that are robust to climate and land‐use change. Cons Biol 31(6):1383-1396. https://doi.org/10.1111/cobi.12943
Ashrafzadeh MR, Khosravi R, Adibi MA, Taktehrani A, Wan HY, Cushman SA (2020) A multi-scale, multi-species approach for assessing effectiveness of habitat and connectivity conservation for endangered felids. Biol Conserv, 245, 108523. https://doi.org/10.1016/j.biocon.2020.108523
Barão I, Queirós J, Vale-Gonçalves H, Paupério J, Pita R (2022) Landscape characteristics affecting small mammal occurrence in heterogeneous olive grove agro-ecosystems. Conservation, 2(1), 51-67. https://doi.org/10.3390/conservation2010005
Beier P, Majka DR, Spencer WD (2008) Forks in the road: choices in procedures for designing wildland linkages. Conserv Biol, 22(4), 836-851. https://doi.org/10.1111/j.1523-1739.2008.00942.x
Benedek AM, Sîrbu I, Lazăr A (2021) Responses of small mammals to habitat characteristics in Southern Carpathian forests. Sci Rep, 11(1), 12031. https://doi.org/10.1038/s41598-021-91488-6
Benedek AM, Sîrbu I (2018) Responses of small mammal communities to environment and agriculture in a rural mosaic landscape. Mamm Biol 90(1):55-65. https://doi.org/10.1016/j.mambio.2018.02.008
Bierwagen, BG (2007) Connectivity in urbanizing landscapes: The importance of habitat configuration urban area size and dispersal. Urban Ecosyst 10(1):29-42. https://doi.org/10.1007/s11252-006-0011-6
Chen X, Kang B, Li M, Du Z, Zhang L, Li H (2023) Identification of priority areas for territorial ecological conservation and restoration based on ecological networks: A case study of Tianjin City, China. Ecol Indic, 146, 109809. https://doi.org/10.1016/j.ecolind.2022.109809
Coetzee BW (2017) Evaluating the ecological performance of protected areas. Biodivers Conserv 26(1):231-236. https://doi.org/10.1007/s10531-016-1235-2
Convention on Conservation of European WildLife and Natural Habitats. (1979) Appendix II
Ćosić N, Říčanová Š, Bruja J, Penezić А, Ćirović D (2013) Do rivers and human-induced habitat fragmentation affect genetic diversity and population structure of the European ground squirrel at the edge of its Pannonian range? Conserv Genet 14:345-354. https://doi.org/10.1007/s10592-013-0466-y
Ćosić N (2015) Vijabilnost populacija tekunica (Spermophilus citellus L.1766) na području Srbije. PhD thesis. Biološki fakultet, Univerzitet u Beogradu. Belgrade, Serbia.
Cushman SA, McRae B, Adriaensen F, Beier P, Shirley M, Zeller K (2013) Biological corridors and connectivity. Chapter 21. In: Macdonald DW; Willis KJ (ed) Key Topics in Conservation Biology 2. Hoboken NJ, Wiley-Blackwell, pp 384-404.
Didham RK, Kapos V, Ewers RM (2012) Rethinking the conceptual foundations of habitat fragmentation research. Oikos 121(2):161-170. https://doi.org/10.1111/j.1600-0706.2011.20273.x
Fahrig L (2003) Effects of habitat fragmentation on biodiversity. Annu Rev Ecol Evol Syst 34(1):487-515. https://doi.org/10.1146/annurev.ecolsys.34.011802.132419
Fahrig L, Baudry J, Brotons L, Burel FG, Crist TO, Fuller RJ, Sirami C, Siriwardena GM, Martin JL (2011) Functional landscape heterogeneity and animal biodiversity in agricultural landscapes: Heterogeneity and biodiversity. Ecol Lett 14:101–112. https://doi.org/10.1111/j.1461-0248.2010.01559.x
Fahrig L (2019) Habitat fragmentation: A long and tangled tale. Glob Ecol Biogeogr 28(1):33-41. https://doi.org/10.1111/geb.12839
Frankham R, Ballou JD, Ralls K, Eldridge M, Dudash MR, Fenster CB, Lacy RC, Sunnucks P (2017) Genetic management of fragmented animal and plant populations. Oxford University Press. https://doi.org/10.1093/oso/9780198783398.001.0001 Green AW, Correll MD, George TL, Davidson I, Gallagher S, West C, Lopata A, Casey D, Ellison K, Pavlacky Jr, DC, Quattrini, L. (2018) Using structured decision making to prioritize species assemblages for conservation. J Nat Conserv 45:48-57. https://doi.org/10.1016/j.jnc.2018.08.003
Hannah L (2011) Climate change, connectivity, and conservation success. Conserv Biol 25:1139–1142. https://doi.org/10.1111/j.1365-2664.2009.01695.x
Hanski I, Ovaskainen O (2000) The metapopulation capacity of a fragmented landscape. Nature 404, 755–758. https://doi.org/10.1038/35008063
Hegyeli, Z (2020) Spermophilus citellus. The IUCN Red List of Threatened Species 2020: e.T20472A91282380. https://dx.doi.org/10.2305/IUCN.UK.2020-2.RLTS.T20472A91282380.en. Downloaded on 23 March 2023.
Heller, NE, Zavaleta ES (2009) Biodiversity management in the face of climate change: a review of 22 years of recommendations. Biol Conserv 142:14–32. https://doi.org/10.1016/j.biocon.2008.10.006
Hilty J, Worboys GL, Keeley A, Woodley S, Lausche B, Locke H, Carr M, Pulsford I, Pittock J, White JW, Theobald DM, Levine J, Reuling M, Watson JEM, Ament R, Tabor GM (2020) Guidelines for conserving connectivity through ecological networks and corridors. Best Practice Protected Area Guidelines Series No. 30. Gland, Switzerland: IUCN pp 120 https://doi.org/10.2305/IUCN.CH.2020.PAG.30.en
Hodgson JA, Moilanen A, Wintle BA, Thomas CD (2011) Habitat area, quality and connectivity: striking the balance for efficient conservation. J Appl Ecol 48(1):148-152. https://doi.org/10.1111/j.1365-2664.2010.01919.x
Howell PE, Muths E, Hossack BR, Sigafus BH, Chandler RB (2018) Increasing connectivity between metapopulation ecology and landscape ecology. Ecology, 99(5):1119-1128. https://doi.org/10.1002/ecy.2189
Jackson HB, Fahrig L (2012) What size is a biologically relevant landscape? Landsc Ecol 27:929–941. https://doi.org/10.1007/s10980-012-9757-9
Jalkanen J, Toivonen T, Moilanen A (2020) Identification of ecological networks for land-use planning with spatial conservation prioritization. Landsc Ecol 35(2):353-371. https://doi.org/10.1007/s10980-019-00950-4
Kalarus K, Nowicki P (2015) How do landscape structure management and habitat quality drive the colonization of habitat patches by the dryad butterfly (Lepidoptera: Satyrinae) in fragmented grassland? PloS One 10(9):e0138557. https://doi.org/10.1371/journal.pone.0138557
Kenyeres Z, Bauer N, Nagy L, Szabó S. (2018) Enhancement of a declining European ground squirrel (Spermophilus citellus) population with habitat restoration. J Nat Conserv 45:98-106. https://doi.org/10.1016/j.jnc.2018.08.006
Keeley AT, Beier P, Creech T, Jones K, Jongman RH, Stonecipher G, Tabor GM (2019) Thirty years of connectivity conservation planning: An assessment of factors influencing plan implementation. Environ Res Lett 14:103001 https://doi.org/10.1088/1748-9326/ab3234
Keeley AT, Beier P, Jenness JS (2021) Connectivity metrics for conservation planning and monitoring. Biol Conserv, 255, 109008. https://doi.org/10.1016/j.biocon.2021.109008
Kim ES, Lee DK, Choi J, Kim JH, Mo Y, Jeon Y, Kim JY (2024) Metapopulation models using landscape connectivity can better reflect landscape heterogeneity. Ecol Inform, 80, 102464. https://doi.org/10.1016/j.ecoinf.2024.102464
Lindtner P, Ujházy K, Svitok M, Kubovčík V (2018) The European ground squirrel increases diversity and structural complexity of grasslands in the Western Carpathians. Mamm Res 63:223-229. https://doi.org/10.1007/s13364-017-0349-6
Lindtner P, Gömöryová E, Gömöry D, Stašiov S, Kubovčík V (2019) Development of physico-chemical and biological soil properties on the European ground squirrel mounds. Geoderma 339:85-93. https://doi.org/10.1016/j.geoderma.2018.12.043
Lindtner P, Svitok M, Ujházy K, Kubovčík V (2020) Disturbances by the European ground squirrel enhance diversity and spatial heterogeneity of plant communities in temperate grassland. Biodivers Conserv 29(3):853-867. https://doi.org/10.1007/s10531-019-01914-5
Littlefield CE, Krosby M, Michalak JL, Lawler JJ (2019) Connectivity for species on the move: supporting climate‐driven range shifts. Front Ecol Environ, 17(5), 270-278. doi:10.1002/fee.2043
Marini L, Bartomeus I, Rader R, Lami F (2019) Species–habitat networks: A tool to improve landscape management for conservation. J App Ecol, 56(4):923-928. https://doi.org/10.1111/1365-2664.13337
Marjakangas EL, Santangeli A, Kujala H, Mammola S, Lehikoinen A (2023) Identifying ‘climate keystone species’ as a tool for conserving ecological communities under climate change. Divers Distrib, 29(11), 1341-1354. 10.1111/ddi.13764
Mateo-Sánchez MC, Balkenhol N, Cushman S, Pérez T, Domínguez A, Saura S (2015) Estimating effective landscape distances and movement corridors: comparison of habitat and genetic data. Ecosphere, 6(4), 1-16. https://doi.org/10.1890/ES14-00387.1
McGarigal K, Wan HY, Zeller KA, Timm BC, Cushman, SA (2016) Multi-scale habitat selection modeling: a review and outlook. Landsc Ecol, 31(6):1161-1175. https://doi.org/10.1007/s10980-016-0374-x
McRae BH, Dickson BG, Keitt TH, Shah VB (2008) Using circuit theory to model connectivity in ecology, evolution and conservation. Ecology 89:2712–2724. https://doi.org/10.1890/07-1861.1
McRae B, Shah VB, Mohapatra TK (2009) Circuitscape v 405.
Mims MC, Drake JC, Lawler JJ, Olden JT (2023) Simulating the response of a threatened amphibian to climate-induced reductions in breeding habitat. Landsc Ecol 38, 1051–1068. https://doi.org/10.1007/s10980-023-01599-w
Mohammadpour N, Jahanishakib F, Asadolahi Z (2023) Development of sustainable protection of biodiversity using modeling corridors of the landscape based on circuit theory. Geogr Environ Sustain 13(4), 55-68. 10.22126/GES.2023.9490.2680
Nikolić T, Radišić D, Ćosić N, Díaz-Delgado R, Milić D, Vujić A, Ćirović D (2019) Landscape heterogeneity effects on keystone rodent species: Agro-ecological zoning for conservation of open grasslands. Biodivers Conserv 28: 3139–3158. https://doi.org/10.1007/s10531-019-01810-y
Nikolić T, Arok M, Radišić D, Mirč M, Velaja L, Milić D, Ćirović D (2020) Endangered species’ trait responses to environmental variability in agricultural settings. Arch Biol Sci 72(1):13-21. https://doi.org/10.2298/ABS190715061N
Opdam P, Verboom J, Pouwels R (2003) Landscape cohesion: an index for the conservation potential of landscapes for biodiversity. Landsc Ecol 18(2):113-126. https://doi.org/10.1023/A:1024429715253
Pazúrová Z, Pouwels R, Ružičková J, Bolliger J, Krokusová J, Oťaheľ J, Pazúr R (2018) Effects of landscape changes on species viability: a case study from northern Slovakia. Sustainability 10(10):3602. https://doi.org/10.3390/su10103602
Pulliam HR (2000) On the relationship between niche and distribution. Ecol Lett 3(4):349-361. https://doi.org/10.1046/j.1461-0248.2000.00143.x
Puzović S, Panjković B, Tucakov M, Stojnić N, Sabadoš K, Stojanović T, Vig L, Marić B, Tešić O, Kiš A, Galamboš L, Pil N, Kicošev V, Stojšić V, Timotić D, Perić R, Bošnjak T, Delić J, Dobretić V, Stanišić J (2015) Upravljanje prirodnom baštinom Vojvodine. Pokrajinski sekretarijat za urbanizam, graditeljstvo i zaštitu životne sredine. Pokrajinski zavod za zaštitu prirode. “Stojkov” Novi Sad.
Rammou DL, Kavroudakis D, Youlatos D (2021) Distribution, population size, and habitat characteristics of the endangered european ground squirrel (Spermophilus citellus, Rodentia, Mammalia) in its southernmost range. Sustainability 13(15):8411. https://doi.org/10.3390/su13158411
Regolin AL, Oliveira-Santos LG, Ribeiro MC, Bailey LL (2021) Habitat quality, not habitat amount, drives mammalian habitat use in the Brazilian Pantanal. Landsc Ecol, 36(9), 2519-2533. https://doi.org/10.1007/s10980-021-01280-0
Riordan‐Short E, Pither R, Pither J (2023) Four steps to strengthen connectivity modeling. Ecography, 2023(11), e06766. https://doi.org/10.1111/ecog.06766
Rybicki J, Hanski I (2013) Species–area relationships and extinctions caused by habitat loss and fragmentation. Ecol Lett, 16:27-38. https://doi.org/10.1111/ele.12065
Službeni glasnik Republike Srbije. Uredba o ekološkoj mreži 102/2010.
The Habitats Directive 92/43/EEC. 1992.
Synes NW, Ponchon A, Palmer SC, Osborne PE, Bocedi G, Travis JM, Watts K (2020) Prioritising conservation actions for biodiversity: Lessening the impact from habitat fragmentation and climate change. Biol Conserv, 252, 108819. https://doi.org/10.1016/j.biocon.2020.108819
Unnithan Kumar S, Cushman SA (2022). Connectivity modelling in conservation science: a comparative evaluation. Sci Rep, 12(1), 16680. https://doi.org/10.1038/s41598-022-20370-w
Van der Sluis TB, Pedroli HB, Corridore G (2003) Ecological networks a challenge for territorial planning – the case of region Abruzzo Italy. In: E Tiezzi CA Brebbia, J-L Uso (ed) Proceedings of Ecosystems and sustainable development ECOSUD. Vol I: Siena, Italy. Wessex Institute of Technology, Southampton, UK, pp 66-79
Van der Sluis TJ, Romanowski J, Matuszkiewiecz I, Bouwma M (2005) Comparison of scenarios for the Vistula River Poland. In: Hong SK, Nakagoshi N, Fu B, Morimoto (ed) Landscape Ecological Application in Man-Influenced Areas Linking Man and Nature Systems. Chapter 25. Springer Netherlands, Dordrecht, pp 417-433
Van der Sluis GT, Gosselink JMJ, Slim PA, Verhagen A, van Keulen H (2009) Restoration of degraded steppe land – opportunities for Lugansk Oblast Eastern Ukraine Wageningen Alterra Alterra report 1935. 62.
Van Teeffelen AJ, Vos CC, Jochem R, Baveco JM, Meeuwsen H, Hilbers JP (2015) Is green infrastructure an effective climate adaptation strategy for conserving biodiversity? A case study with the great crested newt. Landsc Ecol 30(5):937-954. https://doi.org/10.1007/s10980-015-0187-3
Vasudev D, Fletcher R J, Goswami V R, Krishnadas M (2015) From dispersal constraints to landscape connectivity: lessons from species distribution modeling. Ecography 38:967-978. https://doi.org/10.1111/ecog.01306
Verboom J, Foppen R, Chardon P, Opdam P, Luttikhuizen P (2001) Introducing the key patch approach for habitat networks with persistent populations: an example for marshland birds. Biol Conserv 100:89-101. https://doi.org/10.1016/S0006-3207(00)00210-X
Verboom J, Pouwels R (2004) Ecological functioning of ecological networks: a species perspective. In: Jongman RHG, Pungetti G (ed) Ecological networks and greenways: concept, design, implementation. Cambridge University Press, Cambridge UK, pp 65–72
Vos CC, Opdam P, Steingröver EG, Reijnen R (2007) Transferring ecological knowledge into landscape planning: a design method for ecological corridors. In: Key topics and perspectives in landscape ecology. Cambridge University Press, pp227-245.
Waldron A, Adams V, Allan J, Arnell A, Asner G, Atkinson S, Baccini A, Baillie J, Balmford A, Beau JA, Brander L, Brondizio E, Bruner A, Burgess N, Burkart K, Butchart S, Button R, Carrasco R, Cheung W, Christensen V, Clements A, Coll M, Marco MD, Deguignet M, Dinerstein E, Ellis E, Eppink F, Ervin J, Escobedo A, Fa J, Fernandes-Llamazares A, Fernando S, Fujimori S, Fulton B, Garnett S, Gerber J, Gill, D, Gopalakrishna, T, Hahn, N, Halpern, B, Hasegawa, T, Havlik, P, Heikinheimo, V, Heneghan, R, Henry, E, Humpenoder, F, Jonas, H, Jones, K, Joppa, L, Joshi, A, Jung, M, Kingston, N, Klein, C, Krisztin, T, Lam, V, Leclere, D, Lindsey, P, Locke, H, Lovejoy, T, Madgwick, P, Malhi, Y, Malmer, P, Maron, M, Mayorga, J, Meijl, HV, Miller, D, Molnar, Z, Mueller, N, Mukherjee, N, Naidoo, R, Nakamura, K, Nepal, P, Noss, R, O’Leary, B, Olson, D, Abrantes, JP, Paxton, M, Popp, A, Possingham, H, Prestemon, J, Reside, A, Robinson, C, Robinson, J, Sala, E, Scherrer, K, Spalding, M, Spenceley, A, Steenbeck, J, Stehfest, E, Strassborg, B, Sumaila, R, Swinnerton, K, Sze, J, Tittensor, D, Toivonen, T, Toledo, A, Torres, PN, Zeist, WV, Vause, J, Venter, O, Vilela, T, Visconti, P, Vynne, C, Watson, R, Watson, J, Wikramanayake, E, Williams, B, Wintle, B, Woodley, S, Wu, W, Zander, K, Zhang, Y Zhang, Y (2020) Protecting 30% of the planet for nature: costs, benefits and economic implications. Working paper analysing the economic implications of the proposed 30% target for areal protection in the draft post-2020 Global Biodiversity Framework.
Zaharia G, Petrencu L, Baltag EŞ (2016) Site selection of European ground squirrels (Spermophilus citellus) in Eastern Romania and how they are influenced by climate, relief, and vegetation. Turk J Zool 40:917-924. https://doi.org/10.3906/zoo-1505-28
Zeller KA, McGarigal K, Whiteley AR (2012) Estimating landscape resistance to movement: a review. Landsc Ecol 27:777-797. https://doi.org/10.1007/s10980-012-9737-0
Zeller KA, McGarigal K, Beier P, Cushman SA, Vickers TW, Boyce WM (2014) Sensitivity of landscape resistance estimates based on point selection functions to scale and behavioral state: pumas as a case study. Landsc Ecol 29:541-557. https://doi.org/10.1007/s10980-014-9991-4
Zeller KA, McGarigal K, Cushman SA, Beier P, Vickers TW, Boyce WM (2016) Using step and path selection functions for estimating resistance to movement: pumas as a case study. Landsc Ecol, 31:1319-1335. https://doi.org/10.1007/s10980-015-0301-6
Zeller KA, McGarigal K, Cushman SA, Beier P, Vickers TW, Boyce WM (2017) Sensitivity of resource selection and connectivity models to landscape definition. Landsc Ecol 32:835-855. https://doi.org/10.1007/s10980-017-0489-8

No competing interests reported.

AppendixConsEGSRMT.docx

Download PDF

Reviewers agreed at journal
02 Sep, 2024
Reviewers agreed at journal
01 Sep, 2024
Reviewers agreed at journal
30 Aug, 2024
Reviews received at journal
21 Aug, 2024
Reviewers agreed at journal
07 Aug, 2024
Reviewers invited by journal
05 Aug, 2024
Editor assigned by journal
31 Jul, 2024
Submission checks completed at journal
31 Jul, 2024
First submitted to journal
29 Jul, 2024

You are reading this latest preprint version

Combining local data and scientific models to prioritize conservation for European ground squirrel and safeguard grassland habitats

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Study area

Overview of modelling framework

Ecological network of European ground squirrel habitats

Assessing connectivity within and between ecological networks

Prioritizing conservation measures for each network

Results

Viability of habitat networks of EGS

Connectivity within and between ecological networks

Prioritization of conservation measures

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1