Protected Area Network Contributions to Landscape Connectivity under Global Change

doi:10.21203/rs.3.rs-2452023/v1

Download PDF

Research Article

Protected Area Network Contributions to Landscape Connectivity under Global Change

https://doi.org/10.21203/rs.3.rs-2452023/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Context Both anthropogenic change and environmental variability cause spatiotemporal fluctuations in the availability of habitat on a landscape. Land acquisition and restoration for conservation also facilitates dynamic habitat accessibility. Accounting for these spatiotemporal dynamics is critical to the success of conservation planning to increase landscape connectivity.

Objectives We aimed to forecast the individual and interactive effects of anthropogenic change, environmental variability, and conservation action on landscape connectivity over time. In doing so, we sought to evaluate the capacity of conservation actions intended to enhance connectivity to keep pace with global change.

Methods We used a landscape change model to simulate strategies for the expansion of protected area networks on a ~1.5 million ha landscape in North Carolina. We forecasted spatiotemporal changes (2020-2100) to complex habitat networks across multiple scenarios of climate and land-use change, using graph theory to assess the variance in connectivity between conservation strategies and among global change futures.

Results Both climate and land-use change reduced the influence of conservation on landscape connectivity, but through different mechanisms. Climate change increased available longleaf pine habitat, improving overall landscape connectivity and camouflaging connectivity improvements from conservation. Land-use change reduced the connectivity improvements facilitated by conservation as the cost of movement between habitat patches increased.

Conclusion Rather than acting synergistically with land-use change, climate change negated the connectivity declines from land-use change in the study area. More broadly, our approach to forecast landscape-level responses to climate change, land-use change, and management can support the identification of conservation strategies that are more robust to global change, better enhancing landscape connectivity.

connectivity

protected area expansion

global change

conservation forecasting

forest restoration

graph theory

Anthropogenic climate and land-use change represent two of the most significant drivers of habitat loss worldwide (Costanza & Terando 2019) and may act synergistically, exacerbating habitat decline and reducing landscape connectivity, thereby increasing the vulnerability of fragmented populations to extinction (Mantyka-Pringle et al. 2015; Dilts et al. 2016). Connected landscapes facilitate species movement between resource patches (Taylor et al. 1993) and enable dispersal in response to changing environments, allowing species to move following suitable climatic conditions as the climate changes (Perring et al. 2015). By promoting migration, dispersal, and gene flow, connected landscapes support the persistence of wildlife populations and their adaptation to environmental change (Santini et al. 2016). Because of this, strategies to maintain and enhance landscape connectivity, such as the expansion of protected area networks (Andrello et al. 2015), have been increasingly prioritized in conservation efforts to combat biodiversity decline (Heller & Zavaleta 2009; Nuñez et al. 2013). By providing and protecting additional habitat and reducing habitat conversion, the creation of new protected areas can augment connectivity and support adaptation to climate change (Le Saout et al. 2013; Elsen et al. 2020).

However, the design of protected area networks, which is often guided by decisions made to meet specific conservation objectives (Mawdsley 2011), can yield different configurations of protected areas with variable contributions to landscape connectivity (Bergsten et al. 2013; Di Marco et al. 2015, Reside et al. 2018). Additionally, climate and land-use change can reduce the connectedness of existing protected area networks and may create connectivity limits within protected area network expansions, diminishing their capacity to help species respond to environmental change (Asamoah et al. 2021). Some configurations of protected areas may better support climate adaptation and enhance landscape connectivity than others. For example, establishing new protected areas near or adjacent to existing protected areas has been suggested to enhance resilience to anthropogenic change by creating larger core areas of habitat and reducing edge effects, thereby increasing the effective size of populations (Oliver & Morecroft 2014; Herse et al. 2018). Establishing new protected areas on land with high geodiversity characteristics has also been recommended, as these landscapes may result in more environmental niches (Comer et al. 2015) and as such possess a greater capacity to maintain species diversity and ecological function as the climate changes which may confer greater resilience (Anderson et al. 2015). Designing reserves (and linkages) around geodiversity attributes or land facets has also been suggested to support connectivity in a changing climate (Beier and Brost 2010). With renewed emphasis on national and global protected area targets (Dinerstein et al. 2019), the identification of protected area network designs that both enhance connectivity and support adaption to global change is essential to guide conservation investments.

Additionally, anthropogenic change is adding uncertainty to the outcomes of conservation and the ability of protected areas to achieve their intended biodiversity targets, necessitating the identification of protected area network designs that will remain effective under a range of landscape trajectories (Hannah 2010; Albert et al. 2017). Proactively integrating landscape-level interactions between climate change, land-use/land cover change, and environmental variability into conservation planning is critical as their interactive effects will likely change where and how to best allocate resources (de Chazal & Rounsevell 2009). Studies on conservation strategy prioritization often incorporate future climate change or future land-use change but seldom consider both drivers jointly, especially when forecasting future habitat connectivity (Costanza & Terando 2019). Similarly, studies forecasting conservation influences on landscape connectivity have often evaluated conservation at coarse scales (Carlson et al. 2019) or lacked the inclusion of dynamic landscapes across long time horizons (Martensen et al. 2017), though the inclusion of landscape dynamics is gaining momentum (Jennings et al. 2021). As such, forecasting the landscape-level effects of global change and the landscape-level responses to future conservation for landscape connectivity represents an understudied research need for future protected area planning and prioritization.

In this study, we evaluated the individual and interactive influences of environmental variability, climate change (CC), land-use change (LUC), and conservation action on potential landscape connectivity. We used a landscape change model to forecast the trajectory of forests in central North Carolina from 2020 to 2100 under four different global change scenarios, using graph theory to evaluate changes to complex habitat networks over time. We systematically evaluated the connectivity contributions of multiple climate-adaptive conservation strategies for protected area network development (conserving geodiverse lands and conserving lands near already established conservation areas) against a no additional conservation landscape baseline, assessing whether the benefits of aggressive conservation action will be robust to climate and land-use change effects. This spatiotemporally dynamic assessment of connectivity change represents a promising new approach for evaluating the performance of conservation actions under uncertainty, thereby supporting the identification of conservation approaches that are more likely to achieve targeted biodiversity outcomes under global change.

Study area

We focused our study on a 1.5-million-hectare landscape located in central North Carolina, USA (Fig. 1). The climate of this region is humid subtropical with hot, humid summers (July mean temperature: 25°C) and mild winters (January mean temperature: 3.3°C). There is no distinct wet and dry season and mean annual rainfall ranges between 1016–1398 mm (https://www.ncdc.noaa.gov/cag/divisional/mapping). The study area contains three ecoregions: the Atlantic Coastal Plain, the Sandhills, and the Piedmont, each with a combination of longleaf pine (Pinus palustris Mill.) forests, loblolly pine (Pinus taeda L.) plantations, and pine and hardwood mixed forest types and distinct understory plant communities (Jose et al. 2007). Forests cover over half of the study area with urban, suburban, and agricultural land-uses also present throughout. This region is experiencing increasing urbanization pressures as urban cores such as Charlotte and Raleigh/Durham expand (Terando et al. 2014). Climate change is also expected to play a substantial role in driving landscape change, altering the feasibility and timing of management approaches like prescribed burning and potentially shifting forest type composition (Mitchell et al. 2014). The study area has been the center of multi-agency partnerships dedicated to enhancing connectivity for the benefit of threatened and endangered species and to promote longleaf pine restoration (Costanza et al. 2020). These partnerships have been incredibly active in the acquisition, restoration, and management of land for conservation in the study area.

Landscape change modeling

We used LANDIS-II (Scheller et al. 2007), a stochastic, spatially explicit landscape change model, to simulate land acquisition and restoration for protected area network development across the study area from 2020 to 2100 under different global change futures. We forecasted landscape-scale forest change under climate change, land-use change, and conservation action implementation, enabling us to track changes to complex habitat networks over time as existing and restored forests matured for a dynamic assessment of potential connectivity. LANDIS-II simulates forest change as a function of growth and succession via competition for resources (e.g., soil moisture, nitrogen, light) among tree and shrub species age-specific cohorts that experience disturbances and undergo management actions on a landscape of gridded cells, integrating both anthropogenic change and environmental variability (Scheller et al. 2007). LANDIS-II uses species-specific life history traits (e.g., longevity, age of sexual maturity, shade tolerance) to determine landscape and site-level demographics including vegetation regeneration, growth, mortality, and dispersal (Scheller et al. 2011b). Each 1-hectare cell on the landscape was assigned soil characteristics (https://websoilsurvey.nrcs.usda.gov/), climate conditions (Abatzoglou 2013), and disturbance (tropical storm events, Schrum et al. 2020) and management regimes (e.g., forest harvest and thinning practices) obtained from local managers and forestry extension specialists. We based initial landscape conditions on classifications of the landscape as forest stands and other land-uses derived from the 2019 National Landcover Database (Dewitz 2019). We assigned forest stand structure to each forested cell using data imputed from the US Forest Service’s Forest Inventory and Analysis program (Bechtold & Patterson 2015). For a detailed description of LANDIS-II parameterization for the study area, see the Supplementary Information.

To capture current forest management practices and conservation in the study area we simulated three harvest prescriptions and three restoration prescriptions (Table 1). Cells on the landscape were grouped by land ownership type (Dewitz 2019) then divided into discrete “stands” on which management actions were applied. Stands were delineated using an extensive network of roads (https://www.nconemap.gov/datasets/NCDOT::ncdot-road-characteristics) and streams (https://data-nconemap.opendata.arcgis.com/datasets/major-hydrography-streams-rivers/explore). Harvest management prescriptions included clear cutting and replanting loblolly pine stands, targeted thinning in longleaf pine stands emulating the prescribed burning common in natural resource areas throughout the Sandhills ecoregion, and mixed forest thinning. Restoration prescriptions consisting of longleaf pine restoration, mixed pine restoration, and mixed hardwood restoration, were simulated for every conservation strategy (Table 1). In addition to harvest and management, we simulated tropical storm disturbance using the hurricane module parameterized by Schrum et al. (2020).

Table 1

Harvest and restoration prescriptions simulated on the study landscape in central North Carolina.
Prescription	Actions taken	Targets per 3 year harvest rotation
Loblolly clearcut	Complete clearcut of loblolly pine plantation stands followed by replanting of loblolly pine.	4% of private family land, 10% of private corporate land
Longleaf thin/burn	Removal of most of young, fire intolerant hardwood trees with some removal of oak species and non-longleaf pines to simulate prescribed burning.	18% of state land, 2% of private family land, 20% of federal land, and 12% of conservation lands
Mixed forest thinning	Remove half of all cohorts juvenile and mature cohorts in these stands.	9% of private family land
Longleaf restoration	Complete removal of all major hardwood spp. competitors, moderate removal of non-longleaf pine species, replant with longleaf pine.	50% of protected area hectares
Pine mix restoration	Removal of most hardwood species, removal of half of loblolly pine, replant with Virginia, shortleaf, and longleaf pine.	25% of protected area hectares
Hardwood mix restoration	Complete removal of longleaf pine, removal of half of all other species, replant with white oak.	25% of protected area hectares

We used global climate model (GCM) projections derived from the Coupled Model Intercomparison Project Phase 5 Multivariate Adaptive Constructed Analogs (Abatzoglou 2013). From this ensemble, we chose outputs from five GCMs for the Representative Concentration Pathways (RCP) 4.5 and 8.5 that bracketed the temperature and precipitation projections for the study area and served as our model replicates to capture a range of possible future conditions in the study area: bcc-csm1, CNRM-CM5, Had-GEM2, IPSL-CM5A-LR, and Nor-ESM1-M (Grey 2018). These GCMs were selected to encompass a range of temperature and precipitation changes, from warm to hottest and wettest to driest. To understand the influence of land-use change on landscape connectivity, we used two shared socioeconomic pathway (SSP) scenarios from the EPA’s Integrated Climate and Land-Use Scenarios (ICLUS) v2.1.1 (EPA 2016) for spatially explicit projections of land-use change. SSP2 represents a moderate land-use change scenario in which population growth is moderate and plateaus in the second half of the century. SSP5 represents a scenario in which strong market forces and technological progress support both economic and social development leading to greater population growth and urban expansion.

Conservation scenarios

We simulated two protected area network development strategies designed to support global change-adaptive conservation: i) focusing on the acquisition and restoration of lands near existing protected areas (hereafter referred to as the cluster strategy) and ii) emphasizing the acquisition and restoration of areas exhibiting high geodiversity characteristics (hereafter, the geodiversity strategy). Each strategy simulated an aggressive rate of conservation, acquiring and restoring 1% of the study area every five years for a total of 17% of the landscape over 80 years. When added to the existing managed and protected areas in the study area, this rate of acquisition approximates the 30% global terrestrial conservation target established by the Global Deal for Nature (Dinerstein et al. 2019). Areas designated as open water or as high, medium, or low density residential in the 2016 NLCD (Dewitz 2019) were not considered for acquisition, nor were lands designated as already managed or protected, identified using USGS’s (2018) Protected Areas database. Land acquired and subsequently restored for conservation in model simulations remained in conservation status in perpetuity and was not subject to land-use changes.

Our ‘cluster strategy’ simulated the creation new protected areas around land already managed for conservation, creating larger core areas of habitat (Herse et al. 2018). We created buffers at 100, 500, 1000, 2500 meters around established protected areas identified using the US Geological Survey’s Protected Areas database (USGS). Stands whose centers fell within the 100m buffer had the highest probability of being selected for conservation under this strategy, followed by stands that had centers within the 500m buffer, then 1000m, then 2500m using a weighted random sort (see Supplementary Information). Stands with centers that fell outside of these buffers had the lowest probability of selection.

Our ‘geodiversity strategy’ preferentially selected stands with the highest geodiversity attributes, those that possess a diversity of abiotic conditions, for acquisition and restoration (Grey 2008; Lawler et al. 2015). We used zonal statistics to assign an average geodiversity score to each stand using geodiversity data from the Nature Conservancy’s resilient and connected landscapes map (Anderson et al. 2016). We then performed a weighted random sort with stands scoring the highest in geodiversity having the highest probability of being selected (see Supplementary Information).

We compared these conservation strategies to a baseline (no additional conservation) strategy for an evaluation of how conservation influenced forest habitat structure and improved connectivity above a landscape baseline under global change. In total, we simulated five replicates each of four climate and land-use change scenarios (RCP 4.5 and static land-use, RCP 8.5 and static land-use, RCP 4.5 and SSP2 land-use change, RCP 8.5 and SSP5 land-use change) with and without conservation for a total of 60 simulations (4 global change scenarios × 3 conservation scenarios × 5 replicates).

Connectivity analysis

We used graph theory analysis to quantify the influence of conservation strategy selection and anthropogenic change on potential landscape connectivity over time. In graph theory, the landscape is represented as a set of discrete habitat patches, or nodes, connected by links that characterize the ability of an organism to disperse between nodes (Calabrese & Fagan 2004). Because it incorporates both landscape structure and biologically informed movement information into indices of connectivity, graph theory is well-suited for assessing landscape-level trends over time (Galpern et al. 2011; Tulbure et al. 2014). Connectivity was evaluated for a functional species guild of moderate dispersal (5000 m maximum dispersal distance) habitat specialists that preferred above intermediate stand age (> 21-year median stand age) longleaf pine forest habitat. By simulating a theoretical species with a maximum dispersal distance representing a range of common small mammals and reptiles from the southeastern US (Sutherland et al. 2000; Smith & Green 2005), we focused on enhancing landscape connectedness for functional capacity (Barnosky et al. 2017).

We quantified connectivity for the study area by constructing habitat networks from LANDIS-II model outputs based on least cost path resistance distances between all pairs of habitat nodes within 5000 meters pair-wise Euclidean distance (Fig. 2). Resistance values represent the influence of a heterogeneous landscape matrix on movement and assumed that forest species face greater movement difficulty as they move through land cover types increasingly disparate from the forested areas where they reside (Saura et al. 2011). We used a rating scale of 1-100 for resistance values, consistent with the scale of resistances found in previous studies (e.g. Greenwald et al. 2009; Shirk et al. 2015; Blazquez-Cabrera et al. 2016), with 1 representing the lowest cost to movement (e.g. mature longleaf pine forest) and 100 representing the highest (e.g. an urban core). Least-cost paths were calculated using the gdistance R package (van Etten 2017) between the centroids of each pair of nodes and reflect the minimum possible cost accumulated between them (Bishop-Taylor et al. 2015). Because gdistance calculates least-cost paths using conductance instead of resistance surfaces, every cell on the landscape was assigned a conductance value (1/resistance) based on land/forest cover type. By using habitat patch centroids, we were able to simulate species movement between and within habitat patches rather than assuming patch homogenization or an immediate end to species movement at a patch edge (Dickson et al. 2017). We considered a cell on the landscape to be habitat if at least 25% of total biomass was longleaf pine and the median stand age was ≥ 21.

To identify habitat nodes and network changes in response to climate change, land-use change, and conservation, we grouped contiguous habitat cells that co-occurred in both time steps t and t-1 discretely from those that only occurred in time step t for each time step and replicate using an eight-neighbor rule (see Supplementary Information). Doing so allowed us to capture habitat asynchrony and more realistically represent how an organism would perceive spatiotemporal habitat fluctuations, thereby better capturing intra- and inter-patch movement dynamics (Hanski 1999; Wimberly 2006).

We used the Equivalent Connected Area (ECA) index to quantify landscape-level trends in potential connectivity over time across each scenario of global change with and without conservation intervention (Saura et al. 2011; Tulbure et al. 2014). ECA measures the size of a single habitat patch, maximally connected, that if added to the landscape would provide the same probability of connectivity as the existing habitat network (Saura et al. 2011). ECA measures connectivity by accounting for both the habitat area within each node and the habitat made available through dispersal, incorporating the connectivity contribution of nodes and links that serve as stepping-stones between other habitat areas (Saura & Rubio 2010). In this approach, the landscape is represented as a set of discrete habitat patches, or nodes, connected by links (Least Cost Paths, in our analysis) that characterize species dispersal ability (Calabrese & Fagan 2004). ECA indices use area units (e.g. hectares) enabling a direct comparison between changes in total habitat area and network spatial configuration to assess how the addition or loss of habitat, whether from land acquisition or landscape dynamics, affects connectivity (Bishop-Taylor et al. 2018; McIntyre et al. 2018). In our connectivity analysis, nodes were weighted by area and the probability of movement between nodes decayed exponentially as inter-patch least cost path value increased.

We also quantified connectivity for three ECA subcomponents: (i) intra-patch connectivity (ECA_intra), which describes the connectivity contributed by the area within each habitat node; (ii) the connectivity contributed by direct connections between neighboring habitat nodes (ECA_direct), and (iii) the connectivity contributed by intermediate stepping-stones that enabled longer distance dispersal (ECA_step) (Saura and Rubio 2010, Saura et al. 2011, Bishop-Taylor et al. 2018). All indicators were calculated at 10-year time steps using the probabilistic formulation of ECA.

LANDIS-II outputs were processed using the raster library in the statistical software package R (v3.4.3) (Hijmans and van Etten 2012, R Development Core Team 2017). To conduct our connectivity analysis, we used the gDistance (van Etten 2017), spatstat (Baddeley & Turner 2005), igraph (Csardi & Nepusz 2006), and Conefor command line (v2.6- Saura & Torne 2009) R packages. We used a high-performance computing cluster and over 10,000 computing hours to process connectivity indices.

Nodes and links

Habitat networks varied substantially in both node size and distribution between scenarios and across the conservation strategies evaluated (for example, see the variance in network configurations shown in Fig. 3), with greater variation under RCP 8.5 than RCP 4.5. Across all scenarios, the geodiversity strategy facilitated both the greatest number of habitat nodes and the greatest number of links between nodes, followed by the cluster strategy then the no additional conservation landscape baseline (Fig. 4). After 80 years, the geodiversity strategy resulted in a 66% increase in habitat nodes without LUC and a 60% increase with LUC compared to the respective baseline scenarios under RCP 4.5 (Fig. 4a). The cluster strategy yielded a 29% improvement without LUC and a 33% improvement with LUC compared to the respective landscape baselines (Fig. 4a). Conservation yielded a greater increase in available habitat nodes compared to landscape baseline under RCP 8.5 than RCP 4.5, with an average 44% and 62% increase for the cluster strategy and an average 73% and 95% increase for the geodiversity strategy under the no land-use change and land-use change scenarios, respectively (Fig. 4c). The inclusion of land-use change in our simulations reduced the number of habitat nodes on the landscape by 18–23% by 2100 under RCP 4.5 (Fig. 4a) and by 34–41% under RCP 8.5 (Fig. 4c).

While the average number of habitat nodes and links decreased from RCP 4.5 to RCP 8.5, average node size was larger under RCP 8.5 as habitat networks condensed into fewer, larger patches (Table 2). Total habitat area also increased under RCP 8.5 by 30–35% depending on the scenario, compared to RCP 4.5 (Table 2). The addition of land-use change reduced the number of links between habitat nodes under both RCPs (Fig. 4b and d) with larger losses under the more severe land-use changes associated with RCP 8.5. By 2100, LUC reduced links by an average of 12% without conservation, by 11% for the cluster strategy, and by 26% for the geodiversity strategy under RCP 4.5 compared to no LUC scenarios (Fig. 4b). LUC reduced links by an average of 57% without conservation, by 43% for the cluster strategy, and by 41% for the geodiversity strategy under RCP 8.5 (Fig. 4d).

Equivalent Connected Area

Landscape connectivity was strongly influenced by both climate change and land-use change, though their relative influence differed between scenarios. ECA increased more over time under RCP 8.5 than RCP 4.5 for every scenario even when land-use change was considered (Fig. 5a and 5b), reflecting the proliferation of available habitat as warmer climatic conditions favored longleaf pine and likely reduced heterospecific competition. The influence of climate change was especially apparent in our evaluation of landscape connectivity without additional conservation. ECA was ~ 10,000 ha higher by 2100 under RCP 8.5 compared to RCP 4.5 regardless of whether land-use change was included, revealing the influence of more extreme climate change on increasing habitat availability under RCP 8.5 (Table 2).

ECA was higher under all no land-use change scenarios compared to the respective land-use change scenarios, showing that the inclusion of LUC reduced potential connectivity over time (Table 2). Land-use change had more a more substantial influence on landscape connectivity under RCP 4.5 than RCP 8.5. Under RCP 4.5, LUC reduced final ECA by approximately 23%, 26%, and 25% for the landscape baseline, cluster, and geodiversity strategies respectively (Fig. 5a). Under RCP 8.5, LUC reduced ECA by approximately 14%, 13%, and 5% for the landscape baseline, cluster, and geodiversity strategies respectively (Fig. 5b). Though these findings seem to contradict the greater reduction in available links under RCP 8.5 compared 4.5 when LUC was considered as described in the previous section, they in fact highlight the substantial proliferation of habitat availability that occurred, especially under the warmer conditions of RCP 8.5. While links between habitat nodes declined significantly when LUC was considered, especially under RCP 8.5, the increase in total available habitat on the landscape markedly slowed any LUC-related connectivity reductions.

The conservation strategy employed also influenced ECA, though to a lesser degree than climate or land-use change. Under both RCPs and regardless of whether land-use change was included or not, the cluster strategy resulted in the greatest connectivity improvements compared to the geodiversity and no additional conservation strategies, increasing connectivity ~ 192–276% depending on the scenario (Table 2). However, the contribution of either conservation strategy to landscape connectivity above baseline was reduced under RCP 8.5 compared RCP 4.5, especially for the no LUC scenarios. Without land-use change included under RCP 8.5, the geodiversity strategy failed to result in any connectivity improvements above baseline and the cluster strategy performed similarly to the LUC scenario as habitat gains facilitated by climate change dampened the signal of conservation on the landscape (Table 2). This contradicts the trends in conservation strategy performance under RCP 4.5, where LUC reduced the ability of conservation to facilitate connectivity improvements. Across scenarios, early increases in landscape connectivity reflect the expansion of longleaf pine on the landscape, both through conservation and favorable climatic conditions and characterize habitat networks expanding as additional habitat patches became available elsewhere on the landscape (Fig. 5a and 5b).

Table 2

Change in attributes of connectivity, including number of nodes and links, mean node size, total available habitat, and equivalent connected area (ECA) between 2020–2100.
	Cluster / LUC		Cluster / No LUC		Geodiversity / LUC		Geodiversity / No LUC		No Conservation / LUC		No Conservation / No LUC
	RCP 4.5	RCP 8.5	RCP 4.5	RCP 8.5	RCP 4.5	RCP 8.5	RCP 4.5	RCP 8.5	RCP 4.5	RCP 8.5	RCP 4.5	RCP 8.5
Δ # Nodes		4280	1370	4482	1457	5650	2760	6869	4190	1773	-1476	1935	-359
Δ # Links		697282	72778	574194	-105199	921597	334850	1122206	405510	698353	-210832	626393	84945
Δ Mean node size (ha)		-3.1	-0.6	-1.9	0.6	-3.1	-1.3	-2.1	-0.6	-2.3	1.7	-2.0	1.4
Δ Total available habitat (ha)		68.9	10142.5	14878.9	27618.6	3082.8	8974.9	17531.5	19643.2	-10237.0	-640.6	-7356.7	6113.4
Δ ECA (ha)		23191.8	30087.9	31575.0	35026.9	21261.5	28267.0	28810.7	29695.6	15111.2	25974.4	20540.5	30268.6

ECA_intra, ECA_direct, and ECA_step

Intra-patch connectivity (ECA_intra), the contribution of within-patch movement to overall connectivity, was the most important contributor to connectivity under all LUC scenarios for both RCPs (Figs. 5c and 5d). This reflects the higher cost of movement between patches and the reduction in the available habitat nodes on the landscape under land-use change, making movement between nodes less feasible and heightening the importance of connectivity within the nodes themselves. The ability of habitat patches to serve as stepping-stones (ECA_step) was the largest contributor to landscape connectivity for all no LUC scenarios (Figs. 5g and 5h) as more diffuse movement between patches was possible when land-use change pressures were excluded. The importance of ECA_intra under LUC and ECA_step without LUC was more pronounced under RCP 8.5 (Figs. 5d and 5h), showing that though connectivity increased markedly under more extreme climate change scenarios, how connectivity was enhanced on the landscape differed dramatically. Connectivity between neighboring habitat nodes (ECA _direct) was a minor contributor to landscape connectivity, declining in importance from ~ 25% to ~ 5–15% depending on the scenario as habitat networks expanded (Figs. 5e and 5f).

Global change is introducing significant uncertainty into the conservation of biodiversity (Colloff et al. 2017) while heightening the need for larger scale and more adaptive conservation intervention (Reside et al. 2018). Climate and land-use change are expected to have interactive effects, generating conservation outcomes that differ from those when only a single driver is considered (de Chazal & Rounsevell 2009). The magnitude and direction of these interactions can be highly variable depending on the system, and the mechanisms regulating CC/LUC interactions have significant implications for biodiversity change, and therefore, conservation planning (Schulte to Bühne et al. 2021). Integrating landscape change modeling with network theory, we evaluated changes to complex habitat networks over time by relating spatiotemporal dynamics in habitat availability to the conservation strategy employed and assessing connectivity improvements against a landscape baseline across multiple global change scenarios. Connectivity science has yielded exciting advances in recent years, including the use of time-series analysis to assess the influence of environmental variability and ecosystem dynamics on landscape connectivity (e.g., Bishop-Taylor et al. 2018, McIntyre et al. 2018). The use of bioclimatic variables to forecast habitat location and connectivity under global change has also gained momentum (e.g., Littlefield et al. 2017, Huang et al. 2020). Additionally, emergent work has integrated simulation modeling with network theory to assess the influence of conservation strategy selection on potential connectivity (Mozelewski et al. 2022). Our framework builds upon these advances, leveraging a spatially explicit landscape simulation model to forecast connectivity responses to protected area network expansion integrating both anthropogenic change and environmental variability.

Our results showed that connectivity increased over time across all scenarios, even without the acquisition and restoration of additional land for conservation. This supports past studies that found that a changing climate is likely to have positive impacts on longleaf pine, increasing its dominance in the southeast in part because longleaf pine is tolerant of drought and heat stress (Samuelson et al. 2012; Costanza et al. 2015; Peters et al. 2020). Responses to climate change will be variable and species-specific, with some species emerging as ‘winners’, gaining dominance in their respective ecosystems, while others become “losers”, struggling to adapt and experiencing fitness declines (Dyderski et al. 2017; Hoveka et al. 2022). Our projections suggest that longleaf pine may emerge as one such “winner” with climate change conveying a competitive advantage against hardwood species in the study area. However, while a warmer climate could favor the expansion of longleaf pine forests increasing habitat availability and improving connectivity, it could also make the implementation of prescribed burning, which is currently the principal strategy for maintaining and restoring the longleaf ecosystem, more difficult by reducing the available burn window (Mitchell et al. 2014; Kupfer et al. 2020). Urbanization will likely exacerbate limitations on prescribed burning as the wildland-urban interface expands, potentially decreasing longleaf area (Costanza and Moody 2011). Longleaf pine ecosystems are fire-adapted and prescribed burning increases soil nutrient availability, reduces hardwood competition, and improves seedling establishment (Heuberger & Putz 2003). We see the integration of changing burn windows and locations as an important next step in forecasting future connectivity in the study area, as reductions in prescribed burning will likely reduce longleaf pine habitat quality and availability.

Results comparing our two conservation strategies to a landscape baseline revealed that the cluster strategy performed the best across global change scenarios, though more extreme climate change reduced the differences among strategies. Because existing longleaf pine habitat is primarily concentrated on lands owned or managed by federal, state, and local governments and nongovernmental organizations in the southeastern portion of the study area, the cluster strategy better supported progressive dispersal to newly created habitat by targeting lands adjacent to existing conservation areas. Acquiring and restoring land neighboring existing protected areas could have additional conservation benefits under global change, both reducing fragmentation and increasing habitat core area which may increase population sizes and buffer species from edge effects (Herse et al. 2018), supporting species persistence in the face of global change (Dilts et al. 2016). The ability of the cluster strategy to best enhance connectivity supports earlier work comparing connectivity outcomes from a suite of conservation alternatives (Mozelewski et al. 2022), though those findings did not consider global change. These results can also inform ongoing conversations about the relative advantages conferred by fewer large or more small protected areas (e.g., Ovaskainen 2002; Santini et al. 2016) and how habitat configuration effects connectivity (e.g., Jackson and Fahrig 2016; Arroyo-Rodríguez et al. 2020; Hodgson et al. 2022).

Meanwhile, the geodiversity strategy yielded more variable improvements in connectivity, often resulting in the highest and lowest equivalent connected area scores within any given scenario. While longleaf pine forests are often characterized by deep, sandy soils in the study area, geodiverse portions of the study area were commonly affiliated with more clay-dominant and thinner soil substrates (Gilliam et al. 1993; Peet 2007), which could have influenced restoration success and yielded a larger variation in connectivity improvements. We see value in investigating the performance of this strategy on additional landscapes and for different target habitat types, as its relevance as a coarse filter approach to conservation under global change is often suggested (e.g., Beier et al. 2015).

The expansion of protected areas has been recommended as a key defense against biodiversity and habitat loss under global change (Gray et al. 2016). Simultaneously, maintaining and enhancing connectivity is often cited as core strategy for helping species adapt to such change and for enhancing the ability of protected area networks to support population persistence (Brennan et al. 2022). However, while connectivity between protected areas has improved in recent years, it still falls well short of global targets (Saura et al. 2019). Multiple strategies for protected area network expansion to support connectivity under global change have been suggested (e.g., Beier and Brost 2010; Andrello et al. 2015; Albert et al. 2017; Elsen et al. 2020; Lawler et al. 2020). We see our approach combining landscape simulation modeling with network theory as a way to inform these discussions and evaluate strategies in silico, integrating information on global change-related threats into protected area network design to better protect biodiversity in the future (Carroll and Ray 2021).

Most notably, our study highlights contrasting mechanisms by which global change drivers diminished the ability of conservation to improve connectivity on our study landscape. Under moderate climate change, land-use change was the dominant driver, lowering the connectivity potential of the surrounding landscape, thereby weakening the ability of newly established protected area networks to enhance connectivity. Under more extreme climate change, conditions favored climate change resilient longleaf pine (Clark et al. 2018), increasing the connectivity potential of the surrounding landscape and superseding the ability of conservation to make a notable difference. In fact, rather than acting synergistically, climate change reduced the fragmenting effects of land-use change in the study area. Though both drivers dampened the influence of conservation on landscape connectivity, their implications for conservation planning differ greatly. For example, under RCP 4.5 managers might want to prioritize additional acquisition and restoration of land above a 17% area target to counteract the negative influence of land-use change and yield larger connectivity improvements. Under RCP 8.5, substantial connectivity improvements to the landscape baseline might support prioritization of other conservation concerns. This distinction underscores the importance of evaluating multiple drivers of change across multiple change scenarios in conservation planning and forecasting (Mazor et al. 2018).

Limitations

The modeling framework described here is not intended to predict the exact outcome of any conservation action, but instead to evaluate trends in conservation performance across a range of possible futures (Mozelewski and Scheller 2021). This may help inform conservation decision-making and provide support for managers looking to proactively manage their landscapes for change (Hobbs et al. 2014). Although LANDIS-II has been widely used and rigorously tested (e.g., Scheller et al. 2011a; Lucash et al. 2018; Boulanger et al. 2019; Maxwell et al. 2020), uncertainty in parameter estimates and representation of ecological processes exists along with uncertainty surrounding future land management trajectories and approaches. Further, we focused on a limited set of landscape dynamics, excluding insect disturbance (Sturtevant et al. 2015) and species range expansions into the study area as the climate warms (Van Houtven et al. 2019), which have been shown to shift community composition in the eastern United States (Clark et al. 2021).

Although we evaluated the influences of climate change, land-use change, and conservation on landscape connectivity on a single landscape, we believe our modeling framework is broadly applicable to many landscapes. Evaluation of conservation strategies on additional landscapes and with alternative constraints is warranted. For example, rather than prioritizing individual forest stands with high geodiversity, the geodiversity strategy could prioritize stands that collectively increase the representativeness of geodiversity in the management portfolio. Instead of using Euclidean distance, other functions of distance from existing conservation areas could be employed for the cluster strategy. Such alternative approaches could affect model outcomes. Additionally, we recognize that this study does not address protected areas dynamic boundaries, which may be key for protecting species under climate change (Elsen et al. 2020). We see opportunities to expand this framework to include geographically dynamic reserve networks in the future.

In using least cost path resistance distances in our analysis, we assumed that that forest species face greater dispersal difficulty as they move through land cover types increasingly dissimilar from their preferred habitat (Saura et al. 2011). Yet this approach relies on the simplifying assumptions that a single optimal path between habitat patches conveys the movement potential between those two patches, and that an organism has perfect knowledge of the landscape and will choose a path accordingly (Fahrig 2007; McRae et al. 2008; Pinto & Keitt 2009; Bishop-Taylor et al. 2015). Future analyses could use circuit theory (McRae et al. 2013) as a measure of effective distance to replace these assumptions. Additionally, habitat quality metrics could be integrated into habitat node weighting to further account for variation in habitat condition. Although the species guild parameters we used in this analysis were coarse, species-specific requirements could readily be integrated into our framework (Scheller et al. 2011a) to inform connectivity thresholds.

Acknowledgments

We thank Z. Robbins and C. Littlefield for their early reviews of this manuscript. We acknowledge the computing resources provided on Henry2, a high-performance computing cluster operated by North Carolina State University. We also thank L. Lowe for her assistance with HPC project setup and operationalization, which was provided through the Office of Information Technology HPC services at NC State University.

Funding

This research was funded in part by a U.S. Geological Survey Southeast Climate Adaptation Science Center graduate fellowship awarded to T.G.M. and by the Stainable Forestry Initiative Conservation Grant Program. Additionally, North Carolina State University’s Chancellor’s Faculty Excellence Program provided funding and support for this project.

Competing Interests

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

Author Contributions

Both authors contributed to the study conception and design. Model parameterization and connectivity analyses were performed by Tina Mozelewski with supervision and input from Robert Scheller. The first draft of the manuscript was written by Tina Mozelewski and both authors edited previous versions of the manuscript. Both authors read and approved the final manuscript.

Abatzoglou JT (2013) Development of gridded surface meteorological data for ecological applications and modelling. International Journal of Climatology 33:121–131
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. Conservation Biology 31(6):1383-1396
Anderson MG, Barnett A, Clark M, Prince J, Olivero Sheldon A, Vickery B (2016) Resilient and Connected Landscapes for Terrestrial Conservation. The Nature Conservancy, Eastern Conservation Science, Eastern Regional Office. Boston, MA
Anderson MG, Comer PJ, Beier P, Lawler JJ, Schloss CA, Buttrick S, Albano CM, Faith DP (2015) Case studies of conservation plans that incorporate geodiversity. Conservation Biology 29:680–691
Andrello M, Jacobi MN, Manel S, Thuiller W, Mouillot D (2015) Extending networks of protected areas to optimize connectivity and population growth rate. Ecography 38:273–282
Arroyo‐Rodríguez V, Fahrig L, Tabarelli M, Watling JI, Tischendorf L, Benchimol M, Cazetta E, Faria D, Leal IR, Melo FP, Morante‐Filho JC (2020) Designing optimal human‐modified landscapes for forest biodiversity conservation. Ecology letters (9):1404-20
Asamoah EF, Beaumont LJ. Maina JM (2021) Climate and land-use changes reduce the benefits of terrestrial protected areas. Nature Climate Change 11:1105–1110. https://doi.org/10.1038/s41558-021-01223-2
Baddeley A, Turner R (2005) spatstat: An R package for analyzing spation point patterns. Jornal of Statistical Software 12:1-42.
Barnosky AD et al. (2017) Merging paleobiology with conservation biology to guide the future of terrestrial ecosystems. Science 355:eaah4787
Bechtold WA, Patterson PL (2015) The Enhanced Forest Inventory and Analysis Program National Sampling Design and Estimation Procedures (SRS-GTR-80; p. SRS-GTR-80). U.S. Department of Agriculture, Forest Service, Southern Research Station
Beier P, Brost B (2010) Use of land facets to plan for climate change: conserving the arenas, not the actors. Conservation biology 24(3):701-710
Beier P, de Albuquerque FS (2015) Environmental diversity as a surrogate for species representation. Conservation Biology 29(5):1401-1410
Bergsten A, Bodin Ö, Ecke F (2013) Protected areas in a landscape dominated by logging - A connectivity analysis that integrates varying protection levels with competition-colonization tradeoffs. Biological Conservation 160:279–288
Bishop-Taylor R, Tulbure MG, Broich M (2015) Surface water network structure, landscape resistance to movement and flooding vital for maintaining ecological connectivity across Australia’s largest river basin. Landscape Ecology 30:2045–2065
Bishop-Taylor R, Tulbure MG, Broich M (2018) Impact of hydroclimatic variability on regional-scale landscape connectivity across a dynamic dryland region. Ecological Indicators 94:142–150
Blazquez-Cabrera S, Gastón A, Beier P, Garrote G, Simón MÁ, Saura S (2016) Influence of separating home range and dispersal movements on characterizing corridors and effective distances. Landscape Ecology 31:2355–2366
Boulanger Y, Arseneault D, Boucher Y, Gauthier S, Cyr D, Taylor AR, Price DT, Dupuis S (2019) Climate change will affect the ability of forest management to reduce gaps between current and presettlement forest composition in southeastern Canada. Landscape Ecology 34:159–174
Brennan A, Naidoo R, Greenstreet L, Mehrabi Z, Ramankutty N, Kremen C. Functional connectivity of the world’s protected areas (2022) Science: 376(6597):1101-1104
Calabrese JM, Fagan WF (2004) A comparison-shopper’s guide to connectivity metrics. Frontiers in Ecology and the Environment 2:529-536
Carlson M, Browne D, Callaghan C (2019) Application of land-use simulation to protected area selection for efficient avoidance of biodiversity loss in Canada’s western boreal region. Land Use Policy 82:821–831
Carroll C, Ray JC (2021) Maximizing the effectiveness of national commitments to protected area expansion for conserving biodiversity and ecosystem carbon under climate change. Global Change Biology 27(15):3395-414
Clark JS et al. (2021) Continent-wide tree fecundity driven by indirect climate effects. Nature Communications 12:1242
Clark KE, Chin E, Peterson MN, Lackstrom K, Dow K, Foster M, Cubbage F (2018) Evaluating climate change planning for longleaf pine ecosystems in the Southeast United States. J. Southeast. Assoc. Fish Wildl. Agencies 5:160-168
Colloff MJ, Lavorel S, van Kerkhoff LE, Wyborn CA, Fazey I, Gorddard R, Mace GM, Foden WB, Dunlop M, Prentice IC, Crowley J (2017) Transforming conservation science and practice for a postnormal world. Conservation Biology 31(5):1008-17
Comer PJ, Pressey RL, Hunter ML, Schloss CA, Buttrick SC, Heller NE, Tirpak JM, Faith DP, Cross MS, Shaffer ML (2015) Incorporating geodiversity into conservation decisions. Conservation Biology 29:692–701
Costanza JK, Moody A (2011) Deciding where to burn: stakeholder priorities for prescribed burning of a fire-dependent ecosystem. Ecology and Society 16:14
Costanza JK, Terando AJ (2019) Landscape Connectivity Planning for Adaptation to Future Climate and Land-Use Change. Current Landscape Ecology Reports 4:1–13
Costanza JK, Terando AJ, McKerrow AJ, Collazo JA (2015) Modeling climate change, urbanization, and fire effects on Pinus palustris ecosystems of the southeastern U.S. Journal of Environmental Management 151:186–199
Costanza JK, Watling J, Sutherland R, Belyea C, Dilkina B, Cayton H, Bucklin D, Romañach SS, Haddad NM (2020) Preserving connectivity under climate and land-use change: No one-size-fits-all approach for focal species in similar habitats. Biological Conservation 248:108678
Csardi G, Nepusz T (2006) The igraph software package for complex network research. InterJournal Complex Systems 1695
de Chazal J, Rounsevell MDA (2009) Land-use and climate change within assessments of biodiversity change: A review. Global Environmental Change 19:306–315
Dewitz J (2019) National Land Cover Database (NLCD) 2016 Products: U.S. Geological Survey data release, https://doi.org/10.5066/P96HHBIE
Dickson BG, Albano CM, McRae BH, Anderson JJ, Theobald DM, Zachmann LJ, Sisk TD, Dombeck MP (2017) Informing Strategic Efforts to Expand and Connect Protected Areas Using a Model of Ecological Flow, with Application to the Western United States. Conservation Letters 10:564–571
Dilts TE, Weisberg PJ, Leitner P, Matocq MD, Inman RD, Nussear KE, Esque TC (2016) Multiscale connectivity and graph theory highlight critical areas for conservation under climate change. Ecological Applications 26:1223–1237
Dinerstein E, Vynne C, Sala E, Joshi AR, Fernando S, Lovejoy TE, Mayorga J, Olson D, Asner GP, Baillie JEM, Burgess ND, Burkart K, Noss RF, Zhang YP, Baccini A, Birch T, Hahn N, Joppa LN, Wikramanayake E (2019) A global deal for nature: Guiding principles, milestones, and targets. Science Advances 5: eaaw2869
Dyderski MK, Paź S, Frelich LE, Jagodziński AM (2018) How much does climate change threaten European forest tree species distributions?. Global change biology 24(3):1150-63
Elsen PR, Monahan WB, Dougherty ER, Merenlender AM (2020) Keeping pace with climate change in global terrestrial protected areas. Science advances 6:p.eaay0814
EPA Updates to the Demographic and Spatial Allocation Models to Produce Integrated Climate and Land Use Scenarios (ICLUS) Version 2, EPA/600/R-16/366F (National Center for Environmental Assessment, 2016); http://www.epa.gov/ncea
Fahrig L (2007) Non-optimal animal movement in human-altered landscapes. Functional Ecology 21:1003-1015
Galpern P, Manseau M, Fall A (2011) Patch-based graphs of landscape connectivity: A guide to construction, analysis and application for conservation. Biological Conservation 144:44-55
Gilliam FS, Yurish BM, Goodwin LM (1993) Community Composition of an Old Growth Longleaf Pine Forest: Relationship of Soil Texture. Bulletin of the Torrey Botanical Club 120:287
Gray M (2008) Geodiversity: developing the paradigm. Proceedings of the Geologists’ Association 119:287–298
Gray CL, Hill SLL, Newbold T, Hudson LN, Börger L, Contu S, Hoskins AJ, Ferrier S, Purvis A, Scharlemann JPW (2016) Local biodiversity is higher inside than outside terrestrial protected areas worldwide. Nat. Commun. 7:12306
Greenwald KR, Purrenhage JL, Savage WK (2009) Landcover predicts isolation in Ambystoma salamanders across region and species. Biological Conservation 142:2493–2500
Grey GME (2018) Ensemble creation of downscaled climate projections in the Southeast United States. North Carolina State University. [Unpublished master’s thesis]
Hannah, L (2010) A global conservation system for climate‐change adaptation. Conservation Biology 24(1):70-77
Hanski I (1999) Habitat Connectivity, Habitat Continuity, and Metapopulations in Dynamic Landscapes. Oikos 87:209
Heller NE, Zavaleta ES (2009) Biodiversity management in the face of climate change: A review of 22 years of recommendations. Biological Conservation 142:14–32
Herse MR, With KA, Boyle WA (2018) The importance of core habitat for a threatened species in changing landscapes. Journal of Applied Ecology 55:2241–2252
Heuberger KA, Putz FE (2003) Fire in the Suburbs: Ecological Impacts of Prescribed Fire in Small Remnants of Longleaf Pine (Pinus palustris) Sandhill. Restoration Ecology 11:72–81
Hijmans RJ, van Etten J (2012) raster: Geographic analysis and modeling with raster data . R package version 3.4-5. http://CRAN.R-project.org/package=raster
Hobbs RJ et al. (2014) Managing the whole landscape: historical, hybrid, and novel ecosystems. Frontiers in Ecology and the Environment 12:557–564
Hodgson JA, Randle Z, Shortall CR, Oliver TH (2022) Where and why are species' range shifts hampered by unsuitable landscapes?. Global Change Biology 28:4765–4774
Hoveka LN, van der Bank M, Davies TJ (2022) Winners and losers in a changing climate: how will protected areas conserve red list species under climate change?. Diversity and Distributions 28(4):782-92
Jackson ND, Fahrig L (2016) Habitat amount, not habitat configuration, best predicts population genetic structure in fragmented landscapes. Landscape Ecol 31:951–968. https://doi.org/10.1007/s10980-015-0313-2
Jennings MK, Zeller KA, Lewison RL (2021) Dynamic Landscape Connectivity Special Issue Editorial. Land 10(6):555. https://doi.org/10.3390/land10060555
Jose S, Jokela EJ, Miller DL (2007) The longleaf pine ecosystem: ecology, silviculture, and restoration. Springer, New York
Kupfer JA, Terando AJ, Gao P, Teske C, Hiers JK (2020) Climate change projected to reduce prescribed burning opportunities in the south-eastern United States. International Journal of Wildland Fire 29:764
Lawler JJ, Ackerly DD, Albano CM, Anderson MG, Dobrowski SZ, Gill JL, Heller NE, Pressey RL, Sanderson EW, Weiss SB (2015) The theory behind, and the challenges of, conserving nature’s stage in a time of rapid change. Conservation Biology 29:618–629
Lawler JJ, Rinnan DS, Michalak JL, Withey JC, Randels CR, Possingham HP (2020) Planning for climate change through additions to a national protected area network: implications for cost and configuration. Philosophical Transactions of the Royal Society B. 375(1794):20190117.
Le Saout S et al. (2013) Protected areas and effective biodiversity conservation. Science 342:803-805
Littlefield CE, McRae BH, Michalak JL, Lawler JJ, Carroll C (2017) Connecting today’s climates to future climate analogs to facilitate movement of species under climate change. Conservation Biology 31:1397–1408
Lucash MS, Scheller RM, Sturtevant BR, Gustafson EJ, Kretchun AM, Foster JR (2018) More than the sum of its parts: how disturbance interactions shape forest dynamics under climate change. Ecosphere 9:e02293
Mantyka-Pringle CS, Visconti P, Di Marco M, Martin TG, Rondinini C, Rhodes JR (2015) Climate change modifies risk of global biodiversity loss due to land-cover change. Biological Conservation 187:103–111
Martensen AC, Saura S, Fortin M-J (2017) Spatio-temporal connectivity: assessing the amount of reachable habitat in dynamic landscapes. Methods in Ecology and Evolution 8:1253-1264
Mawdsley J (2011) Design of conservation strategies for climate adaptation. Wiley Interdisciplinary Reviews: Climate Change 2(4):498-515
Maxwell CJ, Serra‐Diaz JM, Scheller RM, Thompson JR (2020) Co‐designed management scenarios shape the responses of seasonally dry forests to changing climate and fire regimes. Journal of Applied Ecology 57:1328–1340
Mazor T, Doropoulos C, Schwarzmueller F, Gladish DW, Kumaran N, Merkel K, Di Marco M, Gagic V (2018) Global mismatch of policy and research on drivers of biodiversity loss. Nature Ecology & Evolution 2(7):1071-1074
McIntyre NE, Collins SD, Heintzman LJ, Starr SM, van Gestel N (2018) The challenge of assaying landscape connectivity in a changing world: A 27-year case study in the southern Great Plains (USA) playa network. Ecological Indicators 91:607–616
McRae BH, Dickson BG, Keitt TH, Shah VB (2008) Using circuit theory to model connectivity in ecology, evolution, and conservation. Ecology 89:2712–2724
McRae BH, Shah VB, Mohapatra TK (2013) Circuitscape 4 User Guide. The Nature Conservancy. http://www.circuitscape.org
Mitchell RJ, Liu Y, O’Brien JJ, Elliott KJ, Starr G, Miniat CF, Hiers JK (2014) Future climate and fire interactions in the southeastern region of the United States. Forest Ecology and Management 327:316–326
Mozelewski TG, Robbins ZJ, Scheller RM (2022) Forecasting the influence of conservation strategies on landscape connectivity. Conservation Biology 36(5):e13904
Mozelewski TG, Scheller RM (2021) Forecasting for intended consequences. Conservation Science and Practice 3:e370
Nuñez TA, Lawler JJ, McRae BH, Pierce DJ, Krosby MB, Kavanagh DM, Singleton PH, Tewksbury JJ (2013) Connectivity Planning to Address Climate Change. Conservation Biology 27:407–416
Oliver TH, Morecroft MD (2014) Interactions between climate change and land use change on biodiversity: attribution problems, risks, and opportunities. Wiley Interdisciplinary Reviews: Climate Change 5:317–335
Ovaskainen O (2002) Long-term persistence of species and the SLOSS problem. Journal of Theoretical Biology 218:419–433
Peet RK (2007) Ecological classification of longleaf pine woodlands. In The longleaf pine ecosystem (pp. 51-93). Springer, New York, NY
Perring MP, Standish RJ, Price JN, Craig MD, Erickson TE, Ruthrof KX, Whiteley AS, Valentine LE, Hobbs RJ (2015) Advances in restoration ecology: Rising to the challenges of the coming decades. Ecosphere 6:1-25
Peters MP, Prasad AM, Matthews SN, Iverson LR (2020) Climate change tree atlas, Version 4. U.S. Forest Service, Northern Research Station and Northern Institute of Applied Climate Science, Delaware, OH. https://www.fs.usda.gov/nrs/atlas
Pinto N, Keitt TH (2009) Beyond the least-cost path: Evaluating corridor redundancy using a graph-theoretic approach. Landscape Ecology 24:253–266
R Core Team. R: A Language and Environment for Statistical Computing. (R Foundation for Statistical Computing, 2017)
Reside AE, Butt N, Adams VM (2018) Adapting systematic conservation planning for climate change. Biodiversity & Conservation 27:1-29
Samuelson LJ, Stokes TA, Johnsen KH (2012) Ecophysiological comparison of 50-year-old longleaf pine, slash pine and loblolly pine. Forest Ecology and Management 274:108–115
Santini L, Saura S, Rondinini C (2016) Connectivity of the global network of protected areas. Diversity and Distributions 22:199–211
Saura S, Estreguil C, Mouton C, Rodríguez-Freire M (2011) Network analysis to assess landscape connectivity trends: Application to European forests (1990-2000). Ecological Indicators 11:407–416
Saura S, Rubio L (2010) A common currency for the different ways in which patches and links can contribute to habitat availability and connectivity in the landscape. Ecography 33:523-537
Saura S, Torne J (2009) Conefor Sensinode 2.2: a software package for quantifying the importance of habitat patches for landscape connectivity. Environment Modeling & Software 24:135-139.
Scheller RM, Domingo JB, Sturtevant BR, Williams JS, Rudy A, Gustafson EJ, Mladenoff DJ (2007) Design, development, and application of LANDIS-II, a spatial landscape simulation model with flexible temporal and spatial resolution. Ecological Modelling 201:409–419
Scheller RM, Spencer WD, Rustigian-Romsos H, Syphard AD, Ward BC, Strittholt JR (2011a) Using stochastic simulation to evaluate competing risks of wildfires and fuels management on an isolated forest carnivore. Landscape Ecology 26:1491-1504
Scheller RM, van Tuyl S, Clark KL, Hom J, La Puma I (2011b) Carbon sequestration in the New Jersey pine barrens under different scenarios of fire management. Ecosystems 14:987–1004
Schrum P, Scheller RM, Duveneck MJ, Lucash MS (2020) Base-Hurricane: A new extension for the Landis-II forest landscape model. Environmental Modelling and Software 133:104833
Schulte to Bühne H, Tobias JA, Durant SM, Pettorelli N (2021) Improving Predictions of Climate Change–Land Use Change Interactions. Trends in Ecology & Evolution 36: 29-38. https://doi.org/10.1016/j.tree.2020.08.019
Shirk AJ, Schroeder MA, Robb LA, Cushman SA (2015) Empirical validation of landscape resistance models: insights from the Greater Sage-Grouse (Centrocercus urophasianus). Landscape Ecology 30:1837–1850
Smith AM, M. Green D (2005) Dispersal and the metapopulation paradigm in amphibian ecology and conservation: are all amphibian populations metapopulations? Ecography 28:110–128
Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Soil Survey Geographic (SSURGO) Database. Available online at https://sdmdataaccess.sc.egov.usda.gov
Sturtevant BR, Cooke BJ, Kneeshaw DD, Mac Lean DA (2015) Modeling insect disturbance across forested landscapes: Insights from the spruce budworm. In: Perera A, Sturtevant B, Buse L. (eds) Simulation Modeling of Forest Landscape Disturbances. Springer, Cham
Sutherland GD, Harestad AS, Price K, Lertzman KP (2000) Scaling of natal dispersal distances in terrestrial birds and mammals. Conservation Ecology 4:1-31
Taylor PD, Fahrig L, Henein K (1993) Connectivity Is a Vital Element of Landscape Structure. Oikos 68:57-573
Terando AJ, Costanza J, Belyea C, Dunn RR, McKerrow A, Collazo JA (2014) The Southern Megalopolis: Using the Past to Predict the Future of Urban Sprawl in the Southeast U.S. PLoS ONE 9:e102261
Tulbure MG, Kininmonth S, Broich M (2014) Spatiotemporal dynamics of surface water networks across a global biodiversity hotspot - Implications for conservation. Environmental Research Letters 9:114012
U.S. Geological Survey (USGS) Gap Analysis Project (GAP) (2018) Protected Areas Database of the United States (PAD-US): U.S. Geological Survey data release, https://doi.org/10.5066/P955KPLE.
van Etten J (2017) R package gdistance: Distances and routes on geographical grids. Journal of Statistical Software 76:1–21
Van Houtven G, Phelan J, Clark C, Sabo RD, Buckley J, Thomas RQ, Horn K, LeDuc SD (2019) Nitrogen deposition and climate change effects on tree species composition and ecosystem services for a forest cohort. Ecological Monographs 89:e01345
Wimberly MC (2006) Species dynamics in disturbed landscapes: When does a shifting habitat mosaic enhance connectivity? Landscape Ecology 21:35–46

No competing interests reported.

SupplementaryInformation.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Protected Area Network Contributions to Landscape Connectivity under Global Change

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study area

Landscape change modeling

Conservation scenarios

Connectivity analysis

Results

Nodes and links

Equivalent Connected Area

ECA_intra, ECA_direct, and ECA_step

Discussion

Limitations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Protected Area Network Contributions to Landscape Connectivity under Global Change

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study area

Landscape change modeling

Conservation scenarios

Connectivity analysis

Results

Nodes and links

Equivalent Connected Area

ECAintra, ECAdirect, and ECAstep

Discussion

Limitations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

ECA_intra, ECA_direct, and ECA_step