Abstract
Climate change models predict increases in atmospheric carbon dioxide concentration. As ecosystems equilibrate with the atmosphere, stimulation of photosynthesis is expected to occur. However, growth limitation due to soil nutrients may potentially limit sequestration of carbon. Additionally, changes in producer nutritional quality may cause a decline in grazer populations. Here we extend the WKL model to allow for consideration of the impacts of elevated atmospheric carbon dioxide concentration on producer–grazer dynamics. We do so by explicitly tracking the free carbon in the medium and allowing the producer’s growth rate to be limited by available carbon instead of light. This model is analyzed using primarily local bifurcation analysis. Overall, these analyses show that carbon sequestration due to increased atmospheric carbon dioxide can be limited by insufficient available phosphorus. Furthermore, increased atmospheric carbon dioxide will cause decreased stoichiometric quality of producers where available phosphorus is limiting.
Similar content being viewed by others
References
Ainsworth EA, Rogers A (2007) The response of photosynthesis and stomatal conductance to rising [\(\text{ CO}_2\)]: mechanisms and environmental interactions. Plant Cell Environ 30(3):258–270. https://doi.org/10.1111/j.1365-3040.2007.01641.x
Burkhardt S, Riebesell U (1997) \(\text{ CO}_{{2}}\) availability affects elemental composition (C:N:P) of the marine diatom Skeletonema costatum. Mar Ecol Prog Ser 155:67–76. https://doi.org/10.3354/meps155067
Davies CM, Wang H (2020) Contrasting stoichiometric dynamics in terrestrial and aquatic grazer–producer systems. J Biol Dyn. https://doi.org/10.1080/17513758.2020.1771442
de Graaff MA, van Groenigen KJ, Six J, Hungate B, van Kessel C (2006) Interactions between plant growth and soil nutrient cycling under elevated \(\text{ CO}_2\): a meta-analysis. Glob Change Biol 12(11):2077–2091. https://doi.org/10.1111/j.1365-2486.2006.01240.x
De Kauwe MG, Medlyn BE, Zaehle S et al (2013) Forest water use and water use efficiency at elevated \(\text{ CO}_{{2}}\): a model-data intercomparison at two contrasting temperate forest FACE sites. Glob Change Biol 19(6):1759–1779. https://doi.org/10.1111/gcb.12164
Dhooge A, Govaerts W, Kuznetsov YA, Meijer HGE, Sautois B (2008) New features of the software MatCont for bifurcation analysis of dynamical systems. Math Comput Model Dyn Syst 14(2):147–175. https://doi.org/10.1080/13873950701742754
Diehl S (2007) Paradoxes of enrichment: effects of increased light versus nutrient supply on pelagic producer–grazer systems. Am Nat 169(6):E173–E191. https://doi.org/10.1086/516655
Drake BG, Gonzàlez-Meler MA, Long SP (1997) More efficient plants: A consequence of rising atmospheric \(\text{ CO}_{{2}}\)? Annu Rev Plant Biol 48(1):609–639. https://doi.org/10.1146/annurev.arplant.48.1.609
Du C, Wang X, Zhang M, Jing J, Gao Y (2019) Effects of elevated \(\text{ CO}_{{2}}\) on plant CNP stoichiometry in terrestrial ecosystems: a meta-analysis. Sci Total Environ 650:697–708. https://doi.org/10.1016/j.scitotenv.2018.09.051
Elser JJ, Fagan WF, Kerkhoff AJ, Swenson NG, Enquist BJ (2010) Biological stoichiometry of plant production: metabolism, scaling and ecological response to global change. New Phytol 186(3):593–608. https://doi.org/10.1111/j.1469-8137.2010.03214.x
Hessen DO, Elser JJ, Sterner RW, Urabe J (2013) Ecological stoichiometry: an elementary approach using basic principles. Limnol Oceanogr 58(6):2219–2236. https://doi.org/10.4319/lo.2013.58.6.2219
Körner C (2009) Responses of humid tropical trees to rising \(\text{ CO}_{{2}}\). Annu Rev Ecol Evol Syst 40:61–79. https://doi.org/10.1146/annurev.ecolsys.110308.120217
Leakey ADB, Ainsworth EA, Bernacchi CJ, Rogers A, Long SP, Ort DR (2009) Elevated \(\text{ CO}_{{2}}\) effects on plant carbon, nitrogen, and water relations: six important lessons from FACE. J Exp Bot 60(10):2859–2876. https://doi.org/10.1093/jxb/erp096
Loladze I, Kuang Y, Elser JJ (2000) Stoichiometry in producer–grazer systems: linking energy flow with element cycling. Bull Math Biol 62(6):1137–1162. https://doi.org/10.1006/bulm.2000.0201
Long SP, Drake BG (1991) Effect of the long-term elevation of \(\text{ CO}_{{2}}\) concentration in the field on the quantum yield of photosynthesis of the \(\text{ C}_{{3}}\) sedge, Scirpus olneyi. Plant Physiol 96(1):221–226. https://doi.org/10.1104/pp.96.1.221
Lovelock CE, Kyllo D, Winter K (1996) Growth responses to vesicular-arbuscular mycorrhizae and elevated \(\text{ CO}_{{2}}\) in seedlings of a tropical tree, Beilschmiedia pendula. Funct Ecol. https://doi.org/10.2307/2390177
Low-Decarie E, Fussmann GF, Bell G (2014) Aquatic primary production in a high-\(\text{ CO}_{{2}}\) world. Trends Ecol Evol 29(4):223–232. https://doi.org/10.1016/j.tree.2014.02.006
Pachauri RK, Allen MR, Barros VR et al (2014) Climate change 2014: synthesis report. In: Contribution of Working Groups I, II and III to the fifth assessment report of the Intergovernmental Panel on Climate Change. IPCC
Petrucci RH, Herring FG, Madura JD, Bissonnette C (2011) General chemistry: principles and modern applications, 10th edn. Pearson Canada Inc
Pugh TAM, Müller C, Arneth A, Haverd V, Smith B (2016) Key knowledge and data gaps in modelling the influence of \(\text{ CO}_2\) concentration on the terrestrial carbon sink. J Plant Physiol 203:3–15. https://doi.org/10.1016/j.jplph.2016.05.001
Spijkerman E, de Castro F, Gaedke U (2011) Independent colimitation for carbon dioxide and inorganic phosphorus. PLoS One. https://doi.org/10.1371/journal.pone.0028219
Sterner RW, Elser JJ (2002) Ecological stoichiometry: the biology of elements from molecules to the biosphere. Princeton University Press
Stitt M, Lunn J, Usadel B (2010) Arabidopsis and primary photosynthetic metabolism-more than the icing on the cake. Plant J 61(6):1067–1091. https://doi.org/10.1111/j.1365-313X.2010.04142.x
Urabe J, Togari J, Elser JJ (2003) Stoichiometric impacts of increased carbon dioxide on a planktonic herbivore. Glob Change Biol 9(6):818–825. https://doi.org/10.1046/j.1365-2486.2003.00634.x
van de Waal DB, Verschoor AM, Verspagen JM et al (2010) Climate-driven changes in the ecological stoichiometry of aquatic ecosystems. Front Ecol Environ 8(3):145–152. https://doi.org/10.1890/080178
von Caemmerer S, Quick WP (2000) Rubisco: physiology in vivo. In: Leegood RC, Sharkey TD, von Caemmerer S (eds) Photosynthesis. Springer, Dordrecht, pp 85–113
Wang H, Kuang Y, Loladze I (2008) Dynamics of a mechanistically derived stoichiometric producer–grazer model. J Biol Dyn 2(3):286–296. https://doi.org/10.1080/17513750701769881
Wanninkhof R (2007) The impact of different gas exchange formulations and wind speed products on global air–sea \(\text{ CO}_{{2}}\) fluxes. In: Garbe CS, Handler RA, Jähne B (eds) Transport at the air–sea interface. Springer, Berlin, pp 1–23
Würth MKR, Winter K, Körner C (1998) In situ responses to elevated \(\text{ CO}_{{2}}\) in tropical forest understorey plants. Funct Ecol 12(6):886–895. https://doi.org/10.1046/j.1365-2435.1998.00278.x
Acknowledgements
This research is partially supported by an NSERC Discovery Grant RGPIN-2020-03911 and an NSERC Accelerator Grant RGPAS-2020-00090.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A
This appendix contains the proof of Theorem 1.
Proof
Let \(X(t) \equiv (x(t),y(t),p(t))\) be a solution of Eqs. (6)–(8) with initial conditions in \(\Omega \). Then, \(0< x(0)< \text {min}\{T_P/q, h(T_C)\}, 0< y(0), 0< p(0), p(0) + \theta y(0) < T_P,\) and \(x(0) + y(0) < T_C\). Assume for the sake of contradiction that there is time \(t_1 > 0\) such that X(t) touches or crosses the boundary of \({\bar{\Omega }}\) (closure of \(\Omega \)) for the first time. Therefore, \((x(t),y(t),p(t)) \in \Omega \) for \(0 \le t < t_1\). We now consider cases for which part of the boundary of \({\bar{\Omega }}\) \(X(t_1)\) lies on.
Case 1 \(x(t_1) = 0\) but \(p(t_1) \ne 0\). Since \((x(t),y(t),p(t)) \in \Omega \) for \(0 \le t < t_1\), then \(0 < p(t)\) and \(p(t) + \theta y(t) \le T_P\) for \(0 \le t \le t_1\). Also, all parameters are assumed to be positive. Therefore, \(y(t) \le T_P/\theta - p(t)/\theta < T_P/\theta \) for \(0 \le t \le t_1\). Let \(p_1 = \text {min}\{p(t): t \in [0,t_1]\} > 0\), \(x_1 = \text {max}\{x(t): t \in [0,t_1] \} > 0\), and \(y_1 = \text {max}\{y(t): t \in [0,t_1] \} > 0\). Both \(x_1\) and \(y_1\) are guaranteed to exist since the initial conditions are in \(\Omega \) and thus are positive. Then for \(0 \le t \le t_1\), we have (using Eq. 6)
where \(\mu \) is a constant. Then, \(x(t) \ge x(0)e^{\mu t}\) for \(0 \le t \le t_1\), which implies \(x(t_1) \ge x(0)e^{\mu t_1} > 0\), which is a contradiction.
Case 2 \(x(t_1) = \text {min}\{T_P/q, h(T_C)\}\). Since \((x(t),y(t),p(t)) \in \Omega \) for \(0 \le t < t_1\), we know \(0 \le y(t)\) and \(p(t) + \theta y(t) \le T_P\) for \(0 \le t \le t_1\). Therefore, \(p(t) \le T_P - \theta y(t) \le T_P\) for \(0 \le t \le t_1\). Clearly \(T_C - x(t) - y(t) \le T_C\) for \(x(t), y(t) \ge 0\). Since h(C) is non-decreasing, then \(h(T_C - x(t) - y(t)) \le h(T_C)\) for \(x(t), y(t) \ge 0\). Hence, for \(0 \le t \le t_1\), we have
Note that the right hand side is logistic growth in x with the carrying capacity given by \(\text {min}\{T_P/q, h(T_C)\}\). The standard comparison argument yields that \(x(t) < \text {min}\{T_P/q, h(T_C)\}\) for all \(0 \le t \le t_1\), a contradiction.
Case 3 \(y(t_1) = 0\). For \(0 \le t \le t_1\), we have from Eq. (7)
since all parameters are assumed to be positive; \(f(0) = 0, f'(x) > 0, f''(x) \le 0\) for \(x \ge 0\); and \((x(t),y(t),p(t)) \in \Omega \) for \(0 \le t < t_1\). Hence, \(y(t) \ge y(0)e^{-({\hat{d}}+l_y)t} > 0\) for \(0 \le t \le t_1\), a contradiction.
Case 4 \(p(t_1) = 0\). Since \((x(t),y(t),p(t)) \in \Omega \) for \(0 \le t < t_1\), we know \(p(t) + \theta y(t) \le T_P\) for \(0 \le t \le t_1\). Therefore, \(T_P-p(t)-\theta y(t) \ge 0\) for \(0 \le t \le t_1\), and since we assume that in general \(g(0) = 0\) and \(g'(P) > 0\) for \(P \ge 0\), then \(g(T_P - p(t) - \theta y(t)) \ge 0\) for \(0 \le t \le t_1\).
Also, since \(0 \le p(t)\) and \(p(t) + \theta y(t) \le T_P\) for \(0 \le t \le t_1\), then \(y(t) \le T_P/\theta - p(t)/\theta \le T_P/\theta \) for \(0 \le t \le t_1\), and therefore \(-T_P/\theta \le -y(t)\).
Thus, for \(0 \le t \le t_1\), we have from Eq. (8)
where \(\nu \) is a constant. Thus, \(p(t) \ge p(0) e^{\nu t} > 0\) for \(0 \le t \le t_1\), a contradiction.
Case 5 \(p(t_1) + \theta y(t_1) = T_P\). Let \(z(t) = T_P - p(t) - \theta y(t)\). Since \(p(t_1) + \theta y(t_1) = T_P\), then \(z(t_1) = 0\). Also, since \(t_1\) is assumed to be the first time that X(t) touches or crosses the boundary of \({\bar{\Omega }}\), then \(p(t) + \theta y(t) < T_P\) for \(0 \le t < t_1\) and thus \(z(t) > 0\) for \(0 \le t < t_1\). Then for \(0 \le t \le t_1\), we have
where \({\tilde{u}} > 0\) and \({\tilde{v}} > 0\) are constant. Thus \(z(t) \ge e^{-{\tilde{v}}t} z(0) > 0\) for \(0 \le t \le t_1\), a contradiction.
Case 6 \(x(t_1) + y(t_1) = T_C\). Consider the limit of \(\text {d}x/\text {d}t\) as \(t \rightarrow t_1\):
Since \(h(0) = 0\), then the limit as \(t \rightarrow t_1\) of \(\text {min}\{p/q, h(T_C-x-y) \}\) is 0. However, \(x(t_1) > 0\). Therefore, as \(t \rightarrow t_1\), the limit of \(\text {d}x/\text {d}t\) is \(-\infty \). Clearly this implies that \(\text {d}y/\text {d}t \rightarrow -\infty \) as \(t \rightarrow t_1\). But then as \(t \rightarrow t_1\), \(x(t) + y(t) \rightarrow -\infty \), a contradiction. \(\square \)
Appendix B
This appendix contains all the necessary information to find the equilibria assuming Holling type I functional responses for f and g. In addition, we assume \(\alpha = 0\) and \(\rho (C) = 1\).
Substituting \(f(x) = cx\), \(g(P) = {\hat{c}}P\) and \(h(C) = \gamma C\) into Eqs. (6)–(8), we get the following system:
Equilibria satisfy
Clearly the trivial extinction equilibrium \(E_0 = (0,0,0)\) is a possible solution of the above. We can also explicitly find the forms of the grazer extinction equilibria. The grazer extinction equilibria always take the form \( ({\bar{x}},0,{\bar{p}})\), where \({\bar{x}}\) and \({\bar{p}}\) are decided by what is limiting the producer and other parameter-based conditions.
When the producer is phosphorus limited at equilibrium (\({\bar{p}}/q \le \gamma (T_C-{\bar{x}})\)), a grazer extinction equilibrium is given by \(({\bar{x}}, 0,{\bar{p}})\) where
We assume all parameters are positive. For \({\bar{x}} \ge 0\), we need \({\hat{c}}T_P(r-l_x) - dqr \ge 0\). Then, for \({\bar{p}} \ge 0\), we need \(r - l_x >0\).
When the producer is carbon limited at equilibrium (\(\gamma (T_C-{\bar{x}}) \le {\bar{p}}/q\)), we have \(({\bar{x}}, 0,{\bar{p}})\) where
For this equilibrium to be biologically feasible and not equal to \(E_0\), we require either \(r - l_x > 0\) and \(r + \gamma r - \gamma l_x > 0\), or \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x < 0\).
To determine parameter-based conditions for which grazer extinction equilibria are present, we substitute the equilibria into \({\bar{p}}/q = \gamma (T_C - {\bar{x}})\). With some rearrangement, we find that for both the phosphorus and carbon limited equilibria, \({\bar{p}}/q = \gamma (T_C - {\bar{x}})\) when
A grazer extinction equilibrium takes the phosphorus limited form when \({\bar{p}}/q \le \gamma (T_C - {\bar{x}})\), or equivalently,
A grazer extinction equilibrium takes the carbon limited form when \({\bar{p}}/q \ge \gamma (T_C - {\bar{x}})\), that is,
There may also be coexistence equilibria, which would satisfy
Appendix C
This appendix contains all the necessary information to find the equilibria assuming Holling type II functional responses for f and g. In addition, we assume \(\alpha = 0\) and \(\rho (C) = 1\).
Using \(f(x) = cx/(a+x)\), \(g(P) = {\hat{c}}P/({\hat{a}} + P)\), and \(h(C) = \gamma C\),
Equilibria satisfy
Clearly the trivial extinction equilibrium \(E_0 = (0,0,0)\) is a possible solution of this system. We can explicitly find the forms of the grazer extinction equilibria, dependent upon what is limiting the producer and other parameter-based conditions.
When the producer is phosphorus limited at equilibrium (\({\bar{p}}/q \le \gamma (T_C-{\bar{x}}\))), the grazer extinction equilibrium is given by \(({\bar{x}},0,{\bar{p}})\) where
Consider the equation for \({\bar{x}}\) that includes \({\bar{p}}\). Since we require \({\bar{x}}, {\bar{y}}, {\bar{p}} \ge 0\) for the equilibrium to be biologically feasible, then we require \(r - l_x \ge 0\), since otherwise \({\bar{x}}\) is negative for positive \({\bar{p}}\). The equation for \({\bar{p}}\) would also yield additional conditions for non-negativity, which are not included here.
When the producer is carbon limited at equilibrium (\(\gamma (T_C-{\bar{x}}) \le {\bar{p}}/q\)), we have multiple possible grazer extinction equilibria, given by \(({\bar{x}}, 0, {\bar{p}})\) where
For \({\bar{x}} \ge 0\), we need either \(r - l_x \ge 0\) and \(r + \gamma r - \gamma l_x > 0\), or \(r - l_x \le 0\) and \(r + \gamma r - \gamma l_x < 0\). Note that \(r - l_x > 0\) and \(r + \gamma r - \gamma l_x < 0\) is not possible for positive parameters. Thus, the equilibrium is not biologically feasible if \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\). As in the phosphorus limited case, the equation for \({\bar{p}}\) would also yield conditions for feasibility of the equilibria.
To determine conditions for which grazer equilibria are present dependent on parameter values, we substitute the grazer extinction equilibria found above into \({\bar{p}}/q = \gamma (T_C-{\bar{x}})\). Using the values of \({\bar{x}}\) and \({\bar{p}}\) for the phosphorus limited equilibrium, we require
for \({\bar{p}}/q = \gamma (T_C - {\bar{x}})\), and for phosphorus limited producer growth at equilibrium (\({\bar{p}}/q \le \gamma (T_C-{\bar{x}}))\), we require
Using the values of \({\bar{x}}\) and \({\bar{p}}\) for the carbon limited equilibrium, we require
for \({\bar{p}}/q = \gamma (T_C - {\bar{x}})\), and for carbon limited producer growth at equilibrium (\(\gamma (T_C - {\bar{x}}) \le {\bar{p}}/q\)), we require
There may also be coexistence equilibria, which would satisfy
Appendix D
This appendix contains the proof of Theorem 4.
Proof
Let \(u = x/p\), then
Since \(g(0) = 0\) and \(g^\prime (P) > 0\) for \(P \ge 0\), then \(- u^2g(T_P-p-\theta y) \le 0\). Note that \(\text {min} \{p/q, h(T_C-x-y) \} \le p/q\) if and only if
Thus
Hence, \(u \le \text {min}\{x(0)/p(0), [1 + (d-l_x)/r]/q \} \equiv m\). Consider the equation for \((\text {d}p/\text {d}t)\):
Since \(d > mg(T_P)\) implies \(g(T_P)m - d < 0\), then \(\text {d}p/\text {d}t < 0\) and thus \(p \rightarrow 0\) as \(t \rightarrow \infty \).
Now, consider the equation for \((\text {d}x/\text {d}t)\):
Hence \(\text {lim sup}_{t \rightarrow \infty } x(t) \le p/q\) and \(x \rightarrow 0\) as \(t \rightarrow \infty \). Given the dependence of y on x, this implies that \(y \rightarrow 0\) as \(t \rightarrow \infty \). Therefore, the extinction steady state \(E_0 = (0,0,0)\) is globally asymptotically stable if \(d > mg(T_P)\), where \(m = \text {min}\{x(0)/p(0),[1+(d-l_x)/r]/q\}\). \(\square \)
Appendix E
This appendix contains all the necessary calculations for the stability results in Sect. 3.3 using Holling type I functional responses for f and g.
For \(f(x) = cx\), \(g(P) = {\hat{c}}P\), and \(h(C) = \gamma C\), using F, G, H as defined in “Appendix B”,
We compute the products/sums we need for the Jacobian:
Case 1 Producer nutrient limited and grazer quality limited
The corresponding Jacobian matrix with \(E_P\) from Theorem 2 substituted in is
where
The three eigenvalues for \(A_1\) are (see “Appendix G” for proof)
Case 2 Producer nutrient limited and grazer quantity limited
The corresponding Jacobian matrix with \(E_P\) from Theorem 2 substituted in is
where all entries except \(a_{22}\) are the same as in \(A_1\) (Case 1 above), and
As in Case 1, we can explicitly find the eigenvalues for \(A_2\):
Case 3 Producer carbon limited and grazer quality limited
The Jacobian with \(E_C\) from Theorem 2 substituted in takes the form
where
A 3 x 3 matrix with zeros in these specific entries has its eigenvalues along the main diagonal, since it can be decomposed into a 2 x 2 upper triangular matrix and the eigenvalue given by \(a_{22}\). Thus, in this case
Consider \(r + \gamma r - \gamma l_x\) and \(r - l_x\). We see that for \(r, \gamma , l_x > 0\),
Also, since all parameters are assumed to be positive,
For stability of the grazer extinction equilibrium, we need \(\lambda _2 < 0\), which requires
and either \(r + \gamma r - \gamma l_x < 0\) (in which case \(r-l_x < 0\)) or \(r-l_x > 0\) (in which case \(r + \gamma r - \gamma l_x > 0\)). Then, the grazer extinction equilibrium is a stable node since \(\lambda _1, \lambda _3 < 0\) when \(r + \gamma r - \gamma l_x\) and \(r - l_x\) have the same sign.
For either \(r + \gamma r - \gamma l_x < 0\) or \(r-l_x > 0\), and
the grazer extinction equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
Now, for \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\) we have a few cases. \(\lambda _1\) is guaranteed to be positive, but the signs of \(\lambda _2\) and \(\lambda _3\) depend on additional conditions.
For \(\lambda _2 < 0\), we need either \({\hat{c}}\gamma \theta T_C(r-l_x) + d\theta (r+\gamma r - \gamma l_x) > 0\) or \({\hat{c}}\gamma \theta T_C(r-l_x) + d\theta (r+\gamma r - \gamma l_x) < 0\) and
For \(\lambda _3 < 0\), we need
Hence, for \(r-l_x < 0\) and \(r+\gamma r - \gamma l_x > 0\), we can have either a saddle with a one- or two-dimensional unstable manifold, or an unstable node. However, when \(r-l_x < 0\) and \(r+\gamma r - \gamma l_x > 0\), the equilibrium is not biologically feasible.
Case 4 Producer carbon limited and grazer quantity limited
The Jacobian with \(E_C\) from Theorem 2 substituted in is as follows:
where all entries are the same as \(A_3\) (Case 3), except \(a_{22}\) and
As in the previous case, this matrix has its eigenvalues along the main diagonal. Thus, we have
If \(r + \gamma r - \gamma l_x < 0\), then we know \(r-l_x < 0\). Therefore, if \(r + \gamma r - \gamma l_x < 0\), then \(\lambda _1, \lambda _3 < 0\) and we also need for stability
If \(r-l_x > 0\), then we know \(r + \gamma r - \gamma l_x > 0\) and therefore \(\lambda _1, \lambda _3 < 0\). For \(\lambda _2\) to be less than 0 and therefore for stability we also require
Hence, the grazer extinction equilibrium is a stable node for
and either \(r + \gamma r - \gamma l_x < 0\) or \(r-l_x > 0\).
For either \(r + \gamma r - \gamma l_x < 0\) or \(r-l_x > 0\), and
the grazer extinction equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
For \(r-l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\), we have \(\lambda _1 > 0\) and \(\lambda _2 < 0\). Therefore, we have a saddle. If \(\lambda _3 < 0\), the saddle has a two-dimensional stable manifold and a one-dimensional unstable manifold; if \(\lambda _3 > 0\), it has a one-dimensional stable manifold and a two-dimensional unstable manifold. Note that for \(r-l_x < 0\) and \(r+\gamma r - \gamma l_x > 0\), the equilibrium is not biologically feasible.
Appendix F
This appendix contains all the necessary calculations for the stability results in Sect. 3.3 using Holling type II functional responses for f and g.
For \(f(x) = cx/(a+x)\), \(g(P) = {\hat{c}}P/({\hat{a}} + P)\), and \(h(C) = \gamma C\), using F, G, H as defined in “Appendix C”,
The necessary products and sums for the Jacobian are
Case 1 Producer nutrient limited and grazer quality limited
The corresponding Jacobian matrix with \(E_P\) from Theorem 3 substituted in is
where for \(E_P = ({\bar{x}}, 0, {\bar{p}})\) (Theorem 3)
As in Holling type I Case 1, we can explicitly determine the eigenvalues:
Again, the eigenvalues are not particularly illuminating. Also similar to the Holling type I case, stability does not depend on \(T_C\) or \(\gamma \). Given the producer is nutrient limited at this equilibrium, it is reasonable that the carbon-dependent carrying capacity has no impact on its stability.
Case 2 Producer nutrient limited and grazer quantity limited
The corresponding Jacobian with \(E_P\) from Theorem 3 substituted in is
where all entries are the same as \(B_1\) (Case 1) except \(b_{22}\), which is
for \(E_P = ({\bar{x}}, 0, {\bar{p}})\) (Theorem 3).
The eigenvalues are
The eigenvalues do not provide any stability conclusions in this case. As in Case 2 for Holling type I functional responses, stability does not depend on \(T_C\), \(\gamma \), or \(\theta \). The producer is nutrient limited at equilibrium, so \(T_C\) and \(\gamma \) should not impact stability of the equilibrium. Note that since the grazer is quantity limited at equilibrium, we would also expect the stability of the equilibrium to not depend on the grazer’s P:C ratio, \(\theta \).
Case 3 Producer carbon limited and grazer quality limited
Here we use \(E_{C+}\) or \(E_{C-}\) from Theorem 3 and compute
where
where \({\bar{x}}\) and \({\bar{p}}\) correspond to the possible carbon-limited equilibrium values from Theorem 3.
The corresponding eigenvalues are along the main diagonal, as in the corresponding Holling type I case:
For these eigenvalues, we consider stability cases using \(r- l_x\) and \(r + \gamma r - \gamma l_x\). As in the Holling type I case, we recognize that \(r + \gamma r - \gamma l_x < 0\) implies \(r - l_x < 0\), and \(r - l_x > 0\) implies \(r + \gamma r - \gamma l_x > 0\), when all parameters are assumed to be positive. Therefore we consider three cases.
If \(r + \gamma r - \gamma l_x < 0\) (and therefore \(r - l_x < 0\)), then \(\lambda _1 < 0\) and \(\lambda _3 < 0\). The sign of \(\lambda _2\) depends on the sign of \({\bar{p}}\). Looking at the forms of the equilibria in Sect. 3.2, we observe that for \(r + \gamma r - \gamma l_x < 0\) and \(r - l_x < 0\), both equilibria have a positive \({\bar{p}}\). Hence, the equilibrium is a stable node for \(r + \gamma r - \gamma l_x < 0\) and
If \(r + \gamma r - \gamma l_x < 0\) but the additional condition is not satisfied, then the equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
If \(r - l_x > 0\) (and therefore \(r + \gamma r - \gamma l_x > 0\)), then \(\lambda _1 < 0\) and \(\lambda _3 < 0\). Since \({\bar{p}} > 0\) for \(r - l_x > 0\) and both forms of \({\bar{p}}\), then the equilibrium is a stable node for \(r - l_x > 0\) and
If \(r - l_x > 0\) but \(\lambda _2 > 0\) because the above condition is not satisfied, then the equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
Lastly, if \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\), then \(\lambda _1 > 0\). The signs of \(\lambda _2\) and \(\lambda _3\) depend on additional conditions. For the equilibrium to be a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold when \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\), we require
Note that here the sign of \({\bar{p}}\) is not determined by the signs of \(r - l_x\) and \(r + \gamma r - \gamma l_x\) alone. If either one of these additional conditions is not satisfied (i.e., \(\lambda _2 > 0\) or \(\lambda _3 > 0\)) and the other is satisfied, then the equilibrium is a saddle with a one-dimensional stable manifold and a two-dimensional unstable manifold. If both of these additional conditions are not satisfied, then the equilibrium is an unstable node. However, in this case, the equilibrium is only biologically feasible when the signs of \(r - l_x\) and \(r + \gamma r - \gamma l_x\) are the same.
Case 4 Producer carbon limited and grazer quantity limited
The Jacobian with \(E_{C+}\) or \(E_{C-}\) from Theorem 3 substituted in takes the form
where all entries are the same as in \(B_3\), except
Once again, the eigenvalues are along the main diagonal, using \({\bar{p}}\) from \(E_{C+}\) or \(E_{C-}\) from Theorem 3:
We consider cases based on \(r - l_x\) and \(r + \gamma r - \gamma l_x\). Note that we cannot have \(r - l_x > 0\) and \(r + \gamma r - \gamma l_x < 0\) for positive \(r, \gamma \) and \(l_x\).
If \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x < 0\), then \(\lambda _1 < 0\) and \(\lambda _3 < 0\). Therefore, the equilibrium is a stable node if \(r + \gamma r - \gamma l_x < 0\) and
If this additional condition is not satisfied (i.e., \(\lambda _2 > 0\)), then the equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
Similarly, if \(r - l_x > 0\) and \(r + \gamma r - \gamma l_x > 0\), then \(\lambda _1 < 0\) and \(\lambda _3 < 0\). Hence, the equilibrium is a stable node if \(r - l_x > 0\) and
If the above condition is not satisfied, then the equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold.
If \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\), then \(\lambda _1 > 0\). The equilibrium is a saddle with a two-dimensional stable manifold and a one-dimensional unstable manifold if
If either of these conditions is not met, the equilibrium is a saddle with a one-dimensional stable manifold and a two-dimensional unstable manifold; if both conditions are not met, the equilibrium is an unstable node. Note that if \(r - l_x < 0\) and \(r + \gamma r - \gamma l_x > 0\), then the equilibrium is not biologically feasible.
Appendix G
This appendix proves the form of the eigenvalues for the nutrient limited grazer extinction equilibria.
Claim: For a matrix of the form
where all entries are real numbers, the eigenvalues are
Proof
We know an eigenvalue \(\lambda \) of matrix A satisfies \(det(A - \lambda I) = 0\). Hence, we solve this equation for \(\lambda \).
We see that
Using a cofactor expansion along the second row, we see
We can find solutions to the following using the quadratic formula:
The resulting solutions are
Clearly
and this proves the claim. \(\square \)
Rights and permissions
About this article
Cite this article
Davies, C.M., Wang, H. Incorporating carbon dioxide into a stoichiometric producer–grazer model. J. Math. Biol. 83, 49 (2021). https://doi.org/10.1007/s00285-021-01658-3
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00285-021-01658-3
Keywords
- Carbon dioxide
- Stoichiometric producer–grazer model
- Bifurcation analysis
- Sequestration of carbon
- Phosphorus
- Photosynthesis
- Photorespiration