Tradeoffs in the Evolution of Plant Farming by Ants
Total Page:16
File Type:pdf, Size:1020Kb
Tradeoffs in the evolution of plant farming by ants Guillaume Chomickia,1, Gudrun Kadereitb, Susanne S. Rennerc, and E. Toby Kiersd aDepartment of Bioscience, Durham University, DH1 3LE Durham, United Kingdom; bInstitute of Molecular Physiology, University of Mainz, 55099 Mainz, Germany; cSystematic Botany and Mycology, Department of Biology, University of Munich, 80638 Munich, Germany; and dDepartment of Ecological Science, Vrije Universiteit Amsterdam, 1081 HV Amsterdam, Netherlands Edited by Joan E. Strassmann, Washington University in St. Louis, St. Louis, MO, and approved December 18, 2019 (received for review November 9, 2019) Diverse forms of cultivation have evolved across the tree of life. coevolved traits, such as practiced by leafcutter or plant-farming Efficient farming requires that the farmer deciphers and actively ants, in which ant farmers appear to actively manipulate condi- promotes conditions that increase crop yield. For plant cultivation, tions to increase yield per unit input. Leafcutter ants, for ex- this can include evaluating tradeoffs among light, nutrients, and ample, mediate their local environment in ways that increase protection against herbivores. It is not understood if, or how, productivity, such as by sustaining large populations of Actino- nonhuman farmers evaluate local conditions to increase payoffs. mycete bacteria in specific crypts that suppress parasites in Here, we address this question using an obligate farming mutu- fungal gardens (17). However, as farmers cannot simultaneously alism between the ant Philidris nagasau and epiphytic plants in the optimize all conditions (e.g., herbivore and pathogen defense, genus Squamellaria that are cultivated for their nesting sites and fertilization, modulating environmental variables), it is an open floral rewards. We focused on the ants’ active fertilization of their question as to whether, and how, insect farmers mediate crop crops and their protection against herbivory. We found that ants tradeoffs. benefited from cultivating plants in full sun, receiving 7.5-fold To answer this question, we studied an obligate farming mu- more floral food rewards compared to shade-cultivated plants. tualism between the ant Philidris nagasau (Dolichoderinae) and The higher reward levels correlated with higher levels of crop pro- epiphytic plants in the genus Squamellaria (Hydnophytinae, tection provided by the ants. However, while high-light planting Rubiaceae), occurring in Fijian rainforests on Taveuni and Vanua yielded the greatest immediate food rewards, sun-grown crops Levu islands (11). P. nagasau workers cultivate multiplant colonies contained less nitrogen compared to shade-grown crops. This of Squamellaria epiphytes that can contain 50 or more individuals was due to lower nitrogen input from ants feeding on floral re- whose plant-formed cavities in modified stems (domatia) house ∼ wards instead of insect protein gained from predation. Despite one queen and 250,000 workers (11, 18). In this mutualism, the EVOLUTION this tradeoff, farming ants optimize crop yield by selectively plant- ants control dispersal, fertilization, and defense of the epiphytes. ing their crops in full sun. Ancestral state reconstructions across this They do this by actively collecting Squamellaria seeds, planting ant–plant clade show that a full-sun farming strategy has existed for them under the branch bark of their host tree, and subsequently millions of years, suggesting that nonhuman farmers have evolved protecting both seedlings and adults from herbivory (11, 18). The the means to evaluate and balance conflicting crop needs to their ants also fertilize these nutrient-limited (soilless) epiphytes by own benefit. defecating on specialized and highly absorptive warts within a plant’s tuberous domatium, which is also the ants’ obligatory nest insect agriculture | ants | symbioses | plants | ant-plant interactions site (11, 18). In return, the ants feed on sugar- and amino acid-rich food rewards produced by the flowers’ nectaries. The food rewards cross the tree of life, organisms have evolved the ability to and the domatia, which contain a complex network of inter- Acultivate or “farm” individuals of other species (1–6). This connected cavities (11, 18–21), are key to the ants’ survival. Sim- can involve habitual planting, husbandry, and harvesting of ilar to the other “true” agricultures in attine ants, termites, and “crops,” such as seen in social amoebae rearing bacteria (1), ma- rine snails propagating fungi (2), and damselfish cultivating sea- Significance weed (3). True agriculture, as defined by four key steps—namely habitual planting, cultivation, harvest, and dependence of the In human cultivation systems, farmers increasingly use technol- farmer on the crop (7)—is restricted to social insects (ants, termites, ogy to gather data for evaluating tradeoffs between diverse— beetles) cultivating fungi (7–10) and ants cultivating plants (11). and sometimes conflicting—crop requirements to maximize True agriculture is also practiced by humans and generally in- yield. Some social insects have also evolved agricultural prac- volves the farmer deciphering and actively promoting the condi- tices, but it is unknown how they evaluate local conditions to tions that increase yield. For plant crops, this can involve evaluating balance conflicting crop requirements. In the obligate farming tradeoffs among requirements for light, nutrients, and protection symbiosis between ants and plants in Fijian rainforests, we show against herbivores. The ability to evaluate tradeoffs is important how ant farmers also face key tradeoffs in crop cultivation. because farmers, both human and nonhuman, are unlikely to find While ants cannot simultaneously maximize all services to their conditions in which all crop needs are optimally and consistently crops, our work demonstrates that they cultivate crops in high- met. For example, increasing nutrient conditions may inadvertently light conditions to maximize floral food rewards, despite the increase herbivore pressure (12), whereas increasing crop densities nitrogen costs of this strategy. Evaluation of crop tradeoffs plays may negatively affect light conditions for individual plants (13). a key role in the evolution of farming strategies. While in human farming, technology and modeling are in- creasingly employed to evaluate complex tradeoffs (14, 15), it is Author contributions: G.C. and E.T.K. designed research; G.C., G.K., and E.T.K. performed not understood if, or how, nonhuman farmers evaluate local research; G.C., S.S.R., and E.T.K. contributed new reagents/analytic tools; G.C. analyzed conditions to increase payoffs. There are cases of early insect data; and G.C., S.S.R., and E.T.K. wrote the paper. agriculture in which tradeoffs are largely unresolved, and conflict The authors declare no competing interest. can emerge between farmers selecting for edible hyphae and This article is a PNAS Direct Submission. fungal crops allocating resources to reproductive structures, such Published under the PNAS license. as mushrooms (16). Such conflicts, as seen between phyloge- 1To whom correspondence may be addressed. Email: [email protected]. netically basal attine ants and their basidiomycete fungal crops, This article contains supporting information online at https://www.pnas.org/lookup/suppl/ can result in low-productivity farming (16). In contrast, there doi:10.1073/pnas.1919611117/-/DCSupplemental. are cases of agriculture involving a suite of adaptations and www.pnas.org/cgi/doi/10.1073/pnas.1919611117 PNAS Latest Articles | 1of9 Downloaded by guest on September 25, 2021 ambrosia beetles (7), there is an obligate dependence of the Results and Discussion farmer on the crop (11, 18). Squamellaria crops are likewise ob- Crop Productivity Varies along a Light Gradient and Correlates with ligately dependent on the farmer (11, 18). Herbivore Defense Levels. First, we asked how crop productivity, as By controlling dispersal, fertilization, and defense of the crop, measured by food rewards to ants, varied along a natural light the farmer could potentially maximize crop productivity by mod- gradient. Mature Squamellaria flowers provide P. nagasau workers ulating growing conditions, but this remains poorly understood. with a sugary sap rich in sucrose and amino acids that is only Specifically, it is unknown if farming strategies are tailored to accessible to this ant species (21). We quantified food reward particular environmental conditions. Because of the challenges (number of postanthetic nectaries; ref. 21) on n = 133 shoots in empirically manipulating most farming mutualisms, it is dif- belonging to 50 Squamellaria wilsonii plants spanning a light gra- ficult to test how demands for light, nutrients, and protection are dient in Taveuni rainforest canopies categorized as full shade, evaluated by insect farmers. Here, we address this challenge by midlight-exposed, and full sun, controlling for shoot size (Materials studying the P. nagasau–Squamellaria farming mutualism under and Methods). different natural light conditions. Our aim is to test if ants op- We found that food rewards were 7.5-fold higher in plants timize farming and defense of their crop to match local envi- cultivated in full sun than in plants cultivated in full shade (Fig. ronments. Using productivity metrics, behavioral assays, and 1A): higher light levels consistently led to more flowers and stable isotope analyses, we asked 1) whether crop productivity (i.e., hence food