Ecosystems are complicated and macroscopic living systems with self-organization properties. Ecosystem components are interconnected, interrelated, and form a network. The structure of ecosystem networks determines the makeup of adjacent (direct) and non-adjacent (indirect) pathways over which energy, matter, and information flow. Ecological network analysis (ENA) is a system-oriented methodology used to analyze within-system interactions, which helps researchers identify holistic properties that are otherwise not evident from empirical observations. ENA is introduced in the paper as a promising approach for studying ecosystems. We outline some of the main achievements of ENA, which include network structure, network stability, network ascendancy, network utility and so on. We introduce the steps for one approach in constructing network models and in giving rules for modeling a community. The first step in ENA is to identify the system of interest and place boundaries around it. The second step is to make a list of the major or functional groups in the ecosystem. Third, a unit of currency for the network is selected. Fourth, the adjacency matrix is constructed to document any possible flow interactions. Fifth, field data are obtained related to inputs, outputs and throughflows. Sixth, ENA is applied to the network. Seventh, a sensitivity analysis is conducted. We also provide a brief overview of the algorithmic methods used to construct ecological network typologies. We chose a seven-compartment model of nitrogen flow in the Neuse River Estuary, United States, as an example to explain how D. K. Gattie developed insight into the essence of microdynamic environ flows in an ecological network. The ENA method contributed three significant achievements related to ecosystem study. First, ENA uses a mathematical model and derivations to study the interconnective relationships between different compartments within ecosystem. Second, ENA is a promising method for analyzing microdynamic environ flows in an ecological network. This method indicates indirect effects from non-adjacent network relationships are very important for an ecosystem, and allows exploration of the indirect mechanisms that maintain an ecological network in a steady-state. Third, ENA not only is a system-oriented methodology used to analyze within-system interactions and used to identify holistic properties, but also provides a scientific method different than that built on Newton's principle. Although ENA has evident advantages useful in analyzing ecosystem properties, difficulties remain in constructing ecosystem networks. First, ENA often requires a substantial amount of data and sometimes the data are not available. Second, it is generally difficult to establish the steady-state of a real system based solely on field data. Third, no standard method exists for delimiting an ecosystem, which leads to errors during network analysis. The method has an inherent level of difficulty because ecosystems are intangible by nature since they have no actual boundaries. Our goals are to help other researchers who are considering the construction of network models as well as assist them in further refining the ENA method.