inst/lit_review.md

In noamross/age-infects: Package for manuscript and code of age-infection interaction paper

Theory Development

@Anderson1978 - Poisson and Negative-Binomial models of host-parastite interaction, consequences of aggregation

@Anderson1982 - Distributions of parasites are dynamic rather than static. Distributions modified differently by process. Under-dispersion though parasite mortality, density dependence, parasite-indiuced host death. Over-dispersion through host heterogeneity, parasite reproduction. Combinations of processes can generate nonstandard distributions. In single host cohorts, these forces cause changes in mean parasite burden and variance/mean ratio over time. In older hosts, low population sizes make stochasticity important.

@Pacala1988 - Peaked age-parasite number relationship is evidence of parasite mortality, reduced dispersion over time implies density dependence. Generate ML test for this in data.

@Quinnel1995 - Clumped infection can cause changes in dispersion in the absence of density dependence, as does host variability in infection vulnerability

All the above look at a single cohort, not an age-structured population

• Demonstrate that stability is determined by the relation of the

@Adler1992 - Illustrates importance of mean-variance/mean ratio on stability. Shows cycling in population densities and variance to mean ratio, and some transient dynamics, including third moment around the mean, which shows the system can be approximated by three equations asymptotically but only after a transient.

@Kretzschmar1993b - Proof that stability of a model with variable aggregation is due to aggregation - variance/mean ratio must increase with mean. Qualitative differences between neg. binomial and Neyman type-A distributions. Stability criteria and limit cycles shown.

@Kretzschmar1993ba - Shows that infinite and reduced models are equivalent in respect to existence of asymptotic solutions and stability.

@Pugliese1998 - Parasite model in an infinite system with clumped infection (parasites arrive in Poisson-distributed "clumps"). Clumping creates aggregation and enhances model stability.

@Jaenike1996 - Host fitness increases with aggregation (effects stronger with greater mean infection), but parasite fitness decreases (effects weaker with greater mean infection). Effects can be reversed with sigmoid fitness effects on hosts.

Age Structure

@Kretzschmar1989a - Introduces model with continuous age structure. Proves existence and uniqueness of equilibria in general case. In @Kretzschmar1989b shows persistent solutions with exponential growth but stable distributions.

@Pugliese2004 - Also use continuous age structure in infinite, also introducing logistic growth for the host. Shows that $R_0 < 1$ results is parasite-free stable equilibrium, $R_0 > 1$ gives positive stable, unique positive equilibrium. Shows that $R_0$ depends on age-specific mortality, and collapses to same value as assumed in neg. binomial distribution if it is constant. (Note other age-dependent parameters not explored.)

@Pugliese2008 - Repeats @Pugliese2004 with clumped infection, similar results to @Quinnel1995. $R_0$ reduced with clumping.

@Krasnov2006 showed that parasite counts increased with age in rodents, but the pattern varied by species and life history.

Detection and Sampling

@Wilson2002 - Review

@McCallum1995 - Assumption of negative binomial may hamper diagnosis of parasitism, but comparing young to old hosts might be better.

@Duerr2003 - Interpretation of age/infection relationship ambiguous if multiple processes affecting aggregation

Plants

@Waggoner1981 - Lesion distribution in a variety of diseases better fit with negative binomial than poisson, but some not by neg. binomial either:

@Dobson1994 - Many plant diseases may be treated as macro parasites.

@McRoberts2003 - Review of theoretical basis of severity-incidence relationships in plant disease. Reviews neg-binomial, notes that aggregation tends not to remains constant and the relationship is modeled with a power law. (Taylor 1979,1984). Follows on @Hughes1997 - primarily interested in sampling issues.

Mistletoe

@Overton1994 found that larger trees had more mistletoe infections, but this was not driven by more infections by tree size, but by older trees acquiring more infections.

@Martinez1996 found increasing age-infection probability relationship in cactus mistletoes.

@Medel2004 found that Mimus thenca, an mistletoe of cactus, were highly aggregated, driven by vector behavior and plant defenses. @Rio1995 found that larger cacti had more infections but were not more likely to be visited by vectors.

@DelRio1996 also found that larger cacti were more likely to have infections.

@Fadini2009 found that larger cashew trees acquired more mistletoe infections per unit time than smaller trees, though mistletoe survivorship was not affected by host size or quality.

@Rist2009 found that larger hosts had slightly more mistletoes.

@Aukema2002 found taller mesquite were more often visited by birds.

@Roxburgh2007 found size effects after removing age effects, but age effects mattered, too.

Science paper (Hastings) Metapopulation model. Distribution of times of occupation. (science, '03ish).

TPB V52(3) 1997 - Gyllenberg + Hanski

Aggregation is a general law in parasite ecology: @Poulin2007

noamross/age-infects documentation built on May 23, 2019, 9:30 p.m.