Logistic Regression is a better Method of Analysis Than Linear Regression of Arcsine Square Root Transformed Proportional Diapause Data of<i>Pieris melete</i>(Lepidoptera: Pieridae)

Shi, Penglan; Hu, Hongchang; Xiao, Hai

doi:10.1653/024.096.0361

Cited by 10 publications

(5 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Each OTU count distribution was evaluated for normality with quantile-quantile plots and the Shapiro-Wilk test for normality 66 . A logistic regression with unrarefied count data with an offset was selected over a linear regression with proportion data from the rarefied table since the data were not normally distributed and evidence suggests logistic regressions perform better than arcsine-transformed data in linear regressions 67 . For all models, temperature, nutrients, and scarring were assessed as fixed effects and factorial interaction terms and tank and colony as separate random effects.…”

Section: Methodsmentioning

confidence: 99%

Multiple stressors interact primarily through antagonism to drive changes in the coral microbiome

Maher

Rice

McMinds

et al. 2019

Sci Rep

View full text Add to dashboard Cite

Perturbations in natural systems generally are the combination of multiple interactions among individual stressors. However, methods to interpret the effects of interacting stressors remain challenging and are biased to identifying synergies which are prioritized in conservation. Therefore we conducted a multiple stressor experiment (no stress, single, double, triple) on the coral Pocillopora meandrina to evaluate how its microbiome changes compositionally with increasing levels of perturbation. We found that effects of nutrient enrichment, simulated predation, and increased temperature are antagonistic, rather than synergistic or additive, for a variety of microbial community diversity measures. Importantly, high temperature and scarring alone had the greatest effect on changing microbial community composition and diversity. Using differential abundance analysis, we found that the main effects of stressors increased the abundance of opportunistic taxa, and two-way interactions among stressors acted antagonistically on this increase, while three-way interactions acted synergistically. These data suggest that: (1) multiple statistical analyses should be conducted for a complete assessment of microbial community dynamics, (2) for some statistical metrics multiple stressors do not necessarily increase the disruption of microbiomes over single stressors in this coral species, and (3) the observed stressor-induced community dysbiosis is characterized by a proliferation of opportunists rather than a depletion of a proposed coral symbiont of the genus Endozoicomonas .

show abstract

Section: Methodsmentioning

confidence: 99%

Multiple stressors interact primarily through antagonism to drive changes in the coral microbiome

Maher

Rice

McMinds

et al. 2019

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Proportion data were logit-transformed, following Warton and Hui (2001) and Shi et al (2013). All the data points collected from the last 30 min of each experiment were used.…”

Section: Discussionmentioning

confidence: 99%

Private information alone can trigger trapping of ant colonies in local feeding optima

Czaczkes

Salmane

Klampfleuthner

et al. 2016

Journal of Experimental Biology

View full text Add to dashboard Cite

Ant colonies are famous for using trail pheromones to make collective decisions. Trail pheromone systems are characterised by positive feedback, which results in rapid collective decision making. However, in an iconic experiment, ants were shown to become 'trapped' in exploiting a poor food source, if it was discovered earlier. This has conventionally been explained by the established pheromone trail becoming too strong for new trails to compete. However, many social insects have a well-developed memory, and private information often overrules conflicting social information. Thus, route memory could also explain this collective 'trapping' effect. Here, we disentangled the effects of social and private information in two 'trapping' experiments: one in which ants were presented with a good and a poor food source, and one in which ants were presented with a long and a short path to the same food source. We found that private information is sufficient to trigger trapping in selecting the poorer of two food sources, and may be sufficient to cause it altogether. Memories did not trigger trapping in the shortest path experiment, probably because sufficiently detailed memories did not form. The fact that collective decisions can be triggered by private information alone may require other collective patterns previously attributed solely to social information use to be reconsidered.

show abstract

“…To test for an effect of runway type (vertical or horizontal) on pheromone deposition, the following model formula was used: To test whether ants were more likely to choose the low- or high-effort associated cues on the maze experiment, we analyzed ant decision data using an exact binomial test. To test whether ants spent more time in the olfactometer fields offering low- or high-effort associated cues, we analyzed the logit-transformed (Shi, Sand Hu, & Xiao, 2013; Warton & Hui, 2011) proportion of time ants spent in each field using the following formula, with a Gaussian error structure: …”

Section: Methodsmentioning

confidence: 99%

Greater effort increases perceived value in an invertebrate.

Czaczkes

Brandstetter

Stefano

et al. 2018

Journal of Comparative Psychology

View full text Add to dashboard Cite

Expending effort is generally considered to be undesirable. However, both humans and vertebrates will work for a reward they could also get for free. Moreover, cues associated with high-effort rewards are preferred to low-effort associated cues. Many explanations for these counterintuitive findings have been suggested, including cognitive dissonance (self-justification) or a greater contrast in state (e.g., energy or frustration level) before and after an effort-linked reward. Here, we test whether effort expenditure also increases perceived value in ants, using both classical cue-association methods and pheromone deposition, which correlates with perceived value. In 2 separate experimental setups, we show that pheromone deposition is higher toward the reward that requires more effort: 47% more pheromone deposition was performed for rewards reached via a vertical runway (high effort) compared with ones reached via a horizontal runway (low effort), and deposition rates were 28% higher on rough (high effort) versus smooth (low effort) runways. Using traditional cue-association methods, 63% of ants trained on different surface roughness, and 70% of ants trained on different runway elevations, preferred the high-effort related cues on a Y maze. Finally, pheromone deposition to feeders requiring memorization of one path bifurcation was up to 29% higher than to an identical feeder requiring no learning. Our results suggest that effort affects value perception in ants. This effect may stem from a cognitive process, which monitors the change in a generalized hedonic state before and after reward. (PsycINFO Database Record

show abstract

Logistic Regression is a better Method of Analysis Than Linear Regression of Arcsine Square Root Transformed Proportional Diapause Data ofPieris melete(Lepidoptera: Pieridae)

Cited by 10 publications

References 2 publications

Multiple stressors interact primarily through antagonism to drive changes in the coral microbiome

Multiple stressors interact primarily through antagonism to drive changes in the coral microbiome

Private information alone can trigger trapping of ant colonies in local feeding optima

Greater effort increases perceived value in an invertebrate.

Contact Info

Product

Resources

About