reflecting the importance of wetland hydrologic

Procedia Environmental Sciences 13 (2012) 1315 – 1326

1878-0296 © 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of School of Environment, Beijing Normal University.doi:10.1016/j.proenv.2012.01.124

Available online at www.sciencedirect.com

Procedia Environmental

Sciences Procedia Environmental Sciences 8 (2011) 1342–1353

www.elsevier.com/locate/procedia

The 18th Biennial Conference of International Society for Ecological Modelling

Reflecting the importance of wetland hydrologic connectedness: a network perspective

X.F. Mao, L.J.Cui *. Institute of Wetland Research, Chinese Academy of Forestry, Beijing 100091, China

Abstract

Different types of wetlands are hydraulically interconnected and form specific network structures that display systematic characteristics. The sustainable use of wetlands and water resources requires management approaches that incorporate Wetlands are not isolated spaces but complex habitats, in which many biotic and abiotic connections all around. The sustainable use of wetlands and water resources requires management approaches that incorporate explicitly the spatial and temporal interconnections among different hydrological units around a wetland. However, lateral, vertical and longitudinal hydrological connections have been destroyed heavily due to the impacts of anthropogenic activities. Wetlands tend to be managed in fragmented worldview behind non-sustainable growth. To highlight the importance of maintenance the integrity of interconnected wetlands, we developed a framework to use ecological network analysis in holistic assessment on the Okefenokee wetland system (WS) in USA. Two methods, ascendency analysis and network environ analysis, in ecological network method were used to explore the systematic attributes and components’ independencies of the wetland system. Results indicate the hydrological connections are extremely important for maintaining a wetland system. It is concluded that the proposed methods can provide effective ways to examine the hydrologic organization and components’ interdependencies in a wetland system and contributes to the basin-wide wetlands protection and water resources management. © 2011 Published by Elsevier Ltd. Keywords: Ascendency analysis; Ecological network analysis; Hydrological connections; Network environ analysis; Wetland network

* Corresponding author: Lijuan Cui. E-mail address: [email protected]

© 2011 Published by Elsevier B.V. Selection and/or peer-review under responsibility of School of Environment, Beijing Normal University. Open access under CC BY-NC-ND license.

Open access under CC BY-NC-ND license.

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by Elsevier - Publisher Connector

https://core.ac.uk/display/82097615?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

http://creativecommons.org/licenses/by-nc-nd/3.0/

http://creativecommons.org/licenses/by-nc-nd/3.0/

1316 X.F. Mao and L.J.Cui / Procedia Environmental Sciences 13 (2012) 1315 – 1326 X.F. Mao et al./ Procedia Environmental Sciences 8 (2011) 1342–1353 1343

1. Introduction

Different types of wetlands, such as rivers, lakes, marshes, estuaries, etc., are important ecosystems of the earth, which perform many important ecological functions. For example, they recycle nutrients, store and purify water, augment and maintain stream flow, and provide habitats for a wide variety of living organisms [1]. To protect the basic function of these wetlands, functional and status assessment of wetlands are essential step for ecologists. People develop effective ecosystem-based management strategies for ecosystem restoration and preservation on condition of accurate assessment. To understand the status and variation of ecosystems function under the impacts of human activities, much research has been done to assess the variation of wetlands’ condition in a range of sampling sites [2-5].

However, wetlands are not isolated spaces but complex habitats, in which many biotic and abiotic connections all around. Especially, the flow of water creates permanent or temporary links among and within multiple wetlands, which provide the important role for aquatic ecosystems [6]. These hydrological connections provide significant support (e.g., nutrients, information and energy transport) for maintaining ecosystem function, especially for wetlands within a basin. Due to the closure characteristics [7], many wetlands are, linked by complex hydrological processes in basins, acting as an integral system with specific network structure and functions at certain temporal and spatial scales [8, 9]. Thus, understanding the integral characteristics of these interconnected wetlands, instead of an individual wetland, is also essential to maintain and restore their ecological functions in basins. Many studies have revealed the importance of preserving the ecological integrity of these wetlands rather than only an individual unit in achieving the effective protection and restoration of wetlands [10-13]. And sustainable use of wetlands must explicitly consider the spatial and temporal interconnections among various wetlands in basins [14].

Since wetlands are interconnected in basins, the functional and conditional assessment of individual wetlands seems insufficient in providing enough information for holistic and sustainable ecosystem management. Without considering this pattern of connectivity, one can not assess the ecosystem functions properly [15]. Integral assessment of these interconnected wetlands is necessary but hard to realize. First, the current management practices are characterized by non-sustainable spatial divisions, wetlands tend to be managed on an individual basis because of their size and limited jurisdictional area [4]. Second, and more importantly, most research focus on establishing methods for the functional and conditional assessment of an individual wetland, few methods can deal with the assessment of multiple connected wetlands. Thus, it is urgent to find ways to investigate the interdependencies between system components and thus give accurate assessment to the whole system.

Ecological network analysis (ENA) was developed initially to evaluate food network but has broad applications [16-19]. It can evaluate a system from the viewpoint of connectivity and flows [20-22]. By ENA one can identify many indices of ecosystem functioning (e.g., Ascendency, Total System Throughput, Finn Cycling Index) and analysis the ecological relations of system components, which provide important information for ecosystems management [23].

In this paper, we introduced two methods of ENA to better understand the component’s relationships and integral functional characteristics of a system organized by wetlands and its associated hydrological units. We applied it to a case study of the Okefenokee Swamp wetland to demonstrate the method’s effectiveness. The paper is organized as follows: Section 2 presents the method and data used to measure the holistic structure and function of Okefenokee wetland network. Section 3 reports and interprets the

1317X.F. Mao and L.J.Cui / Procedia Environmental Sciences 13 (2012) 1315 – 13261344 X.F. Mao et al./ Procedia Environmental Sciences 8 (2011) 1342–1353

studied results and Section 4 discusses some considerations of the current study. Section 5 concludes with a simple retrospect to the entire paper.

2. Study area

Okefenokee watershed is a swamp-dominated hydrological unit. It is situated on the Lower Atlantic Coastal Plain of southeastern Georgia in the USA (Fig.1).

Fig.1. The Okefenokee watershed in USA (reconstructed from [27]).

The Okefenokee Swamp is one of the largest natural freshwater wetland systems in the world though human alterations have also affected the amount of water flow and its variability in the wetland [24]. Okefenokee watershed has an area of 3702 km2, of which 1891 km2 or 51% is occupied by Okefenokee Swamp; the remaining 1811 km2 or 49% is pine uplands [25]. Its climate is humid subtropical with a mean annual precipitation of 1285mm: hot and wet during May-September, warm and dry in October-November, cool and moist in December-February, and warm and moist during March-April [26]. More detailed information can be gotten from Patten and Matis [27].

3. Material and methods

3.1. Wetland network models description

In the Okefenokee network model, four hydrologic units including upland surface, upland groundwater, swamp surface and swamp subsurface were considered as main network components. These system components were connected with each others through complicated hydrologic processes.


Fig. 2. Quantified network model of Okefenokee WS (units: 108 m3y-1, 1-upland surface; 2-upland groundwater; 3-swamp surface; 4-swamp subsurface. Reconstruct from [27]).

3.1 Ecological network analysis

Network environ analysis (NEA) and ascendency theory are two important while independent lines of ENA. Each of them has advantages in dealing with complicated ecological networks. NEA can identify the direct and indirect ecological relationships and investigates system level properties of ecosystems. Ascendency ttheory deals with the joint quantification of overall system activity with the organization of component processes and could be used specifically to assess function of a system [16, 18].

3.1.1 Network environ analysis

(1) Network control analysis

Environ concept is introduced first because it is critical to understand the control analysis. Here ‘environ’ is defined as the within system environment of each system component, inclusive of the component, and note that every component has both an input environ and an output environ [28, 29]. They considered each component is influenced by components in its input environ, and influences components with its output environ as part of the network. The interdependence can be achieved by looking at the contribution of each component to the other component’s input and output environs, respectively [28, 29]. Thus, they developed a control matrix CN = (cnij=nij/n’ji) as the ratio of integral flow from compartment j to i to the integral flow from i to j. Compartment j is said to dominate i if its output environ effect on i is larger than input environ effect on j (cnij=nij/nji’>1). Contrarily, if the output environ of j has less influence on i than i receives as input from j, then j is said to be dependent on or controlled by i. The control relationship is further modified such that when nij/nji’<1, cnij=1-nij/nji’, otherwise cnij=0 [29]. Since the control matrix CN was built on the integral flow matrices, it gives an overall result to describe pair-wise control and dependency relationships between two regions/countries.

(2) Network utility analysis

Network utility analysis is an ecological network approach that was firstly introduced by Patten [30]. ‘Utility’ is an economic term to express the relative benefit to cost relationships in networks [31]. They consider that a gain of resources provides positive utility and a loss of resources provides negative utility. The method identifies both direct and indirect qualitative and quantitative ecological relationships in a network, thereby revealing the integration and complexity of ecosystem behaviours [32-34]. In utility


analysis, dij represents the direct utility of net interflow from compartment j to compartment i, and can be expressed as dij= (fij-fji)/Ti. From matrix D, which contains all dij values, a dimensionless integral utility intensity matrix U= (uij) = D0 + D1 + D2 + D3 + ···+Dm could be computed as transitive closure matrix U= (I-D)-1. D is the direct utility because the net flow is expressed relative to the total flow as each component. The matrix U reflects the intensity and pattern of integrated actions between any of compartment in the network, the identify matrix (I=D0) reflects the initial input flows through each compartment, the matrix D1 reflects the direct relationships between any of two compartments in the network, and Dm (m≥2) thus reflects the indirect interactions between compartments.

Based on the pairing signs of the elements in the direct utility matrices (D) or integral utility matrices (U), they determined the pattern of interaction between components in a network [35]. The direct relation between pair of components is given by pairing signs of direct utility matrix. For example, (sd21, sd12) = (+, −) indicates the conceptual relationship that compartment 2 exploits compartment 1, whereas if (sd21, sd12) = (−, +), then compartment 2 is exploited by compartment 1. The other possible relationship is neutralism ((sd21, sd12) = (0, 0)), which means there is no net gain between the components which generally indicates the absence of transactions or the opposing flows are equal. However, some relations arise only from integral utility matrix (U). If (su21, su12) = (−, −), then compartment 1 competes with compartment 2 (leading to negative impacts for both compartments), whereas if (su21, su12) = (+, +), then the relationship between the two compartments is a mutualism (both compartments benefit from their interaction).

Based on the sign matrices, mutualism index M = J (U) =S+/S− is calculated for virtual water trade system to reflect the holistic properties of system. Here, S+ = ijmax (sign (u ), 0) ij max (sign (uij), 0)

and S− = ij(-min (sign (u ), 0) )ij [32]. If the matrices have more positive signs than negative signs, it means that system exhibits mostly positive relationships between compartments and thus represents network mutualism. Conversely, if there are more negative signs than positive signs, the system exhibits mostly negative relationships between compartments and many problematic relationships must be solved or mitigated [22]. During the analysis of components interactions based on the sign distribution and mutualism index, we identified four possible intercompartment ecological relationships (includes competition, exploitation, neutralism and mutualism) and holistic status of virtual water trade network.

3.1.2 Ascendency theory

Here, we focused on 20 network indicators in Table 1. These indicators are divided into three categories: whole-system indicators, component system indicators and ratio-based indicators. Whole-system indicators describe the whole system, including Total System Throughput (TSTP), Average Mutual Information (AMI), Ascendency (A), Overhead (Ø), and Development Capacity (C). TSTP reflects the level of activity which is measured by the sum of the magnitudes of all the flow exchanges occurring in the system. The AMI represents the organization inherent in a system because it capture the average amount of constraint exerted upon an arbitrary amount of mass as it flows from any one compartment to the next [18]. Ascendency is the production of TST and AMI that quantifies both the level of system activity and the degree of the organization [16]. Development Capacity is functions as a mathematical upper bound on the ascendency. It represents the scope of the system for further development. Overhead represents multiplicity of pathways; consequently, when it is high, it is said to


reflect a system under rigorous environmental conditions [35], so it is generated by structural ambiguities deriving from multiplicities in system inputs, exports, dissipations and internal exchanges [36].

Component system indicators are the decomposed Ascendency (A0, Ai, Ae, As), Overhead (Ø0, Redundancy(R), Øe, Øs) and Capacity (C0, Ci, Ce, Cs) measures could provide more detailed information of network in four quarters of the network: import, internal and export, dissipation [35]. A0, Ø0 and C0 are import Ascendency, Overhead from import and import Capacity, respectively. Ai, R and Ci are internal Ascendency, internal Overhead and internal Capacity, respectively. Ae, Øe and Ce describe the Export Ascendency, Overhead from Export and Export Capacity, respectively. As, Øs and Cs correspond to dissipative Ascendency, dissipative Overhead and dissipative capacity, respectively. In our study, we make no difference between export and dissipation. Dissipative Ascendency, Dissipative overhead and Dissipative capacity equals to zero and not considered further.

Ratio-based indicators could be used to quantify ecosystem health and condition [17, 18]. For example, the Ascendency over Capacity (A/C) would describe the network efficiency and optimized at system maturity while the internal ascendency to internal capacity ratio Ai/Ci describes the internal network efficiency. The Overhead over Capacity (Ø/C) might show how the Capacity is limited by the Overhead, and H= C/TST describes the diversity of flows in the system. Other two ratio-based indicators include Link density and Finn’s cycling index (FCI) are used to depict the basic network characteristics of the WS.

Table 1 Network analysis indicator name, symbol and algorithms

NO. Name Symbol Algorithms

1 Total System Throughput TSTP 2 2

1 1

n n

iji j

T

2 Average Mutual Information AMI

2,

logij ij

i j i j

T T TT T T

3 Ascendency A

2log ijij

ij i j

T TT

T T

4 Import Ascendency A0 1,

1, 21 1

logn

n jn j

j n j

T TT

T T

5 Internal Ascendency Ai

21

logn

iji j

ij i j

T TT

T T

6 Export Ascendency Ae , 2

, 2 21 2

logn

j nj n

j n j

T TT

T T


7 Overhead Ø 2

2log ijij

ij i j

TT

T T

8 Overhead from Import Ø0 21,

1, 21 1

logn

n jn j

j n j

TT

T T

9 Redundancy R 2

21

logn

iji j

ij i j

TT

T T

10 Overhead from Export Øe 2

, 2, 2 2

1 2

logn

j nj n

j n j

TT

T T

11 Development Capacity C

2log ijij

ij

TT

T

12 Import Capacity C0 1,

1, 21

logn

n jn j

j

TT

T

13 Internal Capacity Ci

21

logn

ijij

j

TT

T

14 Export Capacity Ce , 2

, 2 21 ..

logn

j nj n

j

TT

T

15 Ascendency/ Capacity A/C =A/C

16 Internal Ascendency/ Internal Capacity

Ai/Ci =Ai/Ci

17 Overhead/ Capacity Ø/C =Ø/C

18 H C/TST =C/TST

19 Link density L/n =L/n

20 FCI TSTc/TSTs TSTc/TSTs

T is the TSTP; (n+1) are import value; (n+2) are export value;


3.2 Data sources

In the Okefenokee wetland system model, Patten and Matis used mean annual data to quantify their model [27]. Mass quality balance method was also used to help their quantification. Considering the difficulties of quantification for such large regions, here the model and its data are accepted as a point of departure.

4. Results and discussion

4.1 System interdependencies between hydrological components

(1) Ecological control relationships

Matrix CN is depicted as Equation (1). As little exchanges occurred between these components, the control matrix is relative simple and clear. The largest value of CN equals 1 in a system, indicating a component is totally controlled by or over another component during their transactions [37]. In the current matrix, the largest value is 1.0, indicating some component depended totally on the other hydrological components. CN13, CN14, CN23 and CN24 belong to the above condition. The other components exhibit some degree of dependence on other regions. For example, cn12=0.86 means the dependence of upland groundwater (component 2) on upland surface (component 1) was 85.9%. Through the results of control analysis, we can quantify the static flow control and dependencies on both donor and recipient-oriented processes in holistic way.

(1) The independency between components within the system is extremely important for wetland system survival although little exchanges took placed. As a result, holistic knowledge of controlling relations is important for the understanding of system structure and function. (2) Ecological utility relationships

Utility matrix D and integral utility matrix U are depicted in Equation (2) and Equation (3), respectively.

0 0.664 0.197 00.860 0 0.001 0.0870.174 0.011 0 0.001

0 0.982 0.012 0

D

(2)

0 0.86 1.00 1.000 0 1.00 1.000 0 0 0.6850 0 0 0

CN


0.641 0.392 0.125 0.0340.507 0.610 0.101 0.0530.112 0.0683 0.978 0.0050.449 0.598 0.083 0.947

U

(3)

Matrix (D) depicts the direct utility between system components and provides us direct information of

the system hydrological characteristics. Water recycles and interactions of among these subsystems will produce indirect relationships and alter their direct relationships, which can only be detected from the integral utility matrix (U).

Mutual relationship between pair-wise components can be elicited from matrix (D) and matrix (U). From the perspective of matrix (D), there are two kinds of ecological relationships including exploitation and neutralism. An example of exploitation can be given as (D21, D12) = (-0.664, 0.860), which indicated upland groundwater got a net positive utility of 0.860 from upland surface and upland surface got a net negative utility of -0.664 from upland groundwater. An example of neutralism can be given by (D41, D14) = (0, 0), indicating there is no direct hydrological exchanges between two components. Integral utility matrix (U) provides a more complete picture of network utility relationships because it integrates both direct and indirect utility. From the perspective of matrix (U), there are also two types of ecological relationships including exploitation and mutualism. An interesting phenomenon is the hydrological relationships between swamp subsurface (component 4) and upland surface (component 1) changed from neutralism to mutualism, indicating two hydrological components benefit from each others from indirect transactions. It is also reflect the fact that hydrological modification should be planned under an integral framework.

4.2 System attributes

The ENA generated total 20 indicators with respect to the current WS. As we can see from Table 2, the basic structure and flow characteristic of the WS can be detected from Connectance indices and Finn’s cycling index. The Okefenokee WS is a well connected system with a link density of 3.25 and a connectance of 1.08. Recycling in the Okefenokee WS is not diverse with a FCI of 0.15. Their distinct nature contributes to great disparity in information indices of the WS.

Table 2 Comparisons of network information indices.

Indicators TSTP AMI A A0 Ai Ae Ø Ø0

value 12.71 1.27 16.17 5.66 4.76 5.74 18.65 5.91 Indicators R Øe C C0 Ci Ce A/C Ø/C

value 5.8 6.93 34.82 11.57 10.56 12.68 46% 54%

Indicators R/C Ai/Ci R/Ci A0/A Ai/A Ae/A Conn FCI

value 17% 45% 55% 35% 29% 36% 1.08 0.15 Whole-level indices are useful for comparing the total activities of ecosystems and the organization

inherent in the flow topology. For example, system activity is measured by the TSTP, which is simply the sum of all the processes occurring in the system. According to the index, the Okefenokee WS (0.34 m3y-

1per m2) is far more active than other wetland system (e.g., Baiyangdian WS 0.21 m3y-1per m2) [8]. Okefenokee WS is located in humid subtropical area with a mean annual precipitation of 1285mm. Hydrological activities should be active in Okefenokee WS. AMI captures topographical constraint based upon the pattern of flow in the network, which measures the overall degree by which one compartment


communicates unambiguously with any other. Results show that AMI and Ascendency of Okefenokee WS are about 1.27 and 16.17 (m3y-1), respectively.

When one regards the component ascendancy, one may find that properties of the WS are reflected more detailed by these indicators. For example, A0, Ai and Ae reflect system ascendency generated by boundary input, interflow and boundary output, respectively. A0 and Ae of the Okefenokee WS are 5.66(m3y-1) and 5.74(m3y-1), respectively, which are larger than Ai (4.76 m3y-1). The Okefenokee WS is a well connected network but with moderately or weak connection (interflow) between system components.

Dimensionless ratio-based indicators represent holistic organization of system or the contribution of different parts of flow to overall system performance. For example, the fraction of the development capacity that appears as ordered flow (A/C) is 46%. Overhead over Capacity (Ø/C) the Okefenokee WS is 54% (Notice that the A and Ø will always vary in opposite direction). The Internal Ascendency over the Internal Capacity (Ai/Ci) of the Okefenokee WS is 45%, which indicates the Okefenokee WS has some Redundancy for the development of its internal organization. The R/Ci and R/C of Okefenokee WS are respectively 55% and 17%. The ratio-based component Ascendency exams how much the system ascendency is generated from inputs, interflows and outputs. One can easily find that the distribution of Ascendency in the Okefenokee WS is somewhat homogeneous. Three indicators, including A0/A, Ai/A and Ae/A in Okefenokee WS are respectively 35%, 29% and 36%.

5. Concluding remarks

In most of cases, wetland is not an isolated hydrological unit but, on the contrary, dynamic habitats with complicated hydrological connections all around. Various hydrological processes produce permanent or temporary links among multiple wetland units and create their distinct network topological structure. Each wetland plays a role in their wetland networks, which operates just like a part of organized wetland systems. We can not understand the whole system by examing only part of the whole system. As a result, it is impossible to restore and manage a wetland system by restoring only one wetland. Thus, an initial step is to develop valid method to support the need for wetlands policies that go beyond conservation and restoration of individual wetland sites.

By applying two analysis methods of NEA on the simple hydrological system model, we intend to introduce both ascendency and network environ analysis to the study of wetland systems. Results demonstrate that intercompartmental transfer are important for the organization of the large hydrological system, even if there was little hydrological exchanges occurred between these subsystems:

(1) Boundary input, output and interflow are playing different roles in the hydrological organization of the large system, which reflected in the input ascendency, output ascendency and inter ascendency.

(2) Each components are controlled or be controlled by other components through interactive hydrological processes.

(3) Network-based management should be considered towards sustainable wetland utilization.

The Okefenokee WS can be given as an example to show the importance of management in holistic view. The current researches can be served as a starting point for further wetland theory study and realistic applications.


Acknowledgements

This work is supported by the National Science & Technology Pillar Program (2011BAC02B03) and Forestry Industry Research Special Funds for Public Welfare Projects (201204201).

References

[1] Rubec CDA, Hanson AR. Wetland mitigation and compensation: Canadian experience. Wetlands Ecol. Manage 2009; 17: 3- 14.

[2] Brinson MM, Rheinhardt RD. The role of reference wetlands in functional assessment and mitigation. Ecol Appl 1996; 6, 69-76.

[3] Patrícioa J, Ulanowicz RE, Pardala MA, Marques JC. Ascendency as an ecological indicator for environmental quality assessment at the ecosystem level: a case study. Hydrobiologia 2006; 555: 19-30.

[4] Christiana RR, Brinsona MM, Damea JK, Johnsonb G, Peterson CH, Baird D. Ecological network analyses and their use for establishing reference domain in functional assessment of an estuary. Ecol Model 2009; 20: 3113-22.

[5] Xu F, Yang ZF, Chen B, Zhao YW. Ecosystem health assessment of the plant-dominated Baiyangdian Lake based on eco-exergy. Ecol Model 2011; 222: 201-9.

[6] Amezaga JM, Santamaría L, Green AJ. Biotic wetland connectivity- supporting a new approach for wetland policy. Acta Oecol 2002; 23: 213-22.

[7] Molle F, Wester P, Hirsch P. River basin closure: Processes, implications and responses. Agr Water Manage 2010; 97(4): 569-77.

[8] Mao XF, Yang ZF, Chen B, Chen H. Examination of wetlands system using ecological network analysis: A case study of Baiyangdian Basin, China. Procedia Environmental Sciences 2010; 2: 427-39.

[9] Mao XF, Yang ZF, Chen B. Wetlands network analysis for Baiyangdian wetlands system in China and comparison with Okefenokee wetlands system in USA. J Env Inform 2011. Doi: 10.1016/j.ecoinf.2011.05.006

[10] Tilley DR, Brown MT. Wetland networks for stormwater management in subtropical urban watersheds. Ecol Eng 1998; 10: 131-58.

[11] Cohen MJ, Brown MT. A model examining hierarchical wetland networks for watershed stormwater management. Ecol. Model 2007; 201: 179-93.

[12] Yang ZF, Sun T, Cui BS, Chen B, Chen GQ. Environmental flow requirements for integrated water resources allocation in the Yellow River Basin, China, Commun. Nonlinear Numer Simul 2009; 14: 2469-81.

[13] Bodin Ö, Saura S. Ranking individual habitat patches as connectivity providers: Integrating network analysis and patch removal experiments. Ecol Model 2010; 221: 2393-405.

[14] Amezaga JM, Santamaría L, Green AJ. Wetland connectedness and policy fragmentation: steps towards a sustainable European wetland policy. Phys Chem Earth (B) 2000; 25(7-8): 635-40.

[15] Leibowitz SG, Loehle C, Li BL, Preston EM. Modeling landscape functions and effects: a network approach. Ecol Model 2000; 132: 77-94.

[16] Cumming GS, Bodin Ö, Ernston H, Elmqvist T. Network analysis in conservation biogeography: challenges and opportunities. Divers Distrib 2010; 16: 414-25.

[17] Ulanowicz RE. A hypothesis on the development of natural communities. J Theor Biol 1980; 85: 223-45. [18] Ulanowicz RE. Growth and Development: Ecosystems Phenomenology. Springer-Verlag, NY; 1986, p. 203. [19] Ulanowicz RE. Ecology, the Ascendency Perspective. Columbia University Press, New York; 1997, 201 pp. [20] Fath BD, Patten BC. Review of the foundations of network environ analysis. Ecosystems1999; 2: 167-79. [21] Schramski JR, Gattie DK, Patten BC, Bata S, Whipple SJ, Borrett SR, Fath BD. Indirect effects and distributed control in

ecosystems: distributed control in the environ networks of a seven-compartment model of nitrogen flow in the Neuse River estuary USA: time series analysis. Ecol Model 2007; 206 (1-2): 18-30.


[22] Borrett SR, Fath BD, Patten BC. Functional integration of ecological networks through pathway proliferation. J Theor Biol 2007; 245: 98-111.

[23] Zhang Y, Yang ZF, Fath BD. Ecological network analysis of an urban water metabolic system: Model development, and a case study for Beijing. Sci Total Environ 2010; 408: 4702-11.

[24] Tobor-Kaplon MA, RHoltkamp H, Scharler UM, Doroszuk A, Kuenenf FJA, Bloem J, De Ruitera PC. Evaluation of information indices as indicators of environmental stress in terrestrial soils. Ecol Model 2007; 208: 80-90.

[25] Loftin CS. Okefenokee Swamp Hydrology. Proceedings of the 1997 Georgia Water Resource Conference, March, 1997. University of Gerogia, Athens, Georgia. 1997.

[26] Rykiel EJ. Okefenokee Swamp watershed: water balance, nutrient budgets, and modeling. In: H.T. Odum and K.C. Ewell, Editors, Cypress Swamps. 1982.

[27] Patten BC, Matis JH. The water envions of Okefenokee Swamp: Applications of static liner environ analysis. Ecol. Model 1982; 16: 1-50.

[28] Patten BC. Systems approach to the concept of environment. Ohio J Sci 1978; 78: 206-22. [29] Patten BC, Auble GT. System theory of the ecological niche. Am Nat 1981; 117: 893-922. [30] Patten BC. Network ecology: indirect determination of the life-environment relationship in ecosystems. In: Higashi, M.,

Burns, T. (Eds.), Theoretical Studies of Ecosystems: The Network Perspective. Cambridge University Press, New York, 1991, pp: 2-351.

[31] Fath BD, Patten BC. Network synergism: emergence of positive relations in ecological systems. Ecol Model 1998; 107: 127-43.

[32] Fath BD. Community-level relations and network mutualism. Ecol Model 2007; 208, 56-67. [33] Lobanova G, Fath BD, Rovenskaya E. Exploring simple structural configurations for optimal network mutualism. Commun.

Nonlinear Numer Simul 2009; 14: 1461-85. [34] Scharler UM, Fath BD. Comparing network analysis methodologies for consumer–resource relations at species and

ecosystems scales. Ecol Model 2009; 220 (22): 3210-8. [35] Ulanowicz RE, Norden JS. Symmetrical overhead in flow networks. Int. J. Syst. Sci 1990. 21: 429-37. [36] Patrícioa J. Ulanowicz RE. Pardal MA, Marques J.C. Ascendency as an ecological indicator: a case study of estuarine pulse

eutrophication. Estuar Coast Shel S 2004; 60: 23-35. [37] Fath BD. Distributed control in ecological networks. Ecol Model 2004;179: 235-46.

reflecting the importance of wetland hydrologic

Documents