Tag Archives: Networks

Plant-pollinator networks in the tropics: a new review just published.


As an ecologist who has carried out field work in the temperate zone (UK), the subtropics (Tenerife and South Africa) and the tropics (parts of South America, Africa and Australia)  I’ve always found the idea that the study of ecology can be divided into “tropical” and “non-tropical” a bit odd.  It’s as if the way that the natural world works somehow changes at about 23 degrees north or south of the equator, making things “different” around the equator.  The tropics are a very special, diverse place, it’s true, but so are many places outside the tropics.

With this in mind I was pleased when I was asked by some of my Brazilian colleagues to contribute to a chapter in a new book entitled Ecological Networks in the Tropics. It was an opportunity to review what is known about plant-pollinator networks in the tropics and the ways in which they are very similar to such networks at lower latitudes. Here’s the details of the chapter, followed by the abstract.  If anyone wants a copy please drop me an email:

Vizentin-Bugoni J, PKM Maruyama, CS Souza, J Ollerton, AR Rech, M Sazima. (2018) Plant-pollinator networks in the tropics: a review. pp 73-91 In Dáttilo W & V. Rico-Gray. Ecological networks in the Tropics. Springer.


Most tropical plants rely on animals for pollination, thus engaging in complex interaction networks. Here, we present a global overview of pollination networks and point out research gaps and emerging differences between tropical and non-tropical areas. Our review highlights an uneven global distribution of studies biased towards non-tropical areas. Moreover, within the tropics, there is a bias towards the Neotropical region where partial networks represent 70.1% of the published studies. Additionally, most networks sampled so far (95.6%) were assembled by inferring interactions by surveying plants (a phytocentric approach). These biases may limit accurate global comparisons of the structure and dynamics of tropical and non-tropical pollination networks. Noteworthy differences of tropical networks (in comparison to the non-tropical ones) include higher species richness which, in turn, promotes lower connectance but higher modularity due to both the higher diversity as well as the integration of more vertebrate pollinators. These interaction patterns are influenced by several ecological, evolutionary, and historical processes, and also sampling artifacts. We propose a neutral–niche continuum model for interactions in pollination systems. This is, arguably, supported by evidence that a high diversity of functional traits promotes greater importance of niche-based processes (i.e., forbidden links caused by morphological mismatching and phenological non-overlap) in determining which interactions occur, rather than random chance of encounter based on abundances (neutrality). We conclude by discussing the possible existence and direction of a latitudinal gradient of specialization in pollination networks.

Polinode – a user-friendly tool for visualising ecological networks

Birds mixed flocks (2) - curvedIt is a general ecological rule that no species exists in isolation; all species interact with other organisms within the communities to which they belong. The collection and analysis of ecological interaction data has burgeoned over the past couple of decades, particularly in my own area of (largely) mutualistic species interactions such as plant-pollinator relationships – see for example this recent post on hummingbird-plant networks.

There are a number of software packages available for analysing and visualising this type of data, including bipartite  and foodweb in R, Food Web Designer, and Gephi.  Tools such as this vary in their flexibility and in the investment of time required to produce good quality graphics, and ultimately it’s down to personal preferences which you use.

Recently I discovered some very user-friendly network visualisation software that is browser/cloud based, free to use (at least the basic version), very flexible, intuitive and quick to learn. Ideal if you are pressed for time and want to generate some quick food webs.

The system is called Polinode and was developed primarily to visualise business and social science data (the “poli” part is nothing to do with pollination, that’s purely coincidental). However there’s no reason why it can’t be used for ecological data, as the image above demonstrates. This is a visualisation of mixed-species flocks of birds feeding together and alone on a local urban park that I’ve discussed previously.  The thickness of the line is proportional to number of interactions observed, and the size of each node is proportional to the number of birds.  Both are scalable in Polinode.

One could also present these data as a straight-line graph, without the loops to indicate single-species feeding:

Birds mixed flocks (5)

As well as these types of networks it’s also possible to produce bipartite (what Polinode terms “hierarchical”) graphs, for example this network of bumblebees feeding on different plant families in a British grassland (click for a closer view – I realised afterwards that I downloaded a rather small version):

Bombus hypnorum with plant families

The system is very flexible and nodes can be grabbed and moved around (as I did above to offset the plant family nodes), recoloured, resized, text resized, etc.

Polinode also calculates a range of network metrics such as degree and Louvain communities (a measure of modularity) which is more limited than some ecologists might require, but which is a good starting point for those new to ecological network statistics.

Data files can be uploaded directly from Excel, and there are example templates showing how to lay out the data.  There is also ample online support including written guidance, videos, and a regular blog. Even in the few months I’ve been playing with the system the developers have added more features, including a graphing facility that generates column and scatter plots from your networks.

There you go, that’s an introduction to Polinode for ecologists; hope it’s useful for your work.