arxiv: 2604.27113 · v1 · submitted 2026-04-29 · 🧬 q-bio.PE · math.DS· nlin.AO

Recognition: unknown

Modeling the impact of host diversity on the evolution of vector feeding preferences and implications for disease control

Shravani Shetgaonkar , Anupama Sharma

Authors on Pith no claims yet

Pith reviewed 2026-05-07 10:16 UTC · model grok-4.3

classification 🧬 q-bio.PE math.DSnlin.AO

keywords vector-borne diseaseshost diversityfeeding preferencesadaptive behaviordisease controlbasic reproduction numbermathematical modelprevalence

0 comments

The pith

Shortening the infectious period of the preferred host limits vector-borne disease persistence even as vectors adapt their feeding preferences.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper constructs a mathematical model linking two host species to a vector population whose feeding preference for one host evolves under selective pressure from disease interventions. It calculates the basic reproduction number R0 to identify when the disease persists and isolates conditions under which control measures succeed or backfire. The central finding is that shortening the infectious period in the preferred host suppresses transmission while preference shifts triggered by host protection reduce prevalence in that host at the possible cost of higher overall prevalence. A reader would care because real-world interventions such as bed nets or treatment often target one host and could unintentionally change vector behavior across multiple reservoirs.

Core claim

In the two-host model, protective actions on the vector's preferred host induce adaptive shifts in feeding preference that lower disease prevalence within that host; however, overall prevalence across both hosts can rise. Shortening the infectious period of the preferred host is shown to be an effective strategy for reducing disease persistence, and a threshold value governs whether a shift toward the non-preferred host amplifies or dampens the burden on the primary host.

What carries the argument

The coupled differential-equation model of transmission that incorporates adaptive rules for vector feeding preference, from which the basic reproduction number R0 and the preference-shift threshold are derived.

If this is right

Shortening the infectious period of the preferred host reduces long-term disease persistence.
A threshold value decides whether preference shifts toward the non-preferred host increase or decrease disease burden on the primary host.
Protective measures on the preferred host produce preference shifts that cut prevalence there yet can raise combined prevalence in both hosts.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Interventions must track resulting preference changes, because ignoring adaptation can enlarge the total reservoir of infection.
Similar dynamics in systems with more than two hosts could produce preference hierarchies that stabilize or expand disease maintenance.
Empirical tests could compare observed preference shifts against the model's threshold after single-host interventions.

Load-bearing premise

Vector feeding preferences evolve according to specific adaptive rules under disease-control pressure, and the system contains exactly two hosts.

What would settle it

Field measurements of vector biting rates and host-specific prevalence before and after protective interventions on the preferred host, checking whether prevalence falls in the preferred host while total prevalence across hosts rises.

Figures

Figures reproduced from arXiv: 2604.27113 by Anupama Sharma, Shravani Shetgaonkar.

**Figure 1.** Figure 1: Effect of recovery rate on the behavior of R0 as a function of vector preference. R0 is varying with αv for different values of µ1 in (a) and δ2 in (b). Black circle indicates the minimum value of R0 which is attained at αvc for all the curves. The value of R0 converges to R0a with further increase in αv. R0a =1.563 for all curves in (b), indicating its value is independent of δ2. Lowest value of R0 is att… view at source ↗

**Figure 2.** Figure 2: Variation of R0 against changes in vector preference and host population for increasing recovery rates. R0 as a function of αv is varying with: Nh1 for different values of µ1 in (a)-(d) and Nh2 for different values of δ2 in (e)-(h). Darker gradient colours correspond to higher values of R0. Green curve represents the threshold value of the host population at which R0 attains maximum. The highest value of R… view at source ↗

**Figure 3.** Figure 3: Effect of threshold R0c on equilibrium disease prevalence with changing vector preference. R0c is varied by increasing βhv in (a)-(d) and µ1 in (e)-(h). The blue, green, and cyan colour curves correspond to equilibrium prevalence in h1, h2, and vector population i.e., y ∗ 1 , y∗ 2 , and y ∗ 3 . The dashed line denotes the value of αv at αvc. Behavior of y ∗ 1 is hump shaped when αv < αvc for lower R0c valu… view at source ↗

**Figure 4.** Figure 4: Pairwise invasibility plots for vector trait to bite h2 for relative encounter rate at (a) Eh2 Eh1 = 0.5, (b) Eh2 Eh1 = 0.95, (c) Eh2 Eh1 = 1.5, (d) Eh2 Eh1 = 5. The growth of the mutant strategy is indicated by the shaded region representing positive invasion fitness. The white area indicates negative invasion fitness. h1 is more abundant in (a)-(b) and h2 is more abundant in (c)-(d). Positive value for σ… view at source ↗

**Figure 5.** Figure 5: Evolutionary Singular strategies σ ∗ h2, obtained for different trade off strength, (1/χ) for various level of encounter rates. The trade-off between performance and preference is varied from strong (χ < 1) to weak (χ > 1). Note that in (a), the value of σ ∗ h2 occurs between 0 and 1 for weak trade of strength values. In (b)-(d), the value of σ ∗ h2 is high for intermediate trade-off values and decreases a… view at source ↗

**Figure 6.** Figure 6: Dependence of evolutionary singular strategy of vector population on the trade-off strength. The parameter values used were such that host encounter rates were equal and the conditions (9) were satisfied. That is, Eh1 = Eh2=1, q2 =0.1, τ2=1 day, τf=0.2 day, frc = 4 and rest parameters were as in view at source ↗

**Figure 7.** Figure 7: The value of evolutionary singular point, R0, equilibrium prevalence in h1, h2, vector population, equilibrium prevalence in host population (h1 and h2) obtained for different χ values. The shaded region denotes the range of χ for which value of αv < 1. For very weak values of trade-off strength, the prevalence decreases in h2 and increases in h1. Overall prevalence among the host population peaks in the s… view at source ↗

**Figure 8.** Figure 8: Behavior of y ∗ 1 − y ∗ 1 (γ = 1, αv = 1) as a function of αv and γ. Values are centered by subtracting y ∗ 1 obtained under conditions of no vector preference and equal transmission ability of the two hosts. Positive and negative deviations are coloured in red and blue to identify the amplification and dilution effects (increase or decrease relative to the reference value). Amplification effect occurs whe… view at source ↗

read the original abstract

Vector-borne diseases often infect multiple host species, increasing the likelihood of disease persistence due to the presence of multiple reservoirs. Vector biting patterns and feeding preferences can shift in response to selective pressures introduced by disease control interventions, altering the dynamics of transmission. In this paper, we develop a mathematical model that couples host diversity and adaptive vector behavior with vector-borne disease transmission dynamics, focusing on a system with two hosts and a vector population exhibiting preference for one host. We derive the basic reproduction number, $R_0$, a threshold that determines the existence of two equilibria in our model, and obtain conditions that can lead to the long-term persistence of the disease. Our analysis suggests that shortening the infectious period of the vector's preferred host is an effective control strategy. We also identified a threshold that determines whether shifting vector preference toward a non-preferred host amplifies or reduces the disease burden on the primary preferred host. Our results show that protective measures for the preferred host can trigger adaptive shifts in vector preferences, reducing disease prevalence in that host. However, this shift may lead to an increase in overall host prevalence.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This two-host model finds that protecting the preferred host can shift vector preferences and raise overall prevalence, but the preference evolution rules are ad hoc and untested.

read the letter

The main thing to know is that the model shows protective measures on one host can backfire by pushing vectors toward the second host, cutting prevalence in the first but lifting it overall, while shortening the preferred host's infectious period looks like a more reliable control move. It also derives an R0 and a threshold separating when preference shifts amplify versus reduce burden on the main host. That coupling of diversity with adaptive preferences is the actual extension here, and it is a legitimate step beyond single-host setups even if it stays within standard ODE frameworks. The derivations of equilibria conditions and the threshold are the parts that hold up on the abstract level. The soft spots are bigger than minor. The rules for how preferences evolve are presented as given functional forms without derivation from data, first principles, or alternative kernels, and the stress-test concern lands because no sensitivity checks appear for those choices or for adding hosts. The two-host limit and lack of parameter values or validation steps make the control recommendations dependent on the specific setup. This is for mathematical epidemiologists who already work on multi-host vector-borne models and want to explore adaptive behavior. A reader could pull the threshold idea for their own extensions, but they would need to re-derive and test the dynamics themselves. It deserves a serious referee because the question of how preference shifts interact with diversity in interventions is worth checking, even with the current gaps in grounding and robustness. Send it for review and ask referees to focus on the evolutionary assumptions and any numerical results.

Referee Report

2 major / 2 minor

Summary. The paper develops a mathematical model for vector-borne disease transmission in a two-host system (one preferred) where vector feeding preferences evolve adaptively in response to selective pressures from disease control. It derives the basic reproduction number R0, identifies a threshold separating amplification versus reduction of disease burden on the preferred host under preference shifts, obtains conditions for disease persistence and equilibria, and concludes that shortening the infectious period of the preferred host is an effective control strategy while protective measures on the preferred host can reduce its prevalence but increase overall host prevalence.

Significance. If the central results hold under justified adaptive dynamics, the work would be significant for highlighting trade-offs in vector-borne disease control when vector behavior evolves, particularly the potential for preference shifts to offset interventions on preferred hosts. It contributes by coupling host diversity with evolutionary vector responses in an ODE framework, offering qualitative insights into persistence thresholds that could guide policy if validated against data.

major comments (2)

[Model formulation] Model formulation section: the adaptive rules governing vector preference evolution are introduced as specific functional forms driven by selective pressure without derivation from first principles, empirical data on feeding behavior, or comparison to alternative kernels (e.g., frequency- vs. density-dependent updating). This is load-bearing because the sign of the identified preference-shift threshold and the net effect of shortening the preferred host's infectious period both depend on these rules; modest changes could reverse the amplification/reduction outcome or the control recommendation.
[Results] Results on R0 and threshold: no sensitivity analysis is reported for the free parameters (preference adaptation rate, infectious period of preferred host) or for relaxing the strict two-host topology. The threshold separating amplification from reduction of burden on the primary host is obtained under the fixed model structure, so its robustness and the persistence conditions cannot be assessed without exploring alternative evolutionary dynamics or additional hosts.

minor comments (2)

[Abstract] Abstract and introduction: equations for R0 and the preference threshold are referenced but not displayed, making it difficult to follow the derivation steps without immediately consulting the main text.
[Discussion] Notation: the distinction between 'overall host prevalence' and prevalence in the preferred host should be clarified with explicit definitions or symbols when discussing the trade-off under preference shifts.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments on our manuscript. We address each major comment below and indicate the revisions that will be incorporated.

read point-by-point responses

Referee: [Model formulation] Model formulation section: the adaptive rules governing vector preference evolution are introduced as specific functional forms driven by selective pressure without derivation from first principles, empirical data on feeding behavior, or comparison to alternative kernels (e.g., frequency- vs. density-dependent updating). This is load-bearing because the sign of the identified preference-shift threshold and the net effect of shortening the preferred host's infectious period both depend on these rules; modest changes could reverse the amplification/reduction outcome or the control recommendation.

Authors: We agree that the specific functional forms for adaptive preference evolution are central to the results and that their justification merits expansion. These forms were chosen to represent selective pressure from control measures in a manner consistent with standard adaptive dynamics frameworks in evolutionary epidemiology. We did not derive them from first principles or empirical feeding data, nor compare them directly to alternatives such as frequency- versus density-dependent updating. In revision we will expand the model formulation section with additional justification, relevant citations on vector behavior, and a discussion of alternative kernels. We will also add a supplementary analysis using an alternative updating rule to confirm that the qualitative thresholds and control recommendations remain robust. A complete first-principles derivation would require new empirical studies outside the scope of this theoretical paper. revision: partial
Referee: [Results] Results on R0 and threshold: no sensitivity analysis is reported for the free parameters (preference adaptation rate, infectious period of preferred host) or for relaxing the strict two-host topology. The threshold separating amplification from reduction of burden on the primary host is obtained under the fixed model structure, so its robustness and the persistence conditions cannot be assessed without exploring alternative evolutionary dynamics or additional hosts.

Authors: The absence of sensitivity analysis is a valid observation. We will add a dedicated sensitivity analysis examining variation in the preference adaptation rate and the preferred host's infectious period, reporting effects on R0, the preference-shift threshold, and equilibrium prevalence. For the two-host assumption, we will extend the discussion section to outline how the core mechanisms extend to additional hosts while noting that quantitative thresholds would require case-specific parameterization. These additions will directly address robustness of the reported thresholds and persistence conditions. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivations are standard model analysis

full rationale

The paper constructs a coupled ODE system for two-host vector-borne transmission with an adaptive preference equation, then applies standard next-generation matrix methods to obtain R0 and bifurcation thresholds for equilibria and preference shifts. These quantities are explicit functions of the model parameters and rates; they do not reduce to the inputs by definition, nor are they obtained by fitting a subset of data and relabeling the fit as a prediction. No load-bearing self-citations or imported uniqueness theorems appear in the provided abstract or description. The adaptive rules are posited as modeling assumptions rather than derived results, so the subsequent threshold analysis remains a direct consequence of the stated equations rather than a tautology.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claims rest on unstated functional forms for preference adaptation, transmission rates between two hosts, and the definition of R0 in terms of those rates; several parameters governing infectious periods and preference thresholds are introduced without independent evidence.

free parameters (2)

preference adaptation rate
Rate at which vectors shift feeding preference in response to host infectiousness or control measures; value not provided but central to the adaptive behavior claim.
infectious period of preferred host
Length of time the preferred host remains infectious; shortening it is claimed as effective control, implying it is a tunable or fitted parameter.

axioms (2)

domain assumption Vector feeding preference evolves adaptively according to selective pressures from disease and control interventions
Core modeling assumption stated in the abstract for the two-host system.
standard math Disease persists or dies out based on a threshold R0 derived from the coupled transmission and preference dynamics
Standard threshold analysis in compartmental models, invoked to determine equilibria.

pith-pipeline@v0.9.0 · 5505 in / 1469 out tokens · 81865 ms · 2026-05-07T10:16:13.371616+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

43 extracted references

[1]

Martinez, A

J. Martinez, A. Showering, C. Oke, R. T. Jones, J. G. Logan, Differential attraction in mosquito-human interactions and implications for disease control, Philosophical Transactions of the Royal Society B 376 (2021) 20190811

2021
[2]

N. M. Sene, B. Diouf, A. Gaye, A. Gueye, F. Seck, C. T. Diagne, I. Dia, D. Diallo, M. Diallo, et al., Blood feeding patterns of aedes aegypti populations in senegal, The American Journal of Tropical Medicine and Hygiene 106 (2022) 1402. 19

2022
[3]

J. Long, K. Maskell, R. Gries, S. Nayani, C. Gooding, G. Gries, Synergistic attraction of west- ern black-legged ticks, ixodes pacificus, to CO2 and odorant emissions from deer-associated microbes, Royal Society Open Science 10 (2023) 230084

2023
[4]

Ferraguti, A

M. Ferraguti, A. D. Martins, Y. Artzy-Randrup, Quantifying the invasion risk of west nile virus: Insights from a multi-vector and multi-host SEIR model, One Health 17 (2023) 100638

2023
[5]

M. M. de Wit, A. Dimas Martins, C. Delecroix, H. Heesterbeek, Q. A. Ten Bosch, Mech- anistic models for west nile virus transmission: a systematic review of features, aims and parametrization, Proceedings of the Royal Society B 291 (2024) 20232432

2024
[6]

Fiatsonu, A

E. Fiatsonu, A. Deka, M. L. Ndeffo-Mbah, Effectiveness of systemic insecticide dog treatment for the control of chagas disease in the tropics, Biology 12 (2023) 1235

2023
[7]

V. M. Schneider, J. E. Visone, C. T. Harris, F. Florini, E. Hadjimichael, X. Zhang, M. R. Gross, K. Y. Rhee, C. Ben Mamoun, B. F. Kafsack, et al., The human malaria parasite plasmodium falciparum can sense environmental changes and respond by antigenic switching, Proceedings of the National Academy of Sciences 120 (2023) e2302152120

2023
[8]

D. O. Escrivani, V. Scheidt, M. Tinti, J. Faria, D. Horn, Competition among variants is pre- dictable and contributes to the antigenic variation dynamics of african trypanosomes, PLOS Pathogens 19 (2023) 1–21

2023
[9]

Vanaerschot, S

M. Vanaerschot, S. Huijben, F. Van den Broeck, J.-C. Dujardin, Drug resistance in vector- borne parasites: multiple actors and scenarios for an evolutionary arms race, FEMS Microbi- ology Reviews 38 (2014) 41–55

2014
[10]

G. H. Palmer, T. Bankhead, H. S. Seifert, Antigenic variation in bacterial pathogens, in: Virulence Mechanisms of Bacterial Pathogens, John Wiley & Sons, Ltd, 2016, pp. 445–480

2016
[11]

Ross, The prevention of malaria, John Murray, 1911

R. Ross, The prevention of malaria, John Murray, 1911

1911
[12]

Macdonald, Epidemiological basis of malaria control, Bulletin of the World Health Orga- nization 15 (1956) 613–626

G. Macdonald, Epidemiological basis of malaria control, Bulletin of the World Health Orga- nization 15 (1956) 613–626

1956
[13]

Chitnis, J

N. Chitnis, J. M. Hyman, J. M. Cushing, Determining important parameters in the spread of malaria through the sensitivity analysis of a mathematical model, Bulletin of Mathematical Biology 70 (2008) 1272–1296

2008
[14]

Chamchod, N

F. Chamchod, N. F. Britton, Analysis of a vector-bias model on malaria transmission, Bulletin of Mathematical Biology 73 (2011) 639–657

2011
[15]

K. S. Rock, D. A. Wood, M. J. Keeling, Age-and bite-structured models for vector-borne diseases, Epidemics 12 (2015) 20–29

2015
[16]

M. N. Anwar, L. Smith, A. Devine, S. Mehra, C. R. Walker, E. Ivory, E. Conway, I. Mueller, J. M. McCaw, J. A. Flegg, et al., Mathematical models of plasmodium vivax transmission: A scoping review, PLOS Computational Biology 20 (3) (2024) e1011931

2024
[17]

S. L. Robertson, K. A. Caillouët, A host stage-structured model of enzootic west nile virus transmission to explore the effect of avian stage-dependent exposure to vectors, Journal of theoretical biology 399 (2016) 33–42. 20

2016
[18]

D. M. Bichara, Global analysis of multi-host and multi-vector epidemic models, Journal of Mathematical Analysis and Applications 475 (2019) 1532–1553

2019
[19]

M. H. Zahid, C. M. Kribs, Decoys and dilution: the impact of incompetent hosts on prevalence of chagas disease, Bulletin of Mathematical Biology 82 (2020) 41

2020
[20]

Yakob, M

L. Yakob, M. B. Bonsall, G. Yan, Modelling knowlesi malaria transmission in humans: vector preference and host competence, Malaria Journal 9 (2010) 1–7

2010
[21]

J. E. Simpson, P. J. Hurtado, J. Medlock, G. Molaei, T. G. Andreadis, A. P. Galvani, M. A. Diuk-Wasser, Vector host-feeding preferences drive transmission of multi-host pathogens: West nile virus as a model system, Proceedings of the Royal Society B: Biological Sciences 279 (2012) 925–933

2012
[22]

R. C. Rivera, S. Bilal, E. Michael, The relation between host competence and vector-feeding preference in a multi-host model: Chagas and cutaneous leishmaniasis, Mathematical Bio- sciences and Engineering 17 (2020) 5561–5583

2020
[23]

Bilal, R

S. Bilal, R. Caja Rivera, A. Mubayi, E. Michael, Complexity and critical thresholds in the dynamics of visceral leishmaniasis, Royal Society Open Science 7 (12) (2020) 200904

2020
[24]

L. Chen, S. Chen, P. Kong, L. Zhou, Host competence, interspecific competition and vector preference interact to determine the vector-borne infection ecology, Frontiers in Ecology and Evolution 10 (2022) 993844

2022
[25]

S. L. Wu, J. M. Henry, D. T. Citron, D. Mbabazi Ssebuliba, J. Nakakawa Nsumba, H. M. Sánchez C, O. J. Brady, C. A. Guerra, G. A. García, A. R. Carter, et al., Spatial dynamics of malaria transmission, PLOS Computational Biology 19 (6) (2023) e1010684

2023
[26]

Miller, A

E. Miller, A. Huppert, The effects of host diversity on vector-borne disease: The conditions under which diversity will amplify or dilute the disease risk, PLoS One 8 (2013) e80279

2013
[27]

Marini, R

G. Marini, R. Rosá, A. Pugliese, H. Heesterbeek, Exploring vector-borne infection ecology in multi-host communities: A case study of west nile virus, Journal of Theoretical Biology 415 (2017) 58–69

2017
[28]

S. A. Geritz, J. A. Metz, É. Kisdi, G. Meszéna, Dynamics of adaptation and evolutionary branching, Physical Review Letters 78 (1997) 2024

1997
[29]

Diekmann, J

O. Diekmann, J. Heesterbeek, M. G. Roberts, The construction of next-generation matrices for compartmental epidemic models, Journal of the Royal Society Interface 7 (2010) 873–885

2010
[30]

Wiggins, Introduction to applied nonlinear dynamical systems and chaos, Springer, 1990

S. Wiggins, Introduction to applied nonlinear dynamical systems and chaos, Springer, 1990

1990
[31]

Stone, K

C. Stone, K. Gross, Evolution of host preference in anthropophilic mosquitoes, Malaria journal 17 (2018) 1–11

2018
[32]

Ravigné, U

V. Ravigné, U. Dieckmann, I. Olivieri, Live where you thrive: joint evolution of habitat choice and local adaptation facilitates specialization and promotes diversity, The American Naturalist 174 (2009) E141–E169. 21

2009
[33]

Le Menach, S

A. Le Menach, S. Takala, F. E. McKenzie, A. Perisse, A. Harris, A. Flahault, D. L. Smith, An elaborated feeding cycle model for reductions in vectorial capacity of night-biting mosquitoes by insecticide-treated nets, Malaria journal 6 (2007) 1–12

2007
[34]

Gandy, E

S. Gandy, E. Kilbride, R. Biek, C. Millins, L. Gilbert, No net effect of host density on tick- borne disease hazard due to opposing roles of vector amplification and pathogen dilution, Ecology and Evolution 12 (2022) e9253

2022
[35]

Keesing, R

F. Keesing, R. D. Holt, R. S. Ostfeld, Effects of species diversity on disease risk, Ecology Letters 9 (2006) 485–498

2006
[36]

E. E. Kemibala, A. Mafra-Neto, T. Dekker, J. Saroli, R. Silva, A. Philbert, K. Nghabi, L. E. Mboera, A zooprophylaxis strategy using l-lactic acid (abate) to divert host-seeking malaria vectors from human host to treated non-host animals, Malaria Journal 19 (2020) 1–7

2020
[37]

Asale, L

A. Asale, L. Duchateau, B. Devleesschauwer, G. Huisman, D. Yewhalaw, Zooprophylaxis as a control strategy for malaria caused by the vector anopheles arabiensis (diptera: Culicidae): a systematic review, Infectious Diseases of Poverty 6 (2017) 1–14

2017
[38]

Keesing, R

F. Keesing, R. S. Ostfeld, Dilution effects in disease ecology, Ecology Letters 24 (2021) 2490– 2505

2021
[39]

R. S. Levine, D. L. Hedeen, M. W. Hedeen, G. L. Hamer, D. G. Mead, U. D. Kitron, Avian species diversity and transmission of west nile virus in atlanta, georgia, Parasites & Vectors 10 (2017) 1–12

2017
[40]

Y. A. Mwalugelo, W. P. Mponzi, L. L. Muyaga, H. H. Mahenge, G. C. Katusi, F. Muhonja, D. Omondi, A. O. Ochieng, E. W. Kaindoa, F. A. Amimo, Livestock keeping, mosquitoes and community viewpoints: a mixed methods assessment of relationships between livestock management, malariavectorbitingriskandcommunityperspectivesinruraltanzania, Malaria Journal 23 (1) ...

2024
[41]

+d3)Y 2 3 −w 1(d1 +δ 1)Z 2 1 . 23 Setw 1 = αvc1y∗ 3 µ1 , then we can derive that˙Wis negative definite if the following are true: 2α2 vc2 1(1−y ∗ 1 −z ∗ 1)2 d3(µ1 +d 1) < w 3 < d3(µ1 +d 1) 2c2 2α2 vN 2 h1(1−y ∗ 3)2 ,(B.1) w32c2 2N 2 h2(1−y ∗ 3)2 (δ2 +d 2)d3 < w 2 < (δ2 +d 2)d3w3 2c2 1(1−y ∗ 2)2 .(B.2) From here, it follows that a positivew3 can be chosen ...
[42]

+u3c2αvNh1(1−y ∗ 3)(y1 −y ∗ 1)(y3 −y ∗ 3) +u 2c1(1−y ∗ 2)(y3 −y ∗ 3)(y2 −y ∗
[43]

+u3c2Nh2(1−y ∗ 3)(y3 −y ∗ 3)(y2 −y ∗ 2)−(µ 1 +d 1)(y1 −y ∗ 1)2 −u 2(d2 +δ 2)(y2 −y ∗ 2)2 −u 3d3(y3 −y ∗ 3)2 −α vc1y3(y1 −y ∗ 1)2 −u 1(d1 +δ 1)(z1 −z ∗ 1)2 −u 2c1y3(y2 −y ∗ 2)2 −u 3c2αvNh1y1(y3 −y ∗ 3)2 −u 3c2Nh2y2(y3 −y ∗ 3)2. By selectingu 1 asu 1 = αvc1y∗ 3 µ1 , ˙Ubecomes negative definite if the following conditions hold true: α2 vc2 1(1−y ∗ 1 −z 1max)...