**5. Conclusions**

Here, it was demonstrated that using 50 or 100 mosquitoes with the rabbit gives similar results with no systematic bias for both pyrethroid and pyrethroid-chlorfenapyr ITNs. The lower density can be used for the WHO tunnel test when testing pyrethroid Interceptor® and pyrethroid-chlorfenapyr Interceptor® G2. Reducing the number of mosquitoes per test decreases its cost and allows a larger number of net samples to be tested at a time. Larger sample sizes will give greater precision when estimating ITN efficacy since the unit of replication in ITNs testing is the bioassay (cone, tunnel, I-ACT, experimental hut) and not the mosquito within that assay. Furthermore, we provide the first evidence that membrane feeding systems can be used as an alternative to rabbit bait in WHO tunnel assays. Membrane assay shows an excellent comparison to the gold-standard WHO tunnel test on both the mortality and feeding success endpoint for the ITNs tested, although control feeding success is lower due to the lower attraction of the membrane to hostseeking mosquitoes. Using membrane feeding systems instead of rabbits or other animals in WHO tunnel assays resolves the ethical issues concerning animal welfare and makes the tests simpler to perform. Further work to improve the feeding success of the membrane

feeding system as a replacement for rabbits in the WHO tunnel test is needed, as mosquito feeding success impacts insecticide induced mortality.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/insects13070562/s1, Figure S1: Mean percentage mortality at (A) 24-h and (B) 72-h with rabbit-100, (C) at 24-h and (D) 72-h with membrane-50 of blood fed and unfed resistant *Anopheles arabiensis* in the WHO tunnel test. The Red dashed line depicts (the WHO mortality threshold ≥ 80%). Figure S2: Bland and Altman plot of (A) blood feeding success (BFS) and (B) 72-h mortality (M72) for Interceptor® and Interceptor® G2 with rabbit or membrane bait against 100-pyrethroid resistant *Anopheles arabiensis* in the WHO tunnel test with a 12 h exposure time. The average value for both methods is plotted on the *x*-axis and the mean difference between methods on the *y*-axis. The solid line in the middle shows the mean difference with a 95% confidence interval of the mean difference represented by the dashed lines. Figure S3: Bland and Altman plot of A-blood feeding success (BFS) and B-72-h mortality (M72) for Interceptor® with susceptible *Anopheles gambiae* and Interceptor® G2 with resistant *Anopheles arabiensis* using rabbit bait and a density of either 100 or 50 mosquitoes in the WHO tunnel test with a 12-h exposure time. The average value for both densities is plotted on the *x*-axis and the mean difference between densities on the *y*-axis. The solid line in the middle shows the mean difference with a 95% confidence interval of the mean difference represented by the dashed lines. Figure S4: Bland and Altman plot of A-blood feeding success (BFS) and B-72-h mortality (M72) for Interceptor® with susceptible *Anopheles gambiae* and Interceptor® G2 with resistant Anopheles arabiensis using rabbit bait and a density of either 100 mosquitoes or membrane bait and a density of 50 mosquitoes in the WHO tunnel test with a 12-h exposure time. The average value for both densities is plotted on the *x*-axis and the mean difference between densities on the *y*-axis. The solid line in the middle shows the mean difference with a 95% confidence interval of the mean difference represented by the dashed lines. Table S1: Mean percentage mortality and 95% confidence interval (95% CI) for the negative control, Interceptor® G2, and Interceptor® at 24 h post exposure (M24) and mortality at 72 h post exposure (M72) and blood feeding success (BFS) or blood feeding inhibition (BFI) of resistant *Anopheles arabiensis* with 12 h of exposure time for rabbit, membrane and human arm and 1 h exposure time for in membrane and human arm in the WHO tunnel test. The negative control thresholds for the WHO tunnel test are blood feeding success ≥ 50% and M24 ≤ 10%. Table S2: Mean percentage mortality and 95% confidence interval (95% CI) for the negative control, susceptible *Anopheles gambiae* with Interceptor® resistant *Anopheles arabiensis* with Interceptor® G2 at 24 h post exposure (M24) and mortality at 72 h post exposure (M72) and blood feeding success (BFS) or blood feeding inhibition (BFI) with the density of 50 or 100 mosquitoes in the WHO tunnel test. The negative control thresholds for WHO tunnel test are blood feeding success ≥ 50% and M24 ≤ 10%. Table S3: Mean percentage mortality and 95% confidence interval (95% CI) for the negative control, resistant *Anopheles arabiensis* with Interceptor ® or Interceptor® G2 at 24 h post exposure (M24) and mortality at 72 h post exposure (M72) and blood feeding success (BFS) or blood feeding inhibition (BFI) with rabbit bait and a density of 100 mosquitoes (rabbit-100) or membrane bait with 50 mosquitoes (membrane-50) in the WHO tunnel test. The negative control thresholds for WHO tunnel test are blood feeding success ≥ 50% and M24 ≤ 10%.

**Author Contributions:** The conceived and design of the experiment study was done by D.S.K., S.J.M. The experiment was performed by D.S.K. Data analysis was conducted by D.S.K., S.J.M. and O.G.O., D.S.K. wrote the manuscript. L.H. drew the diagrams. The manuscript was critically revised by S.J.M., E.M., O.G.O., and L.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research was funded by the Ifakara health Institute Vector Control Product and Testing Unit through MSc project no 02.

**Institutional Review Board Statement:** Ethical approval was obtained from the Ifakara Health Institute Institutional Review Board (IHI/IRB/No25-2021) and the National Institute of Medical Research (NIMR/GQ/R.8a/Vol.IX/3893). Permission to publish was granted by the Tanzania National Institute of Medical Research letter with ref NIMR/HQ/P.12 VOL XXXIV/39.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data set for this study is available on reasonable request from Vector Control Product and Testing Unit of Ifakara Health Institute.

**Acknowledgments:** Our kind thanks to Kafuruki Shubis, IHI, VCPTU administration, insectary and testing teams, and staff of IHI/NM-AIST Bagamoyo Campus for supporting this work. We appreciate the testing and insectary technicians for their willingness to participate in human arm feeding experiments. We thank B.A.S.F. for the donation of the ITNs used in these experiments. Thanks go to Stephen G. Mbwambo for participating in conduct of the experiment and constructive advice, Ahmed Bakari for his support during the conduct of the experiment, Ummi Abdul for assistance in data analysis, Mgeni M. Tambwe for reviewing the early stage of the methods and discussion, and Jason Moore for the resources, administration, and validation of the project activities. We thank Natalie Lissenden of Liverpool School of Tropical Medicine for her helpful comments on an early draft.

**Conflicts of Interest:** D.S.K., S.G.M. and L.H., have no competing interests. S.J.M., O.G.O. and E.M. conduct product evaluations for several vector control product manufacturers including B.A.S.F. that manufacture the ITNs tested. B.A.S.F. was not involved in the work or the decision to publish.
