How can a pollination network show relationships among plants and pollinators?

It shows recorded connections between two kinds of organisms: pollinators and plants. In a bipartite network, those groups appear in separate columns or lists, with a line joining a pollinator to a plant when an interaction was observed. A USGS report uses this format, with bee taxa on one side and host-plant species on the other; its figures represent interactions recorded in the specified surveys. [S1]

For a small, made-up example, imagine observations at one garden during one weekend:

Pollinators           Plants
Bee A  ─────────────   Clover
       └────────────   Aster
Bee B  ─────────────   Aster
Bee C  ─────────────   Clover

This example says Bee A was recorded interacting with clover and aster, while Bee B and Bee C were each recorded with one of those plants. It does not, by itself, tell you how often an interaction happened, whether pollen was transferred, or whether the interaction resulted in successful pollination. Those details would need separate observations and measurements.

First decide the scope: for example, one garden, a defined set of observation days, and the pollinators and plants you can identify reliably. Then make a two-column list and add a connection only when your record supports that pair.

A simple working table could look like this:

Pollinator | Plant observed
Bee A      | Clover
Bee A      | Aster
Bee B      | Aster
Bee C      | Clover

Turn each distinct pollinator–plant pair into one line in the diagram. If you count repeated visits, preserve those counts separately or show line thickness only if you clearly explain what thickness means. A plain line is easier to read and avoids implying that every connection is equally frequent when you have not measured frequency. Label the diagram with its place and observation period so readers can see what its scope is.

Look for patterns in the recorded links, not conclusions the diagram cannot support. You might notice that one observed pollinator connects to several plants, that several pollinators connect to the same plant, or that some recorded pairs are unique. These patterns can help describe how interactions were distributed in your observations.

For example, in the sample, Aster connects to Bee A and Bee B, while Bee B has only one recorded connection. That is a description of this small record, not proof that Bee B depends on aster or that aster is generally more important. The USGS report describes network patterns in relation to its surveys and study scope; its examples do not establish all possible interactions outside those observations. [S1]

A missing line means no link appears in the records represented by that diagram. It does not establish that the pollinator and plant never interact. They may not have been observed together, the interaction may have been missed, or the study may not have covered the relevant place or time. The source report specifies the region, programs, and 2016–17 observation period for its examples, and notes distinctions in how some bee records were collected and identified. [S1]

A useful caption is: “Lines show plant–pollinator interactions recorded during [place and observation period]. The diagram summarizes these observations; unshown pairs are not confirmed absences.” Add how you identified organisms and whether line width or other visual features encode anything. A network is a map of the recorded evidence, not a complete inventory of nature.