Whether an ecological community is controlled from above or below remains a popular framework that continues generating interesting research questions and takes on especially important meaning in agroecosystems. We describe the regulation from above of three coffee herbivores, a leaf herbivore (the green coffee scale, Coccus viridis), a seed predator (the coffee berry borer, Hypothenemus hampei), and a plant pathogen (the coffee rust disease, caused by Hemelia vastatrix) by various natural enemies, emphasizing the remarkable complexity involved. We emphasize the intersection of this classical question of ecology with the burgeoning field of complex systems, including references to chaos, critical transitions, hysteresis, basin or boundary collision, and spatial self-organization, all aimed at the applied question of pest control in the coffee agroecosystem.

Mexico is one of the most biodiverse countries in the world, with an important proportion of endemism mainly because of the convergence of the Nearctic and Neotropical biogeographic regions, which generate great diversity and species turnover at different spatial scales. However, most of our knowledge of the Mexican ant biota is limited to a few well‐studied taxa, and we lack a comprehensive synthesis of ant biodiversity information. For instance, most of the knowledge available in the literature on Mexican ant fauna refers only to species lists by states, or is focused on only a few regions of the country, which prevents the study of several basic and applied aspects of ants, from diversity and distribution to conservation. Our aims in this data paper are therefore (1) to compile all the information available regarding ants across the Mexican territory, and (2) to identify major patterns in the gathered data set and geographic gaps in order to direct future sampling efforts. All records were obtained from raw data, including both unpublished and published information. After exhaustive filtering and updating information and synonyms, we compiled a total of 21,731 records for 887 ant species distributed throughout Mexico from 1894 to 2018. These records were concentrated mainly in the states of Chiapas (n = 6,902, 32.76%) and Veracruz de Ignacio de la Llave (n = 4,329, 19.92%), which together comprise half the records. The subfamily with the highest number of records was Myrmicinae (n = 10,458 records, 48.12%), followed by Formicinae (n = 3,284, 15.11%) and Ponerinae (n = 1,914, 8.8%). Most ant records were collected in the Neotropical region of the country (n = 12,646, 58.19%), followed by the Mexican transition zone (n = 5,237, 24.09%) and the Nearctic region (n = 3,848, 17.72%). Native species comprised 95.46% of the records (n = 20,745). To the best of our knowledge, this is the most complete data set available to date in the literature for the country. We hope that this compilation will encourage researchers to explore different aspects of the population and community research of ants at different spatial scales, and to aid in the establishment of conservation policies and actions. There are no copyright restrictions. Please cite this data paper when using its data for publications or teaching events.

Understanding the ecology of mutualisms becomes a particularly important task when considering agroecosystems, as many ecosystem services are associated with mutualistic interactions. Here we report on experiments associated with an indirect pest control mutualism between the arboreal nesting ant Azteca sericeasur and coffee. This system is particularly interesting because the indirect Azteca-Coffea mutualism emerges from an Aztecascale insect mutualism that takes place on the coffee plant. We describe this interaction structure as a mutualismdependent mutualism and ask whether the density of intermediary mutualist (scale insects, Coccus) that benefits Azteca also influences the benefits provided to coffee plants. We found that indeed Azteca's benefit to Coffea is consistent when Coccus density is high. Furthermore, we also found that at low Coccus density Azteca only benefits Coffea in the beginning of the rainy season, and this effect is likely due to the fact that Coccus produces less sugars with higher precipitation. We suggest a framework for thinking about context-dependency in agroecosystem mutualisms that may provide a more mechanistic way to tease apart the prevalent contextdependent results in ecological literature. Finally, we address some past recommendations as it pertains to the management of the Azteca-Coffea-Coccus complex in coffee agroecosystems.

A team of 79 scientists from more than 50 institutions partnered to gather all available information regarding Mexican ants since 1894, the year in which the first geographical record of an ant is known for the country. In this new study, including more than 21,000 records, we showed that there are ~900 species of ants in Mexico, which are distributed unevenly across the country. Mexico is one of the relatively few countries along the interface of the tropical and temperate zones, and is a megadiverse country.