Abstract

Baird, D. (2009) An assessment of the functional variability of selected coastal ecosystems in the context of local environmental changes. – ICES Journal of Marine Science, 66: 1520–1527.

The functioning of coastal ecosystems is greatly dependent on a wide variety of external pulses (e.g. tides, freshwater influx, seasonal trends in temperature, nutrient input, etc.). Assessments of the effect of a selection of environmental characteristics driven by natural and/or anthropogenic forces on ecosystem function are given using selected ecosystem properties, such as total system throughput, system organization, productivity, recycling, and trophic efficiency, derived from ecological network analysis (ENA) of several coastal ecosystems on monthly, intra-seasonal, seasonal, and interdecadal scales. Each ecosystem was modelled depicting data of standing stocks, the flows between the constituent living and non-living components in the system, exports, and imports. Results from ENA revealed considerable differences of the same property (or properties) resulting from physical changes (e.g. temperature, salinity, oxygen, rate of freshwater inflow) over time. A small temperature increase in a Florida seagrass bed, for example, resulted in increases in system throughput, the P/B ratio, and in the rate of carbon recycling, but also in a significant decrease in system organization. The effect of seasonal increases in water temperature and of measured decrease/increase in river run-off to a few selected estuaries is discussed using ENA.

Introduction

Climate change is a global phenomenon, affecting different geographic regions on the planet differently. Global climate change is considered a direct result of varied human activities and is currently one of the most contentious topics in environmentalism, ecology, and governance (Halpern et al., 2007, 2008). It is predicted that climate change will affect multiple environmental processes and phenomena, such as rainfall, storm events, terrestrial and oceanic primary production, atmospheric and ocean temperature, and ocean currents (Philippart, 2007). Climate change will also influence the magnitude and rate of the direct and indirect goods and services that ecosystems provide. Even more important, some plant and animal species will become extinct, whereas some will proliferate under changing environmental conditions. Wide-scale migrations of plants and animals are envisaged, contributing to a dramatic change in the biodiversity of most kinds of ecosystems as we know them (Naeem, 2006). The question of how a change in biodiversity will affect the functioning of ecosystems [the so-called biodiversity-ecosystem function (BEF) debate] is clearly not easy to answer. Duffy (2006), for example, suggested that investigations of the effect of changing biodiversity on processes in aquatic ecosystems could benefit from extending the biodiversity concept to include landscape diversity, and by focusing on links between diversity and trophic interactions. Ieno et al. (2006) and Naeem (2006) remarked that BEF experiments should adopt a holistic approach to distinguish between the contributions of multiple ecosystem processes to system function, whereas Raffaelli (2006) argued that a system approach is necessary to address issues involving changes in biodiversity and changes in the natural functioning of coastal ecosystems. It was also pointed out, for example, by Raffaelli et al. (2002), Hooper et al. (2005), and Bulling et al. (2006) that relatively few (theoretical) studies have investigated the effect of changes in species richness on ecosystem properties and that most studies have focused on within-trophic group diversity or taxonomic grouping within a much larger foodweb network. Cardinale et al. (2002) deduced from mesocosm studies that changes in species diversity might affect species interactions, which could cause disproportional large changes in the functioning of ecosystems. Hooper et al. (2005) further suggested that changes in the biota could have greater effects on ecosystem properties than changes in abiotic characteristics. There is apparently a strong opinion that changes in biodiversity resulting from climate change could affect ecosystem properties and function and that functional changes in system processes, such as energy-flow pathways, cycling of energy and material, and the efficiency of energy transfer and utilization, can result from changes in both abiotic characteristics and species composition. Experimentation with artificially constructed model systems (based on micro- and mesocosms) to examine the relationships between biological and environmental processes that attempt to address questions relevant to BEF issue have been reviewed by Bulling et al. (2006). Although these experiments have been useful for our understanding of BEF, these model systems are by their nature simplistic, they do not readily incorporate multiple trophic levels and represent reductionist methodology, but are nevertheless useful to shed light on essential characteristics of the effect of changes in the physical–chemical environment on biodiversity and ecosystem function (Bulling et al., 2006).

Although there does not appear to be a universally accepted suite of readily available, measurable, and quantifiable ecosystem characteristics to detect the impact of climate change and changes in biodiversity on ecosystem function, ecological network analysis (ENA) does provide considerable insight into system function. The ideas of Odum (1969) on ecosystem development and maturity initiated considerable interest to characterize ecosystem organization and the identification of properties that are suitable indices of ecosystem function, and which can be used to compare ecosystems on spatial and temporal scales (cf. Ulanowicz, 1986; Wulff et al., 1989; Baird et al., 1991, 2007; Jørgenssen et al., 1992; Christensen, 1994; Baird, 1998; Christian et al., 2005). ENA involves an analysis of foodweb networks that depicts the standing stocks of the living and non-living components and the flows between them within an ecosystem. Assessments of trophic structure through ENA have been done for a wide variety of marine and coastal, particularly estuarine, ecosystems. For example, some have used it to compare trophic structure and system function focusing on temporal scales (Baird and Ulanowicz, 1989; Baird et al., 1998, 2004a), and among ecosystems focusing on spatial scales (Baird and Ulanowicz, 1993; Christensen, 1995; Baird et al., 2007). In these studies, carbon or energy was used as the currency with which to trace the interactions of the foodwebs, although other key elements, such as nitrogen and phosphorus, have also been used in ENA (Baird et al., 1995; Ulanowicz and Baird, 1999; Christian and Thomas, 2003). One of the primary features of ENA is that the interactions between the components of an ecosystem represent rates of flows of energy or matter and not simply their existence as biomass entities.

Foodweb models are static, but allow inference about energy-flow dynamics and consequences of changes within the ecosystem on its various components. For example, Baird et al. (1998) illustrated changes in system properties following a 5°C increase in water temperature over a month in a Florida seagrass bed, and the effect of oxygen depletion on the function of the Neuse River estuary (Baird et al., 2004a).

Such models rely on principles of conservation of mass (or energy in other applications), using known inputs of energy (or carbon as a surrogate for energy), the biomass of all major functional compartments of the ecosystem, the flow of energy between the compartments, and diets to balance the system (Ulanowicz, 1986). The output is a system of flows that permits the description of the fate of carbon (C) or other material as it moves through the foodweb, and of the structural properties of the flows, which imply characteristics of system dynamics. Species are often pooled into compartments on grounds of trophic similarity, or in community components, where species cannot easily be identified taxonomically (e.g. bacteria, phytoplankton, and microzooplankton).

The objectives of this paper are (i) to examine the foodweb dynamics of several coastal ecosystems over temporal scales (i.e. the same system over time) that have changed as a result of a change in some environmental parameter, and to use the output results from ENA to illustrate functional changes at the ecosystem level, and (ii) to illustrate the effect of freshwater input into estuaries on ecosystem function over spatial scales.

The results from ENA can be used to assess changes at the ecosystem level of the systems, in the context of changing environmental parameters and characteristics. In this paper, estuaries are used to illustrate changes in function and structure, because more network analyses have been done on them than on any other kind of ecosystem, and they exhibit stress caused by natural or anthropogenic forcing functions readily and measurably, which allows opportunities to evaluate the controls relevant to trophic structure. Furthermore, the data and results available from ENA are extensive, so reasonable foodwebs may be constructed under different environmental conditions. However, surprisingly few studies have been conducted of the same ecosystem over different temporal scales. Most of the literature refers to annual averaged mass-balanced models.

Study sites and methods

The following coastal ecosystems were studied over a variety of temporal scales: The Neuse River estuary, NC, USA, (indicated by A, Figure 1) from early to late summer in 1997 and 1998 (cf. Baird et al., 2004a); the St Marks Wildlife Refuge, Apalachee Bay, Florida, USA, (see B, Figure 1) during 2 months in 1997 (cf. Baird et al., 1998); the mesohaline region of the Chesapeake Bay, Maryland, USA, (see C, Figure 1) over the four seasons (cf. Baird and Ulanowicz, 1989); and the Kromme River estuary, Eastern Cape, South Africa (see D, Figure 1) over decadal scales (cf. Baird and Heymans, 1996).

Figure 1.

Location of case studies.

Figure 1.

Location of case studies.

At a different spatial scale, I contrast four estuaries in South Africa with different freshwater inflow regimes, which is the only substantial varying physical forcing factor that they have in common. They are the Kromme River estuary (34°09′S 24°51′E), the Gamtoos River estuary (33°09′S 25°04′E), the Swartkops River estuary (33°52′S 25°38′E), and the Sundays River estuary (33°43′S 25°25′E). These systems are in proximity along the Eastern Cape coast, and fall within the same climatic region (indicated collectively by E, Figure 1).

Data on standing stocks, trophic exchanges, and imports and exports for each of the systems discussed were obtained from relevant publications and unpublished information. These sources are given below under the appropriate sections of the paper. Flow networks were constructed for each individual system depicting standing stocks of the living and non-living components and flows between them. Where comparisons within and among systems were made, the individual system(s) consisted of the same number of compartments and the same topological structure.

ENA provides a myriad of output variables and indices. The outputs from network analysis provide many useful indices and system properties of natural ecosystems, relevant information on the interpretation of energy and nutrient flows, how these affect the structure of the ecosystem, and how one may wish to direct management or monitoring actions for the conservation or rehabilitation of biodiversity and ecosystem function (Christian and Thomas, 2003).

The first comprehensive review of the methodologies and use of ENA, and its application in marine ecology, was published by Wulff et al. (1989). The methodology was described in detail by Kay et al. (1989), whereas Ulanowicz (2004) gave a detailed review of the quantitative methods used in network analysis. ENA, which consists of a systematic assessment of flow networks, was used to analyse the networks of the ecosystems listed above. The software package NETWRK 4.2a was developed by Ulanowicz and Kay (1991) and is available with supporting documentation at: www.cbl.umces.edu/~ulan/ntwk/network.hmtl.

Not all the output results from network analysis are presented here, but only those that have direct relevance to the objectives of this paper. Detailed descriptions of the suite of output results that can be derived from ENA can be found in Baird and Ulanowicz (1989), Baird et al. (1998, 2004a, b, 2007), and Christian et al. (2005). The following sections provide brief descriptions of the system properties referred to in this paper, to illustrate change at the system level in the ecosystems discussed here.

  • The Lindeman trophic aggregation routine transforms each complex network of trophic transfers into a linear food chain with discrete trophic levels (i.e. the Lindeman spine). The spine illustrates the amount of material or energy that each level receives from the preceding one, as well as the fraction lost from each level through respiration and export, and the net production passed on to the next higher level. It also illustrates the pool of recycled detrital material, which, together with the inputs to the autotrophs, forms the first trophic level. The Lindeman spine also allows the calculation of the efficiency of trophic transfer for each level (i.e. the efficiency of transfer of energy and material from one level to the next), whereas the trophic efficiency of the whole system can be derived from the logarithmic mean of the efficiencies of each integer trophic level (Baird and Ulanowicz, 1989).

  • The biogeochemical cycle routine of NETWRK4.2 assesses the structure and magnitude of the cycling of material in the system (Finn, 1976). The cycle distribution gives the amount of material that flows through cycles of various length, where a cycle represents a series of transfers between compartments beginning and ending in the same compartment, without passing through the same compartment twice (Baird et al., 2004a). The Finn cycling index (FCI) is derived from the fraction of the sum of flows that is devoted to cycling, and is equal to Tc/TST, where Tc is the amount recycled and total systems throughput (TST) represents the sum of all flows in the ecosystem [see (iii) below]. The FCI is an index of the retentiveness of the system (Baird and Ulanowicz, 1989; Baird et al., 2004a);

  • Various global system indices, based on information theory, describe the developmental and organizational state of the ecosystem (Ulanowicz, 1986, 2004). The TST measures the extent of the total activity of the system and is calculated as the sum of all the flows through all compartments. The system ascendancy (A), which is a single measure of the magnitude and diversity of flows between compartments, reflects on the functional attributes of the system. It incorporates both the size and organization of flows into a single index and is formally expressed as the product of TST and the average mutual information (AMI) inherent in the flow network. A complex trophic structure and high system productivity enhance ascendancy (A). The AMI index, or normalized ascendancy, is indicative of the developmental status of the ecosystem and thus of its inherent organization (i.e. the degree of specialization of flows in the network; Ulanowicz, 2004). The development capacity (DC) is the product of TST and the flow diversity. It measures the potential for a system to develop and is the natural upper limit of A. The total system overheads (i.e. overheads on imports, exports, and dissipation) and redundancy [i.e. a measure of the uncertainty associated with the presence of multiple or parallel pathways among the components of the network (Kay et al., 1989; Ulanowicz and Norden, 1990)] is numerically represented by the difference DC − A and represents the fraction of the DC that does not appear to be organized structure. The magnitudes of imports and exports reflect the self-reliance of a system (i.e. the higher these values, the more dependent the system becomes on external exchanges). A system with low redundancy is considered susceptible to external perturbations, which may affect the trophic interactions between system components. Parallel pathways of energy and material transfers can conversely act as a buffer or reserve should external perturbations occur and in changes in biodiversity. It is postulated that a sustainable system requires a balance between ascendancy and redundancy, for should a perturbation occur, the system can draw from the overhead to keep it in operation, but in a less organized state (Ulanowicz, 1986, 1997; Baird et al., 1991; Scharler and Baird, 2005).

Ascendancy measures the efficiency and definitiveness by which energy transfers are made, whereas the overhead quantifies how inefficient and ambiguous the system performs on average. Higher indices of A reflect increased ecological succession characterized by, for example, species richness, decreased cost of overheads to the system, greater internalization of resources, and finer trophic specialization (Scharler and Baird, 2005). Internal ascendancy (Ai) and internal developmental capacity (DCi) are functions of internal exchanges alone, so exclude exogenous transfers. The ratios A/C and Ai/DCi have been used to compare the organizational status of ecosystems on temporal (Baird and Ulanowicz, 1989; Baird and Heymans, 1996; Baird et al., 1998, 2004a) and spatial (Baird et al., 1991, 2007; Baird and Ulanowicz, 1993; Baird, 1998, 1999) scales. The magnitude of two important system attributes, namely the AMI and flow diversity, are strongly influenced by the total activity, or TST. By dividing theses capacities (DC and A) by TST, the resultant normalized values, presented in the tables, are scaled to eliminate the singular effect of TST (cf. Baird and Ulanowicz, 1989; Baird et al., 1998).

Flow diversity, defined as DC/TST (or normalized DC), measures both the number of interactions and the evenness of flows in the foodweb and is thus a much more dynamic concept than species diversity (Mann et al., 1989; Baird et al., 1998; Ulanowicz, 2004). Comparatively higher values of flow diversity indicate an increase in interactions and a lower degree of unevenness and variability in the flow structure (Baird et al., 2004a, b). The effective number of connections between compartments is given by three connectance indices and is derived from the log-averaged number of links calculated from the systems overhead (Baird et al., 2004b). Here, reference is made to the foodweb connectance index, which refers only to transfers among the living compartments in the system (Ulanowicz, 2004).

Comparisons of system function over temporal scales

The Neuse River estuary, NC, USA

The Neuse River estuary (76°30′N 34°55′E) receives water from a watershed of ∼16 000 km2 and opens into Pamlico Sound. In this paper, attention is focused on the mesohaline portion of the Neuse River estuary, which has a surface area of ∼267 km2, an average depth of 3.8 m, and a total volume of ∼1 × 106 m3. The temperature varies between 5°C in winter and 30°C in summer, and the temperature increases on average from ∼27°C in early summer to ∼30°C in late summer (Christian et al., 1991).

Hypoxia occurs routinely during summer in the mesohaline portion of the Neuse River estuary, where the oxygen concentration can drop from >6 mg l−1 in early summer to <2 mg l−1 during late summer. Oxygen depletion in the Neuse estuary is the result of complex interactions of several environmental phenomena, such as excessive nutrient loading, municipal and industrial wastewater discharges, and stormwater run-off (Paerl et al., 1998). Deteriorating water quality has been associated with high chlorophyll concentrations, nuisance blooms of cyanobacteria and dinoflagellates, oxygen depletion, and fish kills (Paerl et al., 1999; Baird et al., 2004a). Density stratification develops in the Neuse estuary during summer primarily because of temperature and salinity discontinuities, whereas biological oxygen demand is enhanced by higher temperatures, which causes hypoxia below the mixed surface layer (Paerl et al., 1998). Illustrated here is how hypoxia during summer affects trophic transfers and ecosystem properties in the mesohaline region of the Neuse estuary. Quantitative flow models representing the early and late summer conditions of 1997 and 1998 consisted of 25 living and 5 non-living compartments. The variability in standing stocks, productivity, and magnitudes of flows between the compartments, from early to late summer, were incorporated in the models. Table 1 illustrates several environmental variables and system properties. Algal biomass increased over summer, mainly because of increased phytoplankton production, whereas the heterotrophic biomass (mainly benthic invertebrates) declined by ∼38%, because of hypoxic conditions that developed in the bottom waters of the estuary. The increase in system production and in the daily P/B ratio was mainly because of an increase in phytoplankton production, whereas the increase in the system's trophic efficiency can be ascribed to the diversion of energy into microbial pathways and the mass mortality of benthic invertebrates during late summer, when the demand by predatory fish and crabs was high (Baird et al., 2004a). Although the number of cycles increased dramatically over summer, the FCI increased marginally, an indication that the greater number of cycles controlled a relatively small fraction of recycled material during late summer.

Table 1.

System attributes during early and late summer in the Neuse River estuary, NC.

System attribute Early summer Late summer % increase (+), % decrease (−) 
Temperature (°C) 27 30 11 
O2 concentration (mg l−1>6 <2 −66.7 
Algal biomass (mg C m−21 988 2 180 9.7 
Heterotrophic biomass (mg C m−218 149 11 185 −38.4 
System production (mg C m−2 d−13 810 4 260 11.8 
P/B (d) 0.2 0.3 50.0 
System trophic efficiency (%) 4.12 4.82 17.0 
Number of cycles 135 641 374.8 
FCI (%) 14.8 15.3 3.4 
TST (mg C m−2 d−118 404 19 175 4.2 
Ascendancy 37 723 37 141 −1.5 
A/C (%) 47.1 46.6 −1.1 
Ai/Ci 57.1 58.6 2.6 
AMI (A/TST) 2.05 1.94 −5.5 
Redundancy 17 666 15 392 −12.9 
Foodweb connectance 1.35 1.36 0.7 
Flow diversity (D/TST) 4.35 4.16 −4.4 
System attribute Early summer Late summer % increase (+), % decrease (−) 
Temperature (°C) 27 30 11 
O2 concentration (mg l−1>6 <2 −66.7 
Algal biomass (mg C m−21 988 2 180 9.7 
Heterotrophic biomass (mg C m−218 149 11 185 −38.4 
System production (mg C m−2 d−13 810 4 260 11.8 
P/B (d) 0.2 0.3 50.0 
System trophic efficiency (%) 4.12 4.82 17.0 
Number of cycles 135 641 374.8 
FCI (%) 14.8 15.3 3.4 
TST (mg C m−2 d−118 404 19 175 4.2 
Ascendancy 37 723 37 141 −1.5 
A/C (%) 47.1 46.6 −1.1 
Ai/Ci 57.1 58.6 2.6 
AMI (A/TST) 2.05 1.94 −5.5 
Redundancy 17 666 15 392 −12.9 
Foodweb connectance 1.35 1.36 0.7 
Flow diversity (D/TST) 4.35 4.16 −4.4 

Main environmental change: development of hypoxic conditions from early to late summer.

Although the system's activity increased over summer by ∼4%, the ascendancy, the A/C ratio, the AMI, the redundancy, and the flow diversity index decreased. The decrease in A, the A/C ratio, and the AMI indices which reflect on the inherent organizational status and functioning of a system, all indicate a decrease in the function of the Neuse estuary as the environment changed from oxic to anoxic conditions. A decline in the redundancy points to a reduction in parallel energy-flow pathways and less system stability, whereas a lower flow diversity index indicates a decrease in interactions and a greater degree of unevenness and variability in the flow structure and function of the system.

St Marks Wildlife Refuge, FL, USA

The St Marks Wildlife Refuge (30°06′N 84°11′E) is situated in Apalachee Bay in the northeastern Gulf of Mexico and contains a diverse landscape of wetlands, upland forests, and coastal waters dominated by contiguous seagrass beds of Halodule wrightii. Intensive and quantitative sampling of the pelagic and benthic living communities and species in the refuge were conducted during January and February 1994 at six sites, each covering an area of ∼4.5 ha, representing a subunit of the seagrass ecosystem. Flow models, consisting of 51 compartments each and depicting standing stocks and flows, were constructed for each sampling site for January and February, respectively, and analysed using ENA (Baird et al., 1998). The only environmental variable that differed significantly between January and February was an increase of 2°C in the average water temperature in Apalachee Bay, from 15°C in January to 17°C in February.

The system attributes listed in Table 2 indicate an increase in most attributes. Of interest are the dramatic increases in species diversity by ∼58%, number of cycles (66%), the FCI (29%), and TST (24%), pointing to the development of a much more complex and active system (Baird et al., 1998; Luczkovich et al., 2002). However, energy was less efficiently transferred, as reflected by a decrease of 32% in the system's trophic efficiency, whereas the relative and internal A/C ratios both declined by 10% and 5%, respectively (Table 2). One can infer from these dimensionless A/C ratios the development of a less organized system as the temperature increased over time. Furthermore, the decline in the foodweb connectance index, which expresses the magnitude of effective links between the living components only, implies that an increase in temperature affected the connectedness between the living components of the system (Baird et al., 2004a).

Table 2.

System attributes during January and February of the St Marks Wildlife National Refuge, Apalachee Bay, FL.

System attribute January February % increase +, % decrease – 
Temperature (°C) 12 17 41.7 
Species diversity (H′) 0.742 1.174 58.2 
Total biomass (mg C m−211 601 11 641 0.3 
System production (mg C m−2 d−1431 482 11.8 
P/B (d) 0.037 0.041 10.8 
System trophic efficiency (%) 4.91 3.33 –32.2 
Number of cycles 608 1 006 65.5 
FCI (%) 15.53 20 28.8 
TST (mg C m−2 d−11 877 2 321 23.7 
Ascendancy 3 392 4 210 24.1 
A/C (%) 35.6 32.2 −9.6 
Ai/Ci 39.1 37.1 −5.1 
AMI (A/TST) 1.807 1.814 0.4 
Redundancy 2 767 4 078 47.4 
Foodweb connectance 3.6 3.4 −5.6 
Flow diversity 5.1 5.6 9.8 
System attribute January February % increase +, % decrease – 
Temperature (°C) 12 17 41.7 
Species diversity (H′) 0.742 1.174 58.2 
Total biomass (mg C m−211 601 11 641 0.3 
System production (mg C m−2 d−1431 482 11.8 
P/B (d) 0.037 0.041 10.8 
System trophic efficiency (%) 4.91 3.33 –32.2 
Number of cycles 608 1 006 65.5 
FCI (%) 15.53 20 28.8 
TST (mg C m−2 d−11 877 2 321 23.7 
Ascendancy 3 392 4 210 24.1 
A/C (%) 35.6 32.2 −9.6 
Ai/Ci 39.1 37.1 −5.1 
AMI (A/TST) 1.807 1.814 0.4 
Redundancy 2 767 4 078 47.4 
Foodweb connectance 3.6 3.4 −5.6 
Flow diversity 5.1 5.6 9.8 

Main environmental change: an increase in water temperature of 5°C from January to February.

The mesohaline Chesapeake Bay, MD, USA

Chesapeake Bay is situated along the Atlantic coast of the United States, extending from 36°50 to 39°40′E. The salinity in the mesohaline region varies between 6 and 18, and spans ∼48% (or 5.98 × 109 m2) of the total surface area of the bay, whereas its volume equals 47% (or 3.63 × 1010 m3) of the total volume of the estuary. Temperature is the most variable environmental parameter between seasons, when it ranges from 21.4 to 28.9°C in summer, 13.1 to 23.3°C in autumn, 2.3 to 5.7°C in winter, and 6.2 to 16.7°C in spring (Baird and Ulanowicz, 1989).

Network flow models, consisting of 33 compartments, were developed, representing each of the four seasons (Baird and Ulanowicz, 1989). Differences in system attributes between the seasons are presented in Table 3. Most of the indices indicate highest values during summer, except the daily P/B ratio (highest in spring), and the A/C ratio (lowest in summer and highest in winter). The summer/winter difference in the ascendancy component (i.e. the A/C ratio) can be attributed mainly to elevated dissipation overheads (or respiration) and redundancy during summer, from which it can inferred that the carbon flow dynamics in this system during summer at higher temperatures are both more dissipated and less organized than during the less active, cooler seasons (Baird and Ulanowicz, 1989).

Table 3.

System attributes during the four seasons in the mesohaline region of Chesapeake Bay.

System attribute Spring Summer Autumn Winter 
Temperature (°C) 11.6 25.1 18.2 4.0 
Total biomass (mg C m−28 345 14 767 9 819 7 644 
Production (mg C m−1 season−12 766 4 105 1 696 1 198 
P/B (d) 0.33 0.28 0.17 0.16 
System trophic efficiency (%) 9.6 11.3 12.0 8.1 
Number of cycles 36 58 33 36 
FCI (%) 24.2 23.4 22.4 22.6 
TST (mg C m−2 d−114 168 18 247 8 464 6 607 
Ascendancy 29 761 39 301 17 442 13 566 
A/C (%) 44.9 44.3 47.5 49.4 
Ai/Ci (%) 51.8 52.3 53.7 55.1 
AMI (A/TST) 2.10 2.15 2.06 2.05 
Dissipation overheads 11 588 17 260 6 388 4 592 
Redundancy 18 660 25 153 10 427 7 208 
Foodweb connectance 1.62 1.75 1.77 1.6 
Flow diversity 4.67 4.86 4.33 4.11 
System attribute Spring Summer Autumn Winter 
Temperature (°C) 11.6 25.1 18.2 4.0 
Total biomass (mg C m−28 345 14 767 9 819 7 644 
Production (mg C m−1 season−12 766 4 105 1 696 1 198 
P/B (d) 0.33 0.28 0.17 0.16 
System trophic efficiency (%) 9.6 11.3 12.0 8.1 
Number of cycles 36 58 33 36 
FCI (%) 24.2 23.4 22.4 22.6 
TST (mg C m−2 d−114 168 18 247 8 464 6 607 
Ascendancy 29 761 39 301 17 442 13 566 
A/C (%) 44.9 44.3 47.5 49.4 
Ai/Ci (%) 51.8 52.3 53.7 55.1 
AMI (A/TST) 2.10 2.15 2.06 2.05 
Dissipation overheads 11 588 17 260 6 388 4 592 
Redundancy 18 660 25 153 10 427 7 208 
Foodweb connectance 1.62 1.75 1.77 1.6 
Flow diversity 4.67 4.86 4.33 4.11 

Main environmental change: seasonal temperature fluctuations.

The Kromme River estuary, Eastern Cape, South Africa

The Kromme River estuary (34°09′S 24°51′E) is situated along the southeast coast of South Africa, and discharges into St Francis Bay. The total length of the river is ∼95 km, and it meanders through relatively undisturbed rural areas with virtually no agricultural, industrial, or urban development in the catchment area of ∼936 km2. The mean annual run-off (MAR) from the river into the estuary for the period 1924–1980 was ∼116.8 × 106 m3. A large second impoundment was built and completed in 1984. Since then the MAR averaged at 1.3 × 106 m3, which represents a decrease in freshwater inflow of nearly 99%. The effect of the decrease was the establishment of a virtual homogeneous body of water in the estuary, with salinities rarely dropping below 33, and an axial salinity gradient from mouth to the head resembling that of seawater (Baird and Heymans, 1996). The loss of the typical salinity structure of estuaries with moderate freshwater inflows resulted in a change in the species diversity and in the abundance of organisms in the system.

Two detailed carbon flow models were constructed for the estuary; one representing the pre-1984 period and the other based on data collected 10 years later (1994/1995), after the construction of the impoundment and the subsequent reduction in freshwater inflow. Each model depicts the standing stock and flows between 28 compartments (25 living and 3 non-living; Baird and Heymans, 1996). These data were analysed using ENA, and results from these analyses are given in Table 4.

Table 4.

System attributes of four Eastern Cape estuaries, South Africa.

Attributes Kromme Gamtoos Swartkops Sundays 
Air temperature range all seasons (°C) 12–32 12–32 10–34 10–34 
Water temperature (winter min–summer max; °C) 17–27 17–27 14–28 17–25 
Mean annual freshwater inflow (m3 s−10.07 (s.d. = 1.4, n = 36) 1.02 (s.d. = 0.85, n = 36) 0.82 (s.d. = 0.9, n = 48) 2.74 (s.d. = 1.03, n = 36) 
Salinity range (head to mouth) 32–35 12–35 10–35 3.5–35 
Catchment (km2936 34 500 1 360 20 792 
MAR (106 m3105 485 84 186 
Water exchange time (d) 87 26 34 42 
Total dissolved inorganic N (μM) 5.03 13.42 6.39 11.81 
Total dissolved inorganic P (μM) 0.58 0.33 2.43 0.52 
System production (mg C m−2 d−11 571.4 4 474.5 1 761.8 3 030.3 
Total biomass (g C m−2 d−180.7 89.5 134.1 87.5 
P/B day 0.019 0.050 0.013 0.035 
TST (mg C m−2 d−113 641 23 640 11 809 16 385 
DC (mg C m 2 d−1 bits) 58 883 106 680 50 205 72 692 
Ascendancy (A mg C m−2 d−1 bits) 20 491 46 034 18 893 31 161 
AMI (A/TST) 1.5 1.9 1.6 1.9 
Relative ascendancy (A/DC, %) 34.8 43.2 37.6 42.9 
Relative redundancy (R/DC, %) 39.2 31.0 31.0 27.1 
Relative internal ascendancy (Ai/DCi, %) 33.4 43.2 38.1 44.1 
Number of cycles 895 1 117.0 917 1 209 
Fin cycling index (%) 40.8 28.1 26.2 20.1 
Trophic efficiency (log mean) 1.80 3.20 1.33 2.56 
Flow diversity (DC/TST) 4.32 4.51 4.25 4.44 
Foodweb connectance 1.77 1.99 1.83 2.33 
Overall connectance 2.33 2.03 2.03 1.91 
Attributes Kromme Gamtoos Swartkops Sundays 
Air temperature range all seasons (°C) 12–32 12–32 10–34 10–34 
Water temperature (winter min–summer max; °C) 17–27 17–27 14–28 17–25 
Mean annual freshwater inflow (m3 s−10.07 (s.d. = 1.4, n = 36) 1.02 (s.d. = 0.85, n = 36) 0.82 (s.d. = 0.9, n = 48) 2.74 (s.d. = 1.03, n = 36) 
Salinity range (head to mouth) 32–35 12–35 10–35 3.5–35 
Catchment (km2936 34 500 1 360 20 792 
MAR (106 m3105 485 84 186 
Water exchange time (d) 87 26 34 42 
Total dissolved inorganic N (μM) 5.03 13.42 6.39 11.81 
Total dissolved inorganic P (μM) 0.58 0.33 2.43 0.52 
System production (mg C m−2 d−11 571.4 4 474.5 1 761.8 3 030.3 
Total biomass (g C m−2 d−180.7 89.5 134.1 87.5 
P/B day 0.019 0.050 0.013 0.035 
TST (mg C m−2 d−113 641 23 640 11 809 16 385 
DC (mg C m 2 d−1 bits) 58 883 106 680 50 205 72 692 
Ascendancy (A mg C m−2 d−1 bits) 20 491 46 034 18 893 31 161 
AMI (A/TST) 1.5 1.9 1.6 1.9 
Relative ascendancy (A/DC, %) 34.8 43.2 37.6 42.9 
Relative redundancy (R/DC, %) 39.2 31.0 31.0 27.1 
Relative internal ascendancy (Ai/DCi, %) 33.4 43.2 38.1 44.1 
Number of cycles 895 1 117.0 917 1 209 
Fin cycling index (%) 40.8 28.1 26.2 20.1 
Trophic efficiency (log mean) 1.80 3.20 1.33 2.56 
Flow diversity (DC/TST) 4.32 4.51 4.25 4.44 
Foodweb connectance 1.77 1.99 1.83 2.33 
Overall connectance 2.33 2.03 2.03 1.91 

A measure of the changes in the activity of the individual component is reflected in the compartmental throughput. The components that decreased in throughput were mainly the planktonic component of the ecosystem, namely phytoplankton, free-living bacteria, heterotrophic microflagellates, micro- and macrozooplankton, and carnivorous fish. Throughput through these components decreased collectively by ∼81%. Compartments that increased noticeably in throughput include the benthic components of the system, such as submerged benthic macrophytes (specifically eelgrass, Zostera capensis) by ∼130%, benthic invertebrate suspension-feeders by ∼180%, detritivores by ∼118%, and benthic-feeding fish by ∼150%.

Although no large changes in species composition or in the topology of the system were observed, the biomass and throughput of many compartments either decreased or increased significantly. In the absence of changes in any other physical and/or chemical effects, except freshwater inflow, the observed changes can only be ascribed to the altered salinity regime of the estuary. ENA results indicate that the system underwent changes at the system level, with increases in TST, DC, and A. However, declines in the P/B ratio, in the efficiency of energy transfer in the system, in the FCI, and in the A/C ratios (Table 4) suggest that the level of organization and maturity of the system declined.

Comparisons of system function over spatial scales

Although the comparison of geographically separated ecosystems has received considerable attention (Baird et al., 1991, 2007; Christensen, 1995), the great variability in environmental characteristics and prevailing climate regimes make meaningful deductions of how these systems function on a comparative basis tenuous. It would therefore be difficult to compare indices of system function of disparate ecosystems in the context of climate change. However, analyses of ecosystems falling within the same climate regime and represented by network models of the same topology (structure and number of compartments) may shed light on the functional differences between them. The objective of this section is to examine four estuaries in proximity, which are subject to the same rainfall and temperature climate, but differ significantly in the amount of freshwater each receive from their respective catchments. Climate change may alter some common characteristics, and in this case freshwater inflow as a function of precipitation is considered. Freshwater inflow into estuaries is also greatly affected by human, agricultural, and industrial needs, and the building of impoundments to satisfy these needs influence freshwater inputs, and consequently system function, as illustrated above for the Kromme River estuary.

All four estuaries (Kromme, Gamtoos, Swartkops, and Sundays Rivers) are permanently open, are in the warm-temperate region along the southeast coast of South Africa, and discharge into the Indian Ocean. The distance between the Sundays River (at the most eastern location) and Kromme River estuaries (the most western location) is ∼130 km, with the Swartkops and Gamtoos estuaries, respectively, ∼40 and 90 km west of the Sundays River estuary. A 25-compartment model was constructed for each estuary, consisting of 22 living and 3 non-living components and depicting standing stocks and flows between them. The physical and chemical characteristics and the output results from ENA of each are summarized in Table 4.

The four systems differ widely in their physical and chemical characteristics and in the standing stocks of the living components and their rates of productivity. The rates of freshwater inflows range from 0.7 m3 s−1 in the Kromme Estuary to 2.74 m3 s−1 in the Sundays estuary. The difference in inflow rates is largely as a result of impoundments in all estuaries, although the Sundays River receives water by an inter-basin transfer scheme from the Orange River, which itself discharges into the Atlantic Ocean on South Africa's west coast. Concentrations of dissolved inorganic nutrients (presented in Table 4 as total dissolved N and P) are highest in the Gamtoos and Sundays estuaries, because of intensive agricultural activity in their catchments (Scharler and Baird, 2003a, b, 2005; M. Vosloo, pers. comm.). The water temperature in all four systems falls within a narrow band (Table 4), whereas the salinity range differs significantly between them, which can be ascribed to a variable dilution effect of freshwater inputs (Table 4). Differences in the standing biomass and production rates are probably functions of water quality and nutrient inputs, and there seems to be a correlation between these variables and the daily P/B ratio (Table 4).

The variable considered most influenced by climate change is the rate of freshwater inflow as a function of changing weather and precipitation patterns, accompanied by the quality of the run-off. If we interpret the ENA results simplistically, by considering the freshwater inflow regimes as the prominent variable, the TST, DC, A, the AMI, the A/DC and Ai/DCi ratios, number of cycles, the trophic efficiency, flow diversity, and foodweb connectance are highest in the Gamtoos and Sundays River estuaries, which also have higher inflow rates than the other two estuaries (Table 4). These indices reflect on the systems’ activity, organization, and specialization, and appear to indicate that those systems with relatively higher inflow rates are more active (higher TST and P/B ratios) and more organized (higher A/C ratios, AMI, flow diversity, and foodweb connectance indices).

Conclusions

Results from the few examples discussed above illustrate that relatively small changes in environmental parameters do affect the functioning of an ecosystem over time. When hypoxia develops in the Neuse River estuary during summer, the system attributes that reflect on the function and organization of the ecosystem decline, indicating the development of a less organized system, and a diversion of energy from consumers to microbes. In the St Marks Wildlife Refuge, where the temperature increased from one month to the next, and in Chesapeake Bay, where the temperature fluctuated over seasons, changes in functioning and system attributes were observed. Seasonal changes in temperature, as evidenced in the St Marks Refuge and Chesapeake Bay, are natural phenomena; climate change, and water temperature in particular, may affect their functioning to an even greater extent should the variability of temperature increase, or decrease, beyond the natural range. In the Kromme River estuary, the variability in biomass and productivity of the system appears to be caused by reduced freshwater inflow. This changed the estuary into an arm of the sea with a homogeneous salinity regime, as opposed to the previous typical estuarine condition when a salinity gradient was present. These changes are clearly reflected in the altered system-level properties, and thus in the functioning of the system (Table 4).

Seasonal changes in system function were also observed for the Maspalomas lagoon, Gran Canaria, by Almunia et al. (1999). The observed increase there in the A/C ratio over the year of study was attributed to a shift of resources from the benthic to the pelagic subsystem. Conversely, Ulanowicz et al. (1999) reported little change in whole-system indices between wet and dry seasons of the Florida Bay ecosystem, despite a 32% decline in the system's activity, TST, from the wet to the dry season. In an attempt to examine the link between biodiversity and ecosystem function, Raffaelli (2006) compared ENA indices between an undisturbed foodweb of the Ythan estuary in Scotland to those derived from foodwebs from which some species had been removed. He concluded that the removal of the polychaete, Nereis diversicolor, caused system-level effects.

Mesocosm experiments also established that this species play a significant role in the ecosystem-function process of ammonium release from sediments (Raffaelli, 2006). Although the results revealed a rather tenuous relationship between freshwater inputs and system productivity and organization, they nevertheless provide useful information when used comparatively. Results presented here draw attention to differences between systems with different inflow regimes, and augment the view that estuaries are ecosystems that are sensitive and vulnerable to environmental change. It remains an open question how rainfall patterns will be affected by climate change. Either way, an increase or decrease in precipitation will affect the productivity of individual components in the system, the productivity of the whole system, and the functioning of systems such as these coastal rivers and estuaries. It may also alter the species composition by the emigration and/or extinction of some species, and the immigration of others, and thus affect the biodiversity of the system. It would appear, as demonstrated for the Kromme River estuary, that a reduction in freshwater inflow can change a system to such an extent that it becomes a virtual extension of the sea, and loses its typical estuarine characteristics. Conversely, those systems that receive more freshwater, such as the Gamtoos and Sundays River estuaries, can exhibit higher levels of activity and organization. We can only speculate whether these trends are universal in nature.

The complex interactions between physical, chemical, and biological attributes are not readily understood, but results presented here demonstrated that differences between systems over temporal and spatial scales can be quantitatively assessed using ENA, and can provide the benchmark against which future changes can be measured.

References

Almunia
J. G.
Basterretxea
G.
Aristegui
J.
Ulanowicz
R. E.
Benthic–pelagic switching in a coastal subtropical lagoon
Estuarine, Coastal and Shelf Science
 , 
1999
, vol. 
49
 (pg. 
363
-
384
)
Baird
D.
Müller
F.
Leupelt
M.
Case studies: orientors and ecosystem properties in coastal zones
Eco Targets, Goal Functions, and Orientors
 , 
1998
Berlin
Springer
(pg. 
232
-
242
619
Baird
D.
Allanson
B. R.
Baird
D.
Estuaries as ecosystems: a functional and comparative analysis
Estuaries of South Africa
 , 
1999
Cambridge, UK
Cambridge University Press
(pg. 
269
-
288
340
Baird
D.
Asmus
H.
Asmus
R.
Energy flow of a boreal intertidal ecosystem, the Sylt-Rømø Bight
Marine Ecology Progress Series
 , 
2004
, vol. 
b 279
 (pg. 
45
-
61
)
Baird
D.
Asmus
H.
Asmus
R.
Trophic dynamics of eight intertidal communities of the Sylt-Rømø Bight ecosystem, northern Wadden Sea
Marine Ecology Progress Series
 , 
2007
, vol. 
351
 (pg. 
25
-
41
)
Baird
D.
Christian
R. R.
Peterson
C. H.
Johnson
G. A.
Consequences of hypoxia on estuarine ecosystem function: energy diversion from consumers to microbes
Ecological Applications
 , 
2004
, vol. 
a 14
 (pg. 
805
-
822
)
Baird
D.
Heymans
J. J.
Assessment of ecosystem changes in response to freshwater inflow of the Kromme River estuary, St Francis Bay, South Africa: a network analysis approach
Water SA
 , 
1996
, vol. 
22
 (pg. 
307
-
318
)
Baird
D.
Luczkovich
J.
Christian
R. R.
Assessment of spatial and temporal variability in ecosystem attributes of the St Marks National Wildlife Refuge, Apalachee Bay, Florida
Estuarine, Coastal and Shelf Science
 , 
1998
, vol. 
47
 (pg. 
329
-
349
)
Baird
D.
McGlade
J. M.
Ulanowicz
R. E.
The comparative ecology of six marine ecosystems
Philosophical Transactions of the Royal Society London, Series B
 , 
1991
, vol. 
333
 (pg. 
15
-
29
)
Baird
D.
Ulanowicz
R. E.
The seasonal dynamics of the Chesapeake Bay ecosystem
Ecological Monographs
 , 
1989
, vol. 
59
 (pg. 
329
-
364
)
Baird
D.
Ulanowicz
R. E.
Comparative study on the trophic structure, cycling and ecosystem properties of four tidal estuaries
Marine Ecology Progress Series
 , 
1993
, vol. 
99
 (pg. 
221
-
237
)
Baird
D.
Ulanowicz
R. E.
Boynton
W. R.
Seasonal nitrogen dynamics in Chesapeake Bay: a network approach
Estuarine, Coastal and Shelf Science
 , 
1995
, vol. 
41
 (pg. 
137
-
162
)
Bulling
M. T.
White
P. C. L.
Raffaelli
D. G.
Pierce
G. J.
Using model systems to address the biodiversity-ecosystem functioning process
Marine Ecology Progress Series
 , 
2006
, vol. 
311
 (pg. 
295
-
305
)
Cardinale
B. J.
Palmer
M. A.
Collins
S. L.
Species diversity enhances ecosystem functioning through interspecific facilitation
Nature
 , 
2002
, vol. 
415
 (pg. 
426
-
429
)
Christian
R. R.
Baird
D.
Luczkovich
J.
Johnson
J. J.
Scharler
U. M.
Ulanowicz
R. E.
Belgrano
A.
Scharler
U. M.
Dunne
J.
Ulanowicz
R. E.
Role of network analysis in comparative ecosystem ecology
Aquatic Food Webs: an Ecosystem Approach
 , 
2005
New York
Oxford University Press
(pg. 
25
-
40
262
Christian
R. R.
Boyer
J. N.
Stanley
D. W.
Multi-year distribution patterns of nutrients within the Neuse River estuary, North Carolina
Marine Ecology Progress Series
 , 
1991
, vol. 
71
 (pg. 
259
-
274
)
Christian
R. R.
Thomas
C. R.
Network analysis on nitrogen inputs and cycling in the Neuse River estuary, North Carolina, USA
Estuaries
 , 
2003
, vol. 
26
 (pg. 
815
-
828
)
Christensen
V.
On the behaviour of some proposed goal functions for ecosystem development
Ecological Modelling
 , 
1994
, vol. 
75/76
 (pg. 
37
-
47
)
Christensen
V.
Ecosystem maturity–towards unification
Ecological Modelling
 , 
1995
, vol. 
77
 (pg. 
3
-
32
)
Duffy
J. E.
Biodiversity and the functioning of seagrass ecosystems
Marine Ecology Progress Series
 , 
2006
, vol. 
311
 (pg. 
233
-
250
)
Finn
J. T.
Measures of ecosystem structure and function deriver from analysis of flows
Journal of Theoretical Biology
 , 
1976
, vol. 
56
 (pg. 
363
-
380
)
Halpern
B. S.
Selkoe
K. A.
Micheli
F.
Kappel
C. V.
Evaluating and ranking the vulnerability of global marine ecosystems to anthropogenic threats
Conservation Biology
 , 
2007
, vol. 
21
 (pg. 
1301
-
1315
)
Halpern
B. S.
Walbridge
S.
Selkoe
K. A.
Kappel
C. V.
Micheli
F.
D'Agrosa
C.
Bruno
J. F.
, et al.  . 
A global map of human impact on marine ecosystems
Science
 , 
2008
, vol. 
319
 (pg. 
948
-
952
)
Hooper
D. U.
Chapin
F. S.
III
Ewel
J. J.
Hector
A.
Inchausti
P.
Lavorel
S.
Lawton
J. H.
, et al.  . 
Effects of ecosystem functioning: a consensus of current knowledge and needs for future research
Ecological Monographs
 , 
2005
, vol. 
75
 (pg. 
3
-
35
)
Ieno
E. N.
Solan
M.
Batty
P.
Pierce
G. J.
How diversity affects ecosystem functioning: roles of infaunal species richness, identity and density in marine benthos
Marine Ecology Progress Series
 , 
2006
, vol. 
311
 (pg. 
263
-
271
)
Jørgenssen
S. E.
Patten
B. C.
Straskraba
M.
Ecosystems emerging: toward an ecology of complex systems in a complex future
Ecological Modelling
 , 
1992
, vol. 
62
 (pg. 
1
-
27
)
Kay
J. J.
Graham
L. A.
Ulanowicz
R. E.
Wulff
F.
Field
J. G.
Mann
K. H.
A detailed guide to network analysis
Network Analysis in Marine Ecology
 , 
1989
Berlin
Springer
pg. 
284
 
Luczkovich
J.
Ward
G. P.
Johnson
J. C.
Christian
R. R.
Baird
D.
Nickles
H.
Rizzo
W. M.
Determining the trophic guilds of fishes and macroinvertebrates in a seagrass food web
Estuaries
 , 
2002
, vol. 
25
 (pg. 
1143
-
1164
)
Mann
K. H.
Field
J. G.
Wulff
F.
Wulff
F.
Field
J. G.
Mann
K. H.
Network analysis in marine ecology: an assessment
Network Analysis in Marine Ecology
 , 
1989
Berlin
Springer
pg. 
284
 
Naeem
S.
Expanding scales in biodiversity-based research: challenges and solutions for marine systems
Marine Ecology Progress Series
 , 
2006
, vol. 
311
 (pg. 
273
-
283
)
Odum
E. P.
The strategy of ecosystem development
Science
 , 
1969
, vol. 
164
 (pg. 
260
-
270
)
Paerl
H. W.
Pinckney
J. L.
Fear
J. M.
Peierls
B. L.
Ecosystem responses to internal and watershed organic matter loading: consequences for hypoxia in the eutrophying Neuse River estuary, North Carolina, USA
Marine Ecology Progress Series
 , 
1998
, vol. 
166
 (pg. 
17
-
25
)
Paerl
H. W.
Pinckney
J. L.
Fear
J. M.
Peierls
B. L.
Fish kills and bottom-water hypoxia in the Neuse River estuary: reply to Burkholder et al
Marine Ecology Progress Series
 , 
1999
, vol. 
186
 (pg. 
307
-
309
)
Philippart
C. J. M.
Impacts of Climate Change on the European Marine and Coastal Environment. Ecosystems Approach
2007
Marine Board–European Science Foundation
pg. 
88
 
Raffaelli
D. G.
Biodiversity and ecosystem functioning: issues of scale and trophic complexity
Marine Ecology Progress Series
 , 
2006
, vol. 
311
 (pg. 
285
-
294
)
Raffaelli
D.
Van den Putten
W. H.
Perrson
L.
Wardle
D. A.
Petchey
O. L.
Koricheva
J.
Van den Heijden
M. G. A.
, et al.  . 
Loreau
M.
Naeem
S.
Inckhausti
P.
Multi-trophic processes and ecosystem function
Biodiversity and Ecosystem Functioning
 , 
2002
Oxford
Oxford University Press
pg. 
256
 
Scharler
U. M.
Baird
D.
The influence of catchment management on salinity, nutrient stochiometry and phytoplankton biomass of Eastern Cape estuaries, South Africa
Estuarine, Coastal and Shelf Science
 , 
2003
, vol. 
a 56
 (pg. 
735
-
748
)
Scharler
U. M.
Baird
D.
The nutrient status of the agriculturally impacted Gamtoos estuary, South Africa, with special reference to the river-estuary interface
Aquatic Conservation: Marine and Freshwater Ecosystems
 , 
2003
, vol. 
b 13
 (pg. 
99
-
199
)
Scharler
U. M.
Baird
D.
A comparison of selected ecosystem attributes of three South African estuaries with different freshwater inflow regimes, using network analysis
Journal of Marine Systems
 , 
2005
, vol. 
56
 (pg. 
283
-
308
)
Ulanowicz
R. E.
Growth and Development: Ecosystems Phenomenology
1986
New York
Springer
pg. 
203
 
Ulanowicz
R. E.
Ecology, the Ascendant Perspective
1997
New York
Columbia University Press
pg. 
201
 
Ulanowicz
R. E.
Quantitative methods of ecological network analysis
Computational Biology and Chemistry
 , 
2004
, vol. 
28
 (pg. 
321
-
339
)
Ulanowicz
R. E.
Baird
D.
Nutrient controls on ecosystem dynamics: the Chesapeake mesohaline community
Journal of Marine Systems
 , 
1999
, vol. 
19
 (pg. 
159
-
172
)
Ulanowicz
R. E.
Bondavalli
C.
Egnotovich
M. S.
Network analysis of trophic dynamics in south Florida ecosystems, FY 97: the Florida Bay ecosystem
1999
FL
Annual Report to the United States Geological Service Biological Resources Division, University of Miami Coral Gables
pg. 
65
 
Ulanowicz
R. E.
Kay
J. J.
A package for the analysis of ecosystem flow networks
Environmental Software
 , 
1991
, vol. 
6
 (pg. 
131
-
142
)
Ulanowicz
R. E.
Norden
J. S.
Symmetrical overhead in flow networks
International Journal of System Science
 , 
1990
, vol. 
1
 (pg. 
429
-
437
)
Wulff
F.
Field
J. G.
Mann
K. H.
Network Analysis in Marine Ecology: Methods and Applications. Coastal and Estuarine Studies 33
1989
Berlin
Springer
pg. 
284