The study of periphery uniqueness and

balance in ecological networks

Shipeng Xu

College of Agricultural and Biological Sciences, Dali University, Dali 671003, China

[email protected]

https://orcid.org/0009-0002-0921-5318

Abstract:

The study of ecological networks is pivotal in modern conservation biology,

addressing pressing issues such as habitat fragmentation and biodiversity loss,
particularly in geographically complex regions. These networks, comprising
ecological corridors, sources, and nodes, are essential for facilitating species
movement, genetic exchange, and the overall functioning of ecosystems. Traditional
network analysis has primarily focused on central nodes, overlooking the critical role
of peripheral nodes that form the interface with the external environment. Peripheral
nodes, though often less connected, are vital for the network's response to
environmental changes, acting as buffers or filters and influencing the network's
resilience and adaptability. This research introduces the innovative Periphery
Analysis Model (PAM), specifically designed to examine the periphery of ecological
networks. PAM is the first methodological framework dedicated to periphery
analysis, incorporating theories and methodologies from graph theory, complex
network analysis, and landscape ecology. The model is centered around two principal
indices: the Periphery Uniqueness Index (PuI) and the Periphery Balance Index (PbI),
which quantify the distinctive attributes and equilibrium of peripheral nodes within
the ecological network. PAM also includes derived indices for a more nuanced
understanding of the periphery's role and influence. The application of PAM promises
to illuminate the characteristics defining the periphery of ecological networks,
revealing the intricate interdependencies and interactions between peripheral nodes
and the network as a whole. This model is expected to enhance our understanding of
the deeper structural features of ecological networks, offering insights into the
maintenance of biodiversity, ecological connectivity, and ecosystem health. The study
advocates for the integration of PAM in conservation strategies, highlighting its
potential to inform policy and management decisions for preserving ecosystem
stability and resilience in the face of environmental challenges.

Keywords: Ecological Networks, Ecological Networks Periphery, Periphery

Analysis, Periphery uniqueness, Periphery Balance
1. Introduction

Ecological networks(ENs) construction is a crucial approach in modern conservation

biology, addressing challenges like habitat fragmentation and biodiversity loss, partic
ularly in regions with complex geographical environments (Xiao et al., 2022). These
networks, consisting of ecological corridors, sources, and nodes, facilitate species mo
vement, genetic exchange, and overall ecosystem functioning (Geng et al., 2023). Me
thods like Morphological Spatial Pattern Analysis (MSPA) for identifying ecological
sources and the Minimum Cumulative Resistance (MCR) model for determining ecol
ogical corridors are essential tools in this process (Z. Yang et al., 2023; X. Zhou et al.,
2023).MSPA plays a key role in ENs planning by identifying and mapping crucial ha
bitat areas known as "ecological sources," which are vital for maintaining biodiversity
. These sources serve as the building blocks of ecological networks, providing core ar
eas for species to thrive (Chen et al., 2023). The MCR model is another essential met
hod, identifying the most effective routes for ecological corridors that connect differe
nt ecological sources. These corridors are crucial for species movement, allowing the
m to migrate, disperse, and interact across fragmented landscapes (D. Zhou & Song,
2021).Additionally, the gravity model is also employed in ENs planning. This model
assesses the interaction strength between different ecological sources, aiding in the pr
ioritization and optimization of corridors based on the potential for species movement
and interaction (C. Yang et al., 2022).

Research on Ecological Networks (ENs) has primarily concentrated on developing

methods for the construction and subsequent evaluation of these networks, with a
particular emphasis on pinpointing the central nodes that play a crucial role within the
network's framework. The identification of these key nodes is achieved through the
computation of various indices, including node centrality, which measures the
influence of a node within the network; betweenness centrality, which gauges a
node's importance in connecting other nodes; and clustering coefficients, which
indicate the degree of interconnection within a node's neighborhood (Z. Wang, et al.,
2023). These central nodes are deemed to possess a higher level of importance within
the network, exerting a substantial influence on the network's overall stability and
robustness, as they often serve as critical hubs for the flow of information, resources,
and interactions (Li, et al., 2021). Despite the significant progress made in this area,
there exists a notable gap in the literature when it comes to the study of peripheral
nodes within ecological networks. Peripheral nodes, while often overlooked, are an
essential component of the network, as they form the outer layers that interface with
the external environment and facilitate the exchange of materials, energy, and
organisms between the network and its surroundings. The properties and dynamics of
these peripheral nodes are instrumental in shaping the network's response to
environmental changes and stressors, as they can act as buffers or filters, influencing
the resilience and adaptability of the network to various disturbances. Therefore, a
comprehensive understanding of the periphery's characteristics and dynamics is
crucial for gaining insights into the deeper structural features of the entire network, as
they offer an alternative perspective that complements the focus on central nodes. By
examining the peripheral nodes, researchers can uncover patterns and processes that
may not be evident when concentrating solely on the network's core, such as the role
of peripheral nodes in maintaining biodiversity, facilitating ecological connectivity,
and promoting the overall health and functioning of the ecosystem.

In addressing the aforementioned research gaps, our study specifically focuses on the
Ecological Networks Periphery (ENP) as the central object of investigation and
proposes an innovative "Periphery Analysis" model tailored for the examination of
ecological networks. This "Periphery Analysis" model represents a groundbreaking
approach by incorporating a rich tapestry of theories and methodologies from graph
theory, complex network analysis, and landscape ecology. It stands as the inaugural
methodological framework explicitly dedicated to the periphery analysis within the
intricate fabric of complex ecological networks. The "Periphery Analysis" model is
meticulously structured around two principal indices that serve as the cornerstone of
its analytical framework: the Periphery Uniqueness Index and the Periphery Balance
Index. These indices are meticulously crafted to quantify and evaluate the distinctive
attributes of peripheral nodes and their equilibrium within the broader context of the
ecological network. Furthermore, the model expands its analytical prowess through
the inclusion of a suite of derived indices. These derived indices are calculated based
on a nuanced interplay of various network characteristics, thereby offering a more
comprehensive understanding of the periphery's role and influence within the
ecological network. The utilization of the "Periphery Analysis" model is poised to
provide a profound insight into the characteristics that define the periphery of
ecological networks. It will illuminate the nuanced interdependencies and interactions
that exist between peripheral nodes and the rest of the network. Moreover, by delving
into the periphery's attributes, this model will shed light on the structural intricacies
and underlying dynamics that contribute to the overall stability and resilience of the
ecological network.

2. Methods

2.1 Definition and identification of the periphery

The definition of the network periphery is associated with the core-periphery

structure (Elliott et al., 2020). In this structure, the network is divided into two main
parts: the core and the periphery. The periphery typically refers to a collection of
nodes that are connected to the core nodes but have relatively few connections with
each other. In the core-periphery structure, peripheral nodes have the following
characteristics: (1) Connections between peripheral nodes are relatively sparse. (2)
Peripheral nodes are primarily connected to the rest of the network through the core
nodes. (3) The periphery can be seen as the outer part of the network, where the
internal connections of nodes are not as tight as those in the core area. Based on the
core-periphery structure, there are two algorithms that can identify the network
periphery.

2.1.1 Algorithm based on random walks

This method identifies the core-periphery structure by analyzing the behavior of

random walkers. The core-periphery structure is revealed by considering the actions
of random walkers within the network. In this approach, peripheral nodes are
identified by calculating the persistence probability of nodes, which is the probability
that a random walker remains on a particular node for a given amount of time (Rossa
et al., 2013). The algorithm constructs the core-periphery structure through an
iterative process, starting with the most weakly connected nodes and progressively
adding those most likely to belong to the core. The mathematical formula of the
algorithm is as follows:

ai = 1 − (1 − pij)
Ni

Here, ai is the persistence probability of node i, pij is the probability of a connection

existing between nodes i and j, and Ni is the number of neighbors of node i.

2.1.2 Algorithm based on core decomposition

This method identifies the core-periphery structure by iteratively removing the

connections of nodes. In k-shell decomposition, nodes in the network are assigned to
different "shells" based on the "strength" of their connections. Each shell layer
represents a connectivity level of the nodes in the network, with nodes in the outer
layers having fewer connections and those in the inner layers having more
connections (Zhang et al., 2015). This method can reveal the hierarchical structure of
nodes within the network, where core nodes are typically located in the inner layers
and peripheral nodes are in the outer layers. Below is the pseudocode form of the k-
shell decomposition algorithm:
function kShellDecomposition(G):
initialize shellCounter to 1
initialize shells to an empty list

while there exists a node v in G with k-shell equal to

shellCounter:
remove v and all its incident edges from G
for each neighbor u of v:
if u is still present in G:
decrement the k-shell of u by 1
end while

return shells

2.2 Periphery analysis model

Periphery Analysis Model (PAM) is a new ecological network analysis

model proposed by us, focusing on two key metrics: Periphery Uniqueness
(Pu) and Periphery Balance (Pb). Below are the definitions and calculation
methods for these two fundamental measures.

2.2.1 Periphery uniqueness

Periphery uniqueness refers to the extent to which the characteristics of

peripheral nodes in an ecological network differ from those of core nodes.
The Periphery Uniqueness Index (PuI) is the primary quantitative measure
of periphery uniqueness. The Periphery Uniqueness Index (PuI) is further
divided into the Network Periphery Uniqueness Index (N-PuI) and the
Ecologically Weighted Periphery Uniqueness Index (E-PuI). The N-PuI is
the Periphery Uniqueness Index calculated solely based on network
characteristics, while the E-PuI is the Periphery Uniqueness Index obtained
by incorporating weights based on the ecological characteristics of the
peripheral nodes, such as biodiversity indices.

The N-PuI can be expressed as:

1
di − d¯core
∑
N-PuIi =
| Nperipher y | i∈Nperipher y
Where, di is the degree of node i, d¯core is the average degree of core nodes, and
| Nperipher y | is the number of peripheral nodes.

The E-PuI can be expressed as:

we ⋅ fe(i) × di − d¯core
∑
E -PuIi =
e∈E

Where, fe(i) is the quantified value of ecological indicator e for node i, we is the
weight of indicator e, di is the degree of node i, and d¯core is the average degree of
core nodes.

2.2.2 Periphery balance

Periphery balance refers to the extent to which the characteristics of peripheral nodes
in an ecological network are close to or connected with those of core nodes. The
Periphery Balance Index (PbI) is the primary quantitative measure of periphery
balance. The Periphery Balance Index (PbI) is divided into the Network Periphery
Balance Index (N-PbI) and the Ecologically Weighted Periphery Balance Index (E-
PbI). The N-PbI is the Periphery Balance Index calculated solely based on network
characteristics, while the E-PbI is the Periphery Balance Index obtained by
incorporating weights based on the ecological characteristics of the peripheral nodes,
such as the Biodiversity Index. Periphery balance and periphery uniqueness are
relative concepts, but their calculation methods differ.

The N-PbI can be expressed as:

−1
1 1 L
∑ |N | ∑
N-PbIi = lij ×
| Nperipher y | maxi∈Nperipher y ( | N | ∑j∈N lij)
1
i∈Nperipher y j∈N

Where, lij represents the shortest path length from node i to node j, and L is the
maximum possible shortest path length in the network.

The E-PbI can be expressed as:

1 1
∑ ∑
E -PbIi = wS ⋅ Si + wSim ⋅ Simeco(i, j )
| Nperipher y | i∈Nperipher y
| Ncore | j∈N
core
Where, Si represents the number of shortest paths between node i and core nodes,
Simeco(i, j ) represents the similarity of ecological characteristics between node i and
core node j, and wS and wSim are the weights for the number of shortest paths and
ecological characteristic similarity, respectively.

2.3 Comprehensive analysis

After identifying the periphery of ecological networks using specific algorithms,

further analysis of the ecological networks is conducted using the periphery analysis
model. The four indices N-PuI, E-PuI, N-PbI, and E-PbI, collectively provide a
comprehensive framework for analyzing the structural and ecological characteristics
of nodes within ecological networks. These indices have significant practical
implications for understanding the complex dynamics of ecological networks. They
can help identify critical nodes that may be more vulnerable to environmental
changes or disruptions, prioritize areas for conservation efforts, and inform strategies
for managing and preserving ecosystem health. When combined, these indices offer a
multifaceted approach to ecological network analysis. For instance, a node with high
N-PuI and low N-PbI might be a unique but isolated species that could be at risk if
not properly conserved. Conversely, a node with high E-PuI and E-PbI might
represent an ecologically significant species that plays a crucial role in maintaining
the integrity of the ecosystem. By considering both network and ecological
perspectives, researchers can develop more targeted and effective conservation
strategies that take into account the complex interplay of biotic and abiotic factors.

3. Discussion

3.1 The ecological significance of peripheral nodes

Peripheral Nodes in ecological networks refer to those nodes that are

connected to the core nodes of the network but do not occupy a central
position within the network. These nodes typically have a lower degree of
connectivity, yet they play an indispensable role in the overall structure and
function of the network. The presence of peripheral nodes not only increases
the redundancy and robustness of the network but also significantly impacts
the transmission of information and the flow of energy within the network.
Peripheral nodes play a critical role in maintaining the stability of ecological
networks. As they are usually connected to the core nodes, peripheral nodes
can act as a buffer for the transmission of information and the flow of
energy. In the face of external disturbances or internal changes, peripheral
nodes can slow down the propagation of shockwaves through their lower
connectivity, thereby providing a certain protective mechanism for the
network. Moreover, the existence of peripheral nodes also helps to disperse
risks within the network, preventing the failure of a single node from
causing the collapse of the entire network. The diversity of an ecological
network is an important indicator of the health and stability of an ecosystem.
Peripheral nodes facilitate the exchange of biodiversity and the flow of
genes by connecting different ecological regions or species communities.
This exchange of diversity not only aids species in adapting to
environmental changes but also holds significant importance for the long-
term evolution and adaptation of ecosystems. The existence of peripheral
nodes makes ecological networks more complex and dynamic, thereby
enhancing the ecosystem's adaptive capacity and resilience to environmental
changes. Furthermore, peripheral nodes play a key role in the evolution of
ecological networks. Due to their freedom and flexibility, peripheral nodes
can experiment with new behaviors and strategies without being constrained
or influenced by the core nodes. In this way, peripheral nodes can bring new
information and opportunities to the network, thus promoting its evolution
and innovation.

3.2 Comparison of peripheral identification algorithms

In network analysis, identifying the core-periphery structure is crucial for

understanding complex systems. Two key algorithms for this task are the
random walk-based and core decomposition-based methods. The random
walk-based approach uses the behavior of a random walker to assign
coreness values to nodes, revealing the network's global structure and the
role of each node. This method is versatile for various network types,
including directed and weighted networks, and can highlight nodes that
bridge or are peripheral to core activities. In contrast, the core
decomposition-based algorithm targets directed networks, partitioning them
into core and periphery based on edge direction and block model
formulations. This approach is adept at detecting complex core-periphery
structures, especially in networks where the direction of interactions is
pivotal. However, it may be less adaptable to weighted networks and can be
computationally intensive. Overall, the random walk-based method excels in
providing a continuous coreness measure, while the core decomposition-
based method offers detailed insights into directed network structures. The
selection of an algorithm should be based on the network's characteristics
and the research objectives. Future research could integrate both methods to
enhance the understanding of network structures.
3.3 Advantages and limitations of the periphery analysis model

The Periphery Analysis Model (PAM) offers several advantages in the study of
ecological networks. Firstly, it provides a focused examination of peripheral nodes,
which are often neglected in traditional network analysis. By quantifying periphery
uniqueness and balance, PAM allows for a more nuanced understanding of the roles
these nodes play in network stability and ecological connectivity. Secondly, the
model's incorporation of both network characteristics and ecological weights provides
a comprehensive perspective that integrates topological and ecological insights. This
dual focus can lead to more informed conservation strategies that consider the
complex interplay between species and their environment. However, PAM also has its
limitations. The model's reliance on specific algorithms for periphery identification
means that its effectiveness is contingent on the accuracy of these algorithms.
Additionally, the calculation of derived indices can be complex and resource-
intensive, particularly for large and complex networks. Furthermore, PAM may
struggle to capture the full dynamics of peripheral nodes, especially in networks with
high temporal variability or those subject to frequent disturbances.

3.4 Outlook for priphery research in networks

Future research on network peripheries should aim to refine and expand upon the
PAM. One direction is to develop more efficient algorithms for periphery
identification that can handle dynamic networks and provide real-time insights.
Another is to explore the integration of additional ecological data, such as species
interactions and environmental conditions, to provide a more holistic view of
peripheral nodes' roles. Additionally, research could focus on the application of PAM
in various ecological contexts, from urban ecosystems to wilderness areas, to assess
its versatility and applicability across different environmental settings. Moreover, the
integration of PAM with other disciplines such as urban planning and public health
could lead to innovative solutions for managing urban green spaces, mitigating the
impacts of climate change, and enhancing the well-being of urban populations. For
instance, identifying peripheral nodes in urban ecosystems could inform the creation
of green corridors that not only support biodiversity but also improve air quality and
provide recreational spaces for residents. Furthermore, the expansion of PAM to
global and regional scales could contribute to the study of large-scale ecological
patterns and the assessment of environmental changes on a broader level. By
incorporating satellite imagery and geospatial data, researchers can monitor the
dynamics of peripheral nodes over time and across different ecosystems, providing
valuable information for international conservation efforts and environmental
agreements.
4. Conclusions

In conclusion, the PAM emerges as a pivotal tool in the study of ecological networks,
offering a novel perspective on the role and significance of peripheral nodes. By
focusing on the uniqueness and balance of these nodes, PAM provides a
comprehensive framework that integrates network topology with ecological
attributes, leading to a deeper understanding of the structural and functional aspects
of ecological networks. The model's application extends beyond purely ecological
contexts, showing promise in interdisciplinary fields such as ecological-social
systems, urban planning, and even information technology. As we move forward, the
continued development and refinement of PAM will be essential in addressing the
challenges posed by environmental changes and in promoting the sustainable
management of ecosystems. The model's potential to inform policy and conservation
strategies, as well as its ability to contribute to the broader goal of ecological
sustainability, makes it a valuable asset in the arsenal of tools available to researchers
and practitioners alike.

Reference:
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. Ecological Indicators, 146, 109809. https://doi.org/
10.1016/j.ecolind.2022.109809

Elliott, A., Chiu, A., Bazzi, M., Reinert, G., & Cucuringu, M. (2020). Core–periphery structure in
directed networks. Proceedings of the Royal Society A: Mathematical, Physical and Engineering
Sciences, 476(2241), 20190783. https://doi.org/10.1098/rspa.2019.0783

Geng, J., Yu, K., Sun, M., Xie, Z., Huang, R., Wang, Y., Zhao, Q., & Liu, J. (2023).
Construction and Optimisation of Ecological Networks in High-Density Central Urban
Areas: The Case of Fuzhou City, China. Remote Sensing, 15(24), 5666. https://doi.org/
10.3390/rs15245666

Li, Y.-Y., Zhang, Y.-Z., Jiang, Z.-Y., Guo, C.-X., Zhao, M.-Y., Yang, Z.-G., Guo, M.-Y.,
Wu, B.-Y., & Chen, Q.-L. (2021). Integrating morphological spatial pattern analysis and
the minimal cumulative resistance model to optimize urban ecological networks: A case
study in Shenzhen City, China. Ecological Processes, 10(1), 63. https://doi.org/10.1186/
s13717-021-00332-2

Rossa, F. D., Dercole, F., & Piccardi, C. (2013). Profiling core-periphery network structure by
random walkers. Scientific Reports, 3(1), 1467. https://doi.org/10.1038/srep01467
Wang, Z., Shi, Z., Huo, J., Zhu, W., Yan, Y., & Ding, N. (2023). Construction and
Optimization of an Ecological Network in Funiu Mountain Area Based on MSPA and
MCR Models, China. Land, 12(8), 1529. https://doi.org/10.3390/land12081529

Xiao, H., Guo, Y., Wang, Y., Xu, Y., & Liu, D. (2022). Evaluation and Construction of
Regional Ecological Network Based on Multi-Objective Optimization: A Perspective of
Mountains–Rivers–Forests–Farmlands–Lakes–Grasslands Life Community Concept in
China. Applied Sciences, 12(19), 9600. https://doi.org/10.3390/app12199600

Yang, C., Guo, H., Huang, X., Wang, Y., Li, X., & Cui, X. (2022). Ecological Network
Construction of a National Park Based on MSPA and MCR Models: An Example of the
Proposed National Parks of “Ailaoshan-Wuliangshan” in China. Land, 11(11),
1913. https://doi.org/10.3390/land11111913

Yang, Z., Ma, C., Liu, Y., Zhao, H., Hua, Y., Ou, S., & Fan, X. (2023). Provincial-Scale
Research on the Eco-Security Structure in the Form of an Ecological Network of the
Upper Yellow River: A Case Study of the Ningxia Hui Autonomous Region. Land, 12(7),
1341. https://doi.org/10.3390/land12071341

Zhou, D., & Song, W. (2021). Identifying Ecological Corridors and Networks in
Mountainous Areas. International Journal of Environmental Research and Public
Health, 18(9), 4797. https://doi.org/10.3390/ijerph18094797

Zhou, X., Hao, C., Bao, Y., Zhang, Q., Wang, Q., Wang, W., & Guo, H. (2023). Is the
Urban Landscape Connected? Construction and Optimization of Urban Ecological
Networks Based on Morphological Spatial Pattern Analysis. Sustainability, 15(20),
14756. https://doi.org/10.3390/su152014756

Zhang, X., Martin, T., & Newman, M. E. J. (2015). Identification of core-periphery structure in
networks. Physical Review E, 91(3), 032803. https://doi.org/10.1103/PhysRevE.91.032803