You can use Henderson and Tilton formula to remove mortality in control also.

Dear Tigar,

If you used several doses and assessed their mortalities, you can calculate LC50, LC90 values and their fiducial limits with probit analysis. 

You can use mortality corrected by Abbott's formula, W. 1925. A method of computing the effectiveness of an insecticide. J. Econ. Entomol. 18: 265- 267. If you have a reference dose of literature (LC50 and LC90) for the insect, you can compare the few insects (n = 20) a that dose with Chi-square test that dose. We use this method to assess the loss of susceptibility of populations to insecticides where we obtained very few insectos in our study with codling moth in apple orchards (Reyes et al 2015).

Thanks for the answers. please note we tested disturbance not a chemical pesticide, so are looking for an novel approach to the analysis.

Dear Tigar,

In this case not LC50 but EC50 (efficient dose) values can be counted. You should chose and assess a  well  determined feature. The calculation method is the same.  I think the disturbance had some characteristic symptoms. 

You can try Kaplan-Meier Survival Analysis: Log-Rank It was designed for medical studies but I have used to compare the mortalities attributed to different treatments. here is a useful reference 

Rich JT, Neely JG, Paniello RC, Voelker CCJ, Nussenbaum B, Wang E. W. 2010. A practical guide to understanding Kaplan-Meier curves. Otolaryngology–

Head and Neck Surgery 143: 331-336.

Hope it helps

To compare survivorship of two cohorts you should do a Survival Analysis. Several methods are implemented in R (see Chapter 25 in Crawleys R book). I used Cox's proportional hazard model (coxph-function) to compare survival of crickets under predation risk of spiders (Bucher, R., Binz, H., Menzel, F., & Entling, M. H. (2014). Spider cues stimulate feeding, weight gain and survival of crickets. Ecological Entomology, 39(6), 667-673.).

You simply need the time point of death for each individual. The models also allows censoring, in the case you loose an individual during your experiment.

Hi Tigara, I also used Kaplan– Meier estimate and log-rank test to determine insects survivor distribution, I believe it should be suitable for you as well, http://onlinelibrary.wiley.com/doi/10.1111/j.1570-7458.2009.00831.x/abstract

good luck

I noticed that the previous answers have covered the material needed. Hope it works with you

I agree with Roman, you can use cox proportional hazard model, using R it's simple enough. You can censor data (in case one individual left the study before it dies) or use mixed models, in case you have a random variable to take into account. See coxph and coxme packages in R.

Henderson and Tilton's formula would be appropriate to nullify the mortality in control also so as to find exact corrected mortality in the treated one.

Is the cochrane test appropriate to analysis insect mortality rate data?