Research proposal: Predicting coevolutionary patterns on a global scale

Aline S. Guidolin 1, Andrea Paz 2, Andrea Sanchez-Meseguer 3, Diana Ribas Rodrigues 4, Emerson A. S. Bezerra 5, Emilio Lanna 6, Samuel C. Faria 1, Sherif M. Elnagdy 7

1 Universidade de São Paulo, Brazil, 2 Universidad de los Andes, Colombia, 3 Real Jardín Botánico-CSIC, España, 4 Universidade Federal ABC,5 Universidade Federal da Paraíba, Brazil, 6 Universidade Federal de Bahia, Brazil, 7 Cairo University, Egypt


Since Darwin's and Wallace's time, scientists have been discussing the primacy of species interactions as drivers for adaptive evolutionary change and evolutionary diversification. Species constantly engage in strong interactions with other species - parasites, predators, preys, and mutualists. As a result, traits may coevolve and diversify in geographic mosaics. However, the term coevolution was only introduced in 1964 by Ehrlich and Raven [1] discussing interactions between butterflies and their host plants. A broad definition in which there appears to be a consensus is that coevolution involves reciprocal evolutionary changes in two or more interacting entities. Nonetheless since the term was introduced without a clear definition, there has been a continuous argument on what is meant by coevolution and what can or cannot be considered as an example of it [2,3]. The importance of defining coevolution is controversial since some researchers have viewed it as a major factor shaping the world’s biota [3,4,5,6] while others have pointed out that its role may have been overestimated [2,7].

Local coevolutionary selection is an important building block of coevolutionary dynamics, still studies conducted on a wide range of species interactions in the last few decades, demonstrated that coevolution is inherently geographically structured. Thus, hypotheses were drawn that coevolution between two species can progress along very different trajectories in different places, causing trait differences to evolve among different populations of the same species. It is thought that most species are collections of genetically variable populations distributed among environments, and those environments often differ in abiotic factors and/or community composition. Accordingly, the pattern and strength of natural selection imposed by species on each other traits may to be highly variable among those environments. The concept of coevolution as an inherently geographically-structured process was identified by John Thompson as the Geographic Mosaic Theory of Coevolution (GMTC) [8]. However, the geographical aspect of the coevolutionary phenomenon has not been fully addressed, especially concerning global patterns of coevolution. It is known that species diversity covaries with the latitudinal gradient [9]. The Latitudinal Gradient of Diversity (LGD) indicates that lower latitudes would exhibit higher species diversity than higher latitudes. There are different hypothesis to explain this gradient (some of them are interdependent) still LDG is usually associated with geographic, climatic and photosynthetic constraints [10]. Biotic interactions are also believed to play a role in the origin and maintenance of species diversity. There are multiple hypotheses that link the LDG with a presumed gradient in the biotic interactions [11]. Several studies failed to observe consistent changes in ecological interactions with latitude [12,13]. However, some studies about land habitats speculated that the intensity of different interactions may be greater in the tropics [14,11], but this speculation was not fully tested in a statistical framework. Moreover, little information was found in any marine environments.

Mutalism has been known for its potential role in structuring communities [e.g., 15] and promoting coevolution [4]. It is defined as the interaction between individuals of different species that benefit both partners. It has being the focus of many empirical studies for its abundance, diversity. Some species can live without their mutualistic partners and so the relationship is called facultative mutualism. Other species are dependent upon the mutualistic relationship that they cannot live in its absence. Such relationship is an obligate mutualism. The latter kind of mutualism is one of best coevolution examples, since it affects fitness of both entities involved. Indeed, interspecific mutualistic interactions represent some of the most important and widely studied interactions in ecology. Under the global heading of “mutualism”, come partners and interactions such as diverse as hummingbirds and the flowers they pollinate, gut symbionts in the digestive tracts of animals, ants that protect plants from herbivory, and mycorrhizal fungi. Mutualistic interactions are also ubiquitous geographically and evolutionarily, with mutualistic partners from all kingdoms and in all ecosystems [16]. Determination of the benefit gained by each mutualistic partner is necessary to determine whether the interaction is mutalistic or not. The nature of an interaction is important since one partner may cheat the other, and accordingly, the interaction is not a mutualism. One common example of “cheating” occurs when an insect obtains nectar from flowers without pollinating these flowers [17]. The cheater insect consumes nectar, so this interaction is called nectivory (a specialized form of herbivory) and the cheater is often referred to as a "nectar thief." These interaction example can be tricky in the analysis and one way of resolve this problem is to understand the mutualism and evalute each study case.

Moreover, although there is general agreement about the influence of abiotic factors in shaping coevolutionary associations in strength and form [11,18], but Schmeske et al [11] suggested that the question remains whether ‘coevolutionary hotspots’ are promoted by environmental factors or are just a consequence of the latitudinal gradient itself [11]. If there was a latitudinal gradient of mutualism interactions, our second objective will be to differentiate between environmental effects and the diversity gradient. We aim to address these questions using a meta-analysis approach in a statistical framework that will include examples of mutualism associations that are available in the literature.


Figure 1. Hypotheses on the patterns of coevolutionary associations distributions on a global scale. Blue area indicates the hypothesis of a lower number of coevolutionary associations around the equator; red area indicates the hypothesis of high number of coevolutionary events around the equator.


1. To assess if there is a latitudinal gradient of coevolutionary interactions in both marine and land environments. Among all possible interactions we are going to focus on mutualism, a specific association in which both species derive a fitness benefit.

2. If there was a latitudinal gradient of mutualism interactions, our second objective will be to differentiate between environmental effects and the diversity gradient. We aim to address these questions using a meta-analysis approach in a statistical framework that will include examples of mutualism associations that are available in the literature.


Number of coevolutionary events (Response variable)

Our analyses will be based on the data present in the literature. We will perform an exhaustive search for published papers using three different databases ( ISI Web of Science, GoogleScholar, JSTOR), using the keywords “coevolution”, “mutualism”, “latitudinal gradient”, “latitude AND interaction”, “temperature AND biotic interactions” and “climate AND coevolution” . We will also analyze the references cited in these papers to expand our dataset. Then, the quality of the data will be assessed, and only papers showing clear evidence for coevolutionary events will be used. The total number of observations (TNO) might be potentially biased by different causes. The number of studies per region is probably uneven across the different latitudes as it is the number of studies per lineage. Among all potential biases of the study, the phylogenetic bias might be especially significant as it may produce an overestimation of the number of observations. In the example below (Fig. 2) the mutualistic association that occurred in the ancestral node promoted diversification in lineage 1 and 2, with the result of four new mutualistic interactions in the descendant lineages. However, these four new interactions cannot be considered as totally independent observations [19]. The best way to account for the phylogenetic dependence bias is to establish a range in which a minimum and maximum number of observations will be considered in the analysis. In the example above, four mutualistic events is the maximum number while one is the minimum number of observations (Fig. 2). However, we expect that molecular phylogenies would be only available for a small fraction of the observations [11]. Additionally, ancestral area reconstructions would be performed to assess the geographic origin of the association. Therefore, only the maximum number of observations will be considered in this project. As previously explained, this might produce an overestimation of the mutualistic interactions, but we expect that this bias to be equally distributed across the whole latitudinal gradient.
Accordingly, we are proposing to normalize our data as a proportion of mutualistic events per species richness (NTNO/NTOTALof species) in the studied region. We will compare this proportion with other two normalizing strategies: a) TNO divided per the total number of papers published for each climatic zone of the globe, and b) To account for the unobserved mutualistic events (negative observations) the TNO number will be divided by the number of existing interactions that are not considered mutualistic. For this latter case we will focus on the association between insects and flowers in pollination events - pollinators vs. non-pollinators interactions among insects and flower plants.


Figure 2. Schematic representation of the phylogenetic dependence observation bias.

Environmental data (explanatory variables)

We are assuming that mutualistic organisms share the geographic distribution. For every single mutualistic interaction, we will record coordinates of the centroid of that distribution, and this will constitute a single observation. The localities for which coevolutionary interactions are found will be georeferenced and mapped, and the latitude data for each mutualistic interaction event will be recorded in order to contrast with the latitudinal diversity gradient (LDG). We will compile a set of environmental variables stored as geographic information systems (GIS) layers using data publicly available. For marine environments we will use layers available at the N.O.A.A.database (National Oceanic and Atmospheric Administration: that includes climatic and productivity variables specifically we will use information for temperature, pH, dissolved organic matter (DOM), particulate organic matter (POM), primary productivity (clorophyl a) and bacterial abundance. For terrestrial environments we will use the 19 bioclimatic variables derived from temperature and precipitation data available at the WorldClim database [20] and altitude information obtained from a digital elevation model [21]. Data for all georeferenced locality points will be extracted from these GIS layers to derive explanatory variables for this study.

Statistical analyses

The association between TNO and the latitude will be tested with a regression analyses by generalized least squares (GLS), showing the slope and the level of significance (p-value) [22].
To determine whether any parameter or combination of parameters explains the variation observed in the NCE, we will run the stepwise general linear models using a corrected Akaikes’s information criterion (AICc) [22,23]. We want to evaluate the magnitude of the contribution of each variable included within our model. To compute a value for the effect size of our variables we will use partial Eta squared values, this number ranges from 0 to 1 with 0 meaning no effect on the response variable [24].

Expected results and implications

The importance of the environment in shaping ecological associations has been assessed in several groups and interaction types as diverse as plant-pollinator and predator-prey [e.g., 25, 26]. Some studies also establish that coevolutionary interactions are strongly shaped by abiotic factors [18].

If coevolutionary associations are promoted by environmental factors then climate change is prone to alter the existence and mechanisms of these associations [18]. It is important to make an extensive research on how environmental factors are affecting coevolutionary dynamics and thus how these dynamics might be affected in the future [18].

Understanding environmental features that drives coevolutionary interactions is essential to effectively generate better conservation area plans in a global scale. The potential of our pattern findings to address organism’s interactions is particularly important in the face of global climate change, which is coupled with accelerating habitat loss and degradation [29].

Aside the extinction probabilities, species should not be equally protected. Instead, evolutionary processes render each species unique with a characteristic history that can be quantified for the purpose of conservation prioritization [28]. Assessments that integrate phylogenetic distinctiveness and extinction threat have been performed mainly for mammalian groups, drawing attention to extraordinary species from lesser known localities and lineages [27]. However, it is not easily found in the literature conservation plans that considerate coevolved species as an important characteristic to increase their level of extinction threat.

Time table

Activity Month1 Month2 Month3 Month4 Month5 Month6 Month7 Month8 Month9 Month10 Month11 Month12
Review of the literature and construction of database X X X X - - - - - - - -
Collection of environmental data - - - - X X - - - - - -
Collection of organisms interaction data - - - - - - X X - - - -
Statistical analysis - - - - - - - - X X - -
Manuscript writing - - - - - - - - - - X X
1. Ehrlich PR and Raven PH (1964) Butterflies and plants: a study in coevolution. Evolution 18: 586–608.
2. Janzen DH (1980) When is it coevolution? Evolution 34: 611–612.
3. Thompson JN (1989) Concepts of Coevolution. Trends in Ecology and Evolution 4 (6):179-183
4. Thompson, JN (1994) The Coevolutionary Process. Chicago, IL: University of Chicago Press.
5. Grimaldi D (1999) The co-radiations of pollinating insects and angiosperms in the Cretaceous. Ann. Mo. Bot. Gard. 86:373-406
6. Janzen DH (1983) The natural history of mutualism. Pp 40-99 in D.H. Boucher, ed. The biology of mutualism: ecology and evolution, Croom Helm, London.
7. Schemske DW (1983) Limits to specialization and coevolution in plant animal mutualisms. Pp. 67-109 in M.H Nitecki, ed. Coevolution. University of Chicago Press, Chicago, IL.
8. Thompson, JN (2005) The Geographic Mosaic of Coevolution. Chicago, IL: University of Chicago Press.
9. Fischer AG (1960) Latitudinal variations in organic diversity. Evolution 14:64–81
10. Cardillo, M., C. D. L. Orme, and I. P. F. Owens (2005) Testing for latitudinal bias in diversification rates: An example using New World birds. Ecology 86:2278-2287
11. Schmeske DW, Mittelbach GG, Cornell HV, Sobel JM, Roy K (2009) Is there a latitudinal gradient in the importance if biotic interactions? Annual Reviews in Ecology and Evolution 40:245-269
12. Lambers, J. H. R., J. S. Clark, and B. Beckage (2002) Density-dependent mortality and the latitudinal gradient in species diversity. Nature 417:732-735.
13. Hillebrand, H. 2004. On the generality of the latitudinal diversity gradient. The American Naturalist 163:192-211.
14. Pianka, E. R. (1966) Latitudinal gradients in species diversity: A review of concepts. American Naturalist 100: 33-46.
15. Van der Heijden, MGA, Horton, TR (2009) Socialism in soil? The importance of mycorrhizal fungal networks for facilitation in natural ecosystems. Journal of Ecology 97, 1139-1150.
16. Hoeksema JD, Bruna EM (2000) Pursuing the big questions about interspecific mutualism: a review of theoretical approaches. Oecologia 125:321-330.
17. Michener, CD (2007) The Bees of the World, 2nd ed. Baltimore, MD: The John Hopkins University Press.
18. Toju H, Abe H, Ueno S, Miyazawa Y, Taniguchi F, Sota T, Yahara T (2011) Climatic gradients of arms race coevolution. American Naturalist 177(5): 562-573.
19. Futuyma D (2009) Evolution. Sinauer Associates Inc. Second Edition.
20. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25, 1965-1978.
21. Jarvis A, Reuter HI, Nelson A, Guevara E (2008) Hole-filled SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database (
22. Sokal, R. R., & Rohlf, F. J. (1995). Biometry, 3rd edn. New York: Freeman
23. Burnaham and Anderson (1998) Model Selection and Multimodel Inference A Practical Information-Theoretic Approach. Second Edition
24. Olejnik, S. & Algina, J (2003) Generalized Eta and Omega Squared Statistics: Measures of Effect Size for Some Common Research Designs Psychological Methods. 8:(4)434-447.
25. Hegland, S. J., A. Nielsen, A. Lazaro, A. L. Bjerknes, and O. Totland (2009) How does climate warming affect plant-pollinator interactions? Ecology Letters 12:184–195.
26. Durant, J. M., D. Hjermann, G. Ottersen, and N. Stenseth (2007) Climate and the match or mismatch between predator require- ments and resource availability. Climate Research 33:271–283.
27. de Queiroz K (2011) Branches in the lines of descent: Charles Darwin and the evolution of the species concept. Biol. J. Linn. Soc. 103: 19-35.
28. Huang, D. (2012) Threatened reef corals of the world. Plos One, 7(3): e34459.
29. Segelbacher G, Cushman SA, Epperson, BK, Fortin M-J, Francois O, Hardy OJ, Holderegger R, Taberlet P, Waits LP, Manel S (2010) Applications of landscape genetics in conservation biology: concepts and challenges. Conserv. Genet. 11:375-385.
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License