What happens when you spend an entire year watching hummingbirds and the plants they visit in one of Brazil’s most unique ecosystems? You begin to unravel the complex, ever-changing relationships that tie together birds, flowers, and the environment they share.

In a new study jointly led by Steffani Queiroz and Marsal Amorim — and part of my ongoing collaboration with a brilliant team of Brazilian hummingbird researchers — we explored how plant–pollinator interactions shift over time in the Campo Rupestre, a montane tropical ecosystem rich in biodiversity and endemic species. Despite the region’s relatively stable climate, we found that the relationships between hummingbirds and flowers are anything but static.

Over the course of 624 hours of observation spread across a full year, we recorded over 9,000 hummingbird visits involving nine bird species and 47 plant species. Many of these plants — and one of the most frequent visitors, the stunning Hyacinth Visorbearer (Augastes scutatus) — are found nowhere else on Earth.

Our goal was to understand how the structure of this ecological network — which plants interact with which hummingbirds, and how often — changes over time, and what drives those changes. Are they shaped by morphological fit (the match between beak and flower shape)? By phenology (when plants bloom)? By nectar characteristics such as the amount produced and its sugar concentration?

What we discovered is that different factors dominate in different seasons. During the rainy season, when hummingbirds are more abundant, interactions were shaped mostly by morphological matching — suggesting that competition leads to greater niche partitioning. In contrast, during the dry season, the network became sparser and was more influenced by nectar sugar content and flowering patterns.

Interestingly, while the overall annual network wasn’t especially nested (a common pattern in mutualistic networks), it was highly modular — meaning that it contained distinct clusters of species that mostly interacted among themselves. This structure changed significantly across months, highlighting the dynamic nature of tropical plant-pollinator interactions, even in environments with relatively little climate variation.

This work highlights the importance of long-term, fine-scale studies in uncovering how interactions among species shift through time. It also underscores the remarkable biodiversity and ecological complexity of the Campo Rupestre — and the need to understand and protect it.

Here’s the reference – if anyone wants a copy, drop me a message via my Contact page:

Queiroz, S.N.P., Amorim, M.D., Lopes, S.A., Vizentin-Bugoni, J., Jorge, L.R., Ollerton, J., Santos, T. & Rech, A.R. (2025) Temporal dynamics of a Neotropical plant-hummingbird interaction network. Austral Ecology 50:e70089

And here’s the full abstract:

Species interaction networks are expected to vary following temporal changes in the environment and the composition of the local community. However, there are still gaps in our knowledge about temporal variation in networks in tropical areas, where less variable climates are expected to produce more stable community structures over time. Here we describe a plant-hummingbird network in the Brazilian Campo Rupestre ecosystem and investigate multiscale temporal variation of interactions in this community as well as the possible mechanisms underlying the frequencies of species interactions. Plants visited by hummingbirds were observed monthly for a year and each species had morphology, phenology and nectar traits measured. During 624 h of observation we recorded nine hummingbird species visiting 47 plant species, amounting to 9015 visits to flowers. Most plants (28 species) were endemic to the Campo Rupestre and mostly visited by the also endemic hummingbird Augastes scutatus (the Hyacinth Visorbearer). The annual network was not nested but presented high modularity and intermediate specialisation. While the overall (annual) frequencies of interaction were primarily defined by morphological matching and phenological overlap, we found a remarkable temporal change in community structure over the year, with different processes underlying interactions among plants and hummingbirds at different seasons. The interaction pattern during the rainy season was more similar to the annual network than the dry season (when nectar sugar content and plant phenology were also important), with more links per species and lower specialisation. The higher importance of morphology to predict interactions during the rainy season suggests higher niche partitioning when more hummingbird species are present in the community. Our results exemplify the importance of considering the temporal dynamics of the community to advance the understanding of the processes defining species interactions over time in the tropics.

My sincere thanks to Sinzinando Albuquerque-Lima for the photograph, which was taken in the Amazon, not where the research described above was conducted.