Measuring and Linking the Missing Part of Biodiversity and Ecosystem Function: The Diversity of Biotic Interactions
Round 1
Reviewer 1 Report
The idea of partitioning interactions into different measures of diversity, analogous to the partitioning of species diversity, is a very good one, and I applaud the goal of taking this on.
I am unsure, however, about the notation used to create the interaction matrices (e.g., Fig. 1b, Fig. 2a, 2b). The authors are illustrating only bipartite networks (e.g., plants, pollinators), but their interaction matrices are more like those developed for multitrophic level food webs. Thus, the interaction matrices do not correspond to either their webs or the related degree distributions (e.g., Fig. 1d, 1e).
As a result, I don't understand the subsequent computations, all of which depend on the value NZ derived from the interaction matrices. Until the discordance between the interaction matrices and the illustrated bipartite webs is resolved, I cannot evaluate the validity of the method or the interpretations and conclusions of the authors.
Author Response
Reviewer #1 (Comment to Author):
Comment #1 from the Reviewer #1: The idea of partitioning interactions into different measures of diversity, analogous to the partitioning of species diversity, is a very good one, and I applaud the goal of taking this on.
Response by the authors: We thank the reviewer for the feedback and criticism, we followed all the suggestions you made. We apologize for the misleading information that our figures transmitted. We hope that the figures are fixed and that you feel comfortable with the manuscript.
Comment #2 from the Reviewer #1: I am unsure, however, about the notation used to create the interaction matrices (e.g., Fig. 1b, Fig. 2a, 2b). The authors are illustrating only bipartite networks (e.g., plants, pollinators), but their interaction matrices are more like those developed for multitrophic level food webs. Thus, the interaction matrices do not correspond to either their webs or the related degree distributions (e.g., Fig. 1d, 1e). As a result, I don't understand the subsequent computations, all of which depend on the value NZ derived from the interaction matrices. Until the discordance between the interaction matrices and the illustrated bipartite webs is resolved, I cannot evaluate the validity of the method or the interpretations and conclusions of the authors.
Response by the authors: We thank the reviewer for his time spent and effort in revising our manuscript. Apparently, reviewer # 1 was confused (perhaps due to our lack of clarity) in the approach used by us. Here we use interaction matrices A to show the relationships between two trophic levels (e.g. plants and pollinators) where aij = number of interactions between the pollinator species i and the plant species j (Dáttilo and Rico-Gray 2018). Note that this same matrix organization can also be used to represent multitrophic level food webs (Pimm 1982). It is important to mention that there are other ways to represent biotic interactions, such as interaction lists. However, here we show the most used way to represent the interactions between two trophic levels in the programs and literature available. Thus, in order to avoid any confusion, we have clarified in our figure captions, that each dimension of the matrices (i.e. rows or columns) represents only one trophic level differently from those used for food web analysis.
We apologize for the misleading information in our hypothetical examples/figures. Following your criticism, we adjusted all the figures and recalculated network descriptors in order to make our examples clear to the reader.
Pimm, S. L. (1982). Food webs. In Food webs (pp. 1-11). Springer, Dordrecht.
Dáttilo, W., & Rico-Gray, V. (2018). Ecological networks in the tropics. Cham, Switzerland: Springer.
Reviewer 2 Report
Review of Luna et al. Measuring and Linking the Missing Part of Biodiversity – Diversity January 2020
In general I agree with the approach of the authors: measures of biodiversity must go beyond just counting species or habitats and include measures of interactions. So I certainly welcome their contribution to this discussion. However what they are proposing is not as unique as they suggest in the manuscript. There is at least one key paper that has tried to define a metric for the biodiversity of species interactions – Dyer et al. 2010: https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1744-7429.2009.00624.x
The authors should look at that paper and those that cite it to get a better understanding of where their work fits into our current understanding of how to measure the biodiversity of species interactions. Going even further back, the idea that the “biodiversity of species interactions” needed to be taken more seriously and measured appropriately is probably at least 20 years old. Certainly, I was using that phrase in the early 2000s – see: Ollerton, J. & Cranmer, L. (2002) Latitudinal trends in plant-pollinator interactions: are tropical plants more specialised? Oikos 98: 340-350
Other than the Dyer et al. paper the field has not made a huge amount of progress since then, which is why I welcome this new paper, but it needs to be placed in its proper historical context.
There are some other smaller issues with the manuscript as its stands which I’ve highlighted below in relation to line numbers. Some are just typos but others will need more consideration. I hope that the authors find this useful.
L36 – in general, phylogeneticists don’t like the term “basal”, use “early emerging” instead.
LL38-40 – this idea is a little dated now and needs to be more nuanced, e.g. we know that insect pollination was common in pre-angiosperm floras – for a short review, see reference 7 that you cite.
LL45-46 – “symbiotic” means close living and can be applied to parasitism and commensalism, as well as mutualism. Most plant-pollinator interactions are therefore not “symbiotic”; the exceptions are things like fig-fig wasp relationships. See also L12, L45, L77, L413, L542.
L202 – should that be “continuum”?
L205 - broken sentence?
L236 – typo – “the more the interactions”
LL283-286 – the references that are cited in relation to loss of species are both in silico assessments that may not reflect what is happening in nature. More recently experimental approaches that remove species in the field have challenged some of the older in silico conclusions. For example, see this paper and some of those it cites: Biella P., Akter A., Ollerton J., Tarrant S., Janeček Š., Jersáková J. & Klecka J. (2019) Experimental loss of generalist plants reveals alterations in plant-pollinator interactions and a constrained flexibility of foraging. Scientific Reports 9: 1-13
L337 – should this be “interactions”?
L367 – typo – “in terms of”
L373 – typo – “trophic”
L384 – should this be “loss or phenological”?
LL384-386 – the idea that most flower visitors are not pollinators and that pollination is “a rare event” is wrong. It very much depends on the species concerned. It might be true for some plants in some places in some years, but it’s certainly not universally true.
L396 – this should read “pollen vector”. A “pollinator vector” would be something that moved a pollinator.
L397 – should be “filtered”.
Author Response
Reviewer #2 (Comment to Author):
Comment #1 from the Reviewer #2: In general I agree with the approach of the authors: measures of biodiversity must go beyond just counting species or habitats and include measures of interactions. So I certainly welcome their contribution to this discussion. However what they are proposing is not as unique as they suggest in the manuscript. There is at least one key paper that has tried to define a metric for the biodiversity of species interactions – Dyer et al. 2010: https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1744-7429.2009.00624.x
Response by the authors: We appreciate the feedback and comments by the reviewer. Following his or her criticism we have now improved the manuscript and added the references recommended by the reviewer. We also lowered some of our statements and tried to place them under the correct context.
Comment #2 from the Reviewer #2: The authors should look at that paper and those that cite it to get a better understanding of where their work fits into our current understanding of how to measure the biodiversity of species interactions. Going even further back, the idea that the “biodiversity of species interactions” needed to be taken more seriously and measured appropriately is probably at least 20 years old. Certainly, I was using that phrase in the early 2000s – see: Ollerton, J. & Cranmer, L. (2002) Latitudinal trends in plant-pollinator interactions: are tropical plants more specialised? Oikos 98: 340-350
Response by the authors: Thank you for recommending previous work that supports our postulates We have added some phrases following this recommendation. We also double checked the literature used by us to avoid any misinterpretation of its content.
Now it reads:
Comment #3 from the Reviewer #2: Other than the Dyer et al. paper the field has not made a huge amount of progress since then, which is why I welcome this new paper, but it needs to be placed in its proper historical context. There are some other smaller issues with the manuscript as its stands which I’ve highlighted below in relation to line numbers. Some are just typos but others will need more consideration. I hope that the authors find this useful.
Response by the authors: We acknowledge that the field of interaction diversity has had slow development, this is why we fell motivated to write this new proposal particularly for the special issue of Diversity journal. Interactions are relevant processes and a misunderstood diversity component, indeed ecologist always relate interactions to explain ecosystem functioning. As Janzen said: What escapes the eye, however, is a much more insidious kind of extinction: the extinction of ecological interactions” (Janzen 1971). We have tried to place our manuscript in the proper historical context. In addition, we followed all of the reviewer recommendations, and we hope that the new version of the manuscript is clearer, if any further concerns rises we are open to answer and fix it. Thank you.
Janzen, D. H. (1971). The deflowering of central America.
Comment #4 from the Reviewer #2: L36 – in general, phylogeneticists don’t like the term “basal”, use “early emerging” instead.
Response by the authors: Changed, thanks for the suggestion.
Comment #5 from the Reviewer #2: LL38-40 – this idea is a little dated now and needs to be more nuanced, e.g. we know that insect pollination was common in pre-angiosperm floras – for a short review, see reference 7 that you cite.
Response by the authors: Thanks for pointing out this detail, now we have added information to introduction, now it reads:
Lines 43-45 “However, pollination can be traced back to pre-angiosperm floras as there are growing evidence that angiosperms evolved in habitats in which many gymnosperms where pollinated by insects [7].”
Comment #6 from the Reviewer #2: LL45-46 – “symbiotic” means close living and can be applied to parasitism and commensalism, as well as mutualism. Most plant-pollinator interactions are therefore not “symbiotic”; the exceptions are things like fig-fig wasp relationships. See also L12, L45, L77, L413, L542.
Response by the authors: We have now corrected this issue, thanks for clarifying the meaning of “symbiotic”.
Comment #7 from the Reviewer #2: L202 – should that be “continuum”?
Response by the authors: Changed, thanks for the suggestion.
Comment #8 from the Reviewer #2: L205 - broken sentence?
Response by the authors: Corrected, thanks for the suggestion.
Comment #9 from the Reviewer #2: L236 – typo – “the more the interactions”
Response by the authors: Corrected, thanks for the suggestion.
Comment #10 from the Reviewer #2: LL283-286 – the references that are cited in relation to loss of species are both in silico assessments that may not reflect what is happening in nature. More recently experimental approaches that remove species in the field have challenged some of the older in silico conclusions. For example, see this paper and some of those it cites: Biella P., Akter A., Ollerton J., Tarrant S., Janeček Š., Jersáková J. & Klecka J. (2019) Experimental loss of generalist plants reveals alterations in plant-pollinator interactions and a constrained flexibility of foraging. Scientific Reports 9: 1-13
Response by the authors: We acknowledge that our references actually are theoretical approaches, following the reviewer comment we have added more references including the experimental removal of species. In addition, we have added a phrase that highlights the lack of empirical evidence, and some examples of experimental studies that removed generalist species from networks.
Now it reads:
Lines 318 – 326 “It is worth to mention that despite the potential use of network robustness (R) to evaluate the tolerance of interactive communities to species extinction, the amount of empirical information available is still limited and controversial in the literature. On the one hand, most of theoretical studies have shown that when generalist species are experimentally removed from a community, its extinction curves rapidly decrease. On the other hand, in an empirical study, when the generalist harvester ant species was locally extinct, other submissive ant species with fewer interactions replaced its generalist role in the community [timoteo]. However, in a plant-pollinator system when highly visited plants were removed, pollination and insect foraging were jeopardized [ollerton].”
Comment #11 from the Reviewer #2: L337 – should this be “interactions”?
Response by the authors: Corrected, thanks for the suggestion.
Comment #12 from the Reviewer #2: L367 – typo – “in terms of”
Response by the authors: Done, thanks for the suggestion.
Comment #13 from the Reviewer #2: L373 – typo – “trophic”
Response by the authors: Done, thanks for the suggestion.
Comment #14 from the Reviewer #2: L384 – should this be “loss or phenological”?
Response by the authors: The reviewer is right, thanks for the suggestion.
Comment #15 from the Reviewer #2: LL384-386 – the idea that most flower visitors are not pollinators and that pollination is “a rare event” is wrong. It very much depends on the species concerned. It might be true for some plants in some places in some years, but it’s certainly not universally true.
Response by the authors: We appreciate the explanation of this issue as it is a common opinion in some ecological literature, we have lowered our statement. We wanted to highlight that in interaction systems there are always cheaters, therefore rewiring measures include the variation made by such species. Thereby, it is important to identify those species that are generating rewiring to better understand the spatial and temporal variation of interactions, and how cheaters and effective interaction partners contribute to such variation.
Now it reads:
Lines 416-421 “This is of evolutionary relevance, since floral visitors can be divided into cheaters (i.e. robbers and thieves of flower rewards) and effective pollinators and, therefore, the turnover of interactions (i.e., rewiring) could include the variation made by those non-effective pollinators. Hence, if we identify which species are the main drivers of interaction rewiring, we will understand how cheaters and effective interaction partners contribute to rewiring and their role in maintaining the stability and function of ecological networks.”
Comment #16 from the Reviewer #2: L396 – this should read “pollen vector”. A “pollinator vector” would be something that moved a pollinator.
Response by the authors: Thanks we have corrected this part.
Comment #17 from the Reviewer #2:L397 – should be “filtered”.
Response by the authors: Fixed.
Reviewer 3 Report
I carefully read the manuscript entitled “Measuring and linking the missing part of biodiversity and ecosystems function: the diversity of biotic interactions”. The authors provide tools for estimating the diversity of biotic interactions within ecosystems, with the aim of including ecological networks as measures of biodiversity.
The study is interesting and timely, and is an attempt to simplify and deal with biotic interactions, that are very complex and difficult to be analyzed, due to multiple factors that occur and act simultaneously.
I have only some minor comments to clarify some points:
Figure 1: does the matrix in figure 1(b) corresponds to the interactions shown in figure (a)? If this is the case, I don’t understand why the row sum of rows B, D, C, E, are 4, 3, 2, and 1. From figure (a) I see that species B, D, C, E interacts only with species a. Why the cell (B,b), (B,c), and (B,d) are filled in with the value “1”? As the authors state in line 70, no interaction should be shown with the value “0”.
Lines 178-180: this sentence is not clear. I don’t why abundant of species are expected to have a higher number of links than rare species. The numbers of links are related to species ecology (e.g. the ability of a pollinator to visits different kinds of flowers, due to its shape, tongue features, and so on, or the ability of a disperser to disperse different kinds of seeds), rather than species abundance. On the contrary, species abundance is related to interaction frequency (how often the same event can occur), e.g. how many times a pollinator visits a species.
Line 234: aij is not in formula (2). Should it be moved to the description of formula (3)?
Line 417: replace “habitats” with “habitat”
Line 436: It is not clear what do you mean for “focal local network”. Please provide a broader definition.
Line 469-471. The sentence is not clear. As highlighted for lines 178-180, is not clear why dominant and abundant species have a higher number of links than other species. A higher number of links may occur when traits are not specialized (e.g. seed traits which are adapted to a wide range of dispersers) but not necessary to species abundance.
Author Response
Reviewer #3 (Comment to Author):
Comment #1 from the Reviewer #3: I carefully read the manuscript entitled “Measuring and linking the missing part of biodiversity and ecosystems function: the diversity of biotic interactions”. The authors provide tools for estimating the diversity of biotic interactions within ecosystems, with the aim of including ecological networks as measures of biodiversity. The study is interesting and timely, and is an attempt to simplify and deal with biotic interactions, that are very complex and difficult to be analyzed, due to multiple factors that occur and act simultaneously. I have only some minor comments to clarify some points:
Response by the authors: We thank the reviewer for emphasizing the potential importance of this work and for his or her comments, which improved the quality of our manuscript. Below, we explain how we addressed the reviewer’s specific concerns and comments.
Comment #2 from the Reviewer #3: Figure 1: does the matrix in figure 1(b) corresponds to the interactions shown in figure (a)? If this is the case, I don’t understand why the row sum of rows B, D, C, E, are 4, 3, 2, and 1. From figure (a) I see that species B, D, C, E interacts only with species a. Why the cell (B,b), (B,c), and (B,d) are filled in with the value “1”? As the authors state in line 70, no interaction should be shown with the value “0”.
Response by the authors: We apologize for this misunderstanding, in the beginning this figure was meant to represent independent networks and examples, but as noted by you and another reviewer this was highly misleading. Following this criticism, we have synchronized the frame in the Figure 1. We hope that the figure now is clearer to the reader.
Comment #3 from the Reviewer #3: Lines 178-180: this sentence is not clear. I don’t why abundant of species are expected to have a higher number of links than rare species. The numbers of links are related to species ecology (e.g. the ability of a pollinator to visits different kinds of flowers, due to its shape, tongue features, and so on, or the ability of a disperser to disperse different kinds of seeds), rather than species abundance. On the contrary, species abundance is related to interaction frequency (how often the same event can occur), e.g. how many times a pollinator visits a species.
Response by the authors: The reviewer argument is right. However, both stochastic (abundance-based) and deterministic factors (niche-based) determine the number and frequencies of interaction among species. In some cases, abundance is the main factor but in others behavior stands out (Krishna et al. 2008). To avoid any confusion, we have re-written this phrase.
Now it reads:
Lines 195-197 “In this sense, each species within a network establishes a number of links with other species; for example, one highly competitive bee species may establish five links, whereas one specialized bee species only establishes one or two links.”
Krishna, A., Guimaraes Jr, P. R., Jordano, P., & Bascompte, J. (2008). A neutral‐niche theory of nestedness in mutualistic networks. Oikos, 117(11), 1609-1618.
Comment #4 from the Reviewer #3: Line 234: aij is not in formula (2). Should it be moved to the description of formula (3)?
Response by the authors: The term aij makes reference to how pij is calculated, despite it is not used for formula 2, it is required so the reader understands how to obtain pij. We modified this phrase to make it clearer to the reader.
Now it reads:
Lines 250-252 “In which pij is the proportion of the frequency of interactions in a network A. Network A is defined as aij = number of interactions between the higher trophic level i and the lower trophic level j.”
Comment #5 from the Reviewer #3: Line 417: replace “habitats” with “habitat”
Response by the authors: Done, thanks for the suggestion.
Comment #6 from the Reviewer #3: Line 436: It is not clear what do you mean for “focal local network”. Please provide a broader definition.
Response by the authors: Thanks for pointing out this situation, we have re-written this part of the manuscript to avoid confusion.
Now it reads:
Lines 465-471 “If we study a local network which has 10 realized interactions and the metaweb has 100, its dissimilarity is equal to 10/100, so the β’OS value of the studied network is 0.1, and this means that the interactions of studied network are highly dissimilar compared with the other networks that form the metaweb. In another case, if the studied network has 10 realized interactions and the metaweb has 12, the β’OS value will be of 10/12 = 0.83, which means that the interactions of the studied network are similar to the other networks in the metaweb.”
Round 2
Reviewer 1 Report
Thank you for clarifying the various network metrics and calculations. My only comment at this point is that you should be clear in the paper that your approach as detailed here is applicable to bipartite networks, but that the extension to multitrophic food webs remains an open question/open area of research.
Author Response
We thank the reviewer for her/his feedback, which was fundamental to improve the quality of our manuscript.
Now, we have rewritten some parts in the introduction, figure legend and conclusion in order to make clearer to the reader that our approach is applicable to bipartite networks. Now it reads:
Introduction:
"The main objective of this work is to explain our basic knowledge about the diversity of interactions involving two trophic levels (e.g., plant-pollinator and plant-disperser), and how i) is measured; ii) such measures should be biologically interpreted; and iii) this knowledge is useful for both ecological theory and conservation applications. We start by explaining the local diversity of interactions (α-diversity) and then how to measure the turnover of species interactions over space and time (β-diversity), and we will conclude with how to measure regional pools of species interactions (γ-diversity)."
Figure legend:
"Figure 1. (a) Local network formed by nodes denoting five plant species (left) and five floral visitor species (right); the links represent the interactions between the species. In this bipartite network, nodes can be subdivided in two trophic levels (i.e., plants and pollinators) and interactions only occur between nodes of different trophic levels. If we count the number of different lines we obtain the interaction richness SI. (b) Interaction matrix from which the network (a) was drawn, the numbers in the right represent the number of links per species k (i.e., degree). If we multiply the number of plant by the number of floral visitors we obtain the network size NZ. Each dimension of the matrix represents just one trophic level (e.g, rows = plants; columns = pollinators). (c) Hypothetical representation of the interaction of two communities (pasture and jungle), each network represents the interactions of each habitat, and the values below them describe its local properties SI, NZ and C. (d) Degree distribution plot of the network: (a) we can see that we have a species with a high degree (5 links) and some species with a low degree (1 or 2 links). (d) Cumulative degree distribution following a truncated power-law."
Conclusion:
"Our study is an effort to further understand an unknown component of biodiversity: species interactions. Conversely, the approach to measure the diversity of interactions that we present in this work applies for the study of bipartite networks, thereby the extension of this approach to measure the diversity in interactions in multitrophic food webs remains an open area of research, and this topic remains to be elucidated in detail (but see [46]). In short, we have highlighted the different and newest approaches to studying species interactions and the biodiversity-ecosystem function relationship by measuring the diversity of interactions over space, time or both together."
Once again, thank you for your time and effort in dealing with our manuscript.
