Fig 1.
Diagram of the experimental facility from Laboratorio de Ecología Acuática (LEA), Universidad San Francisco de Quito (USFQ), located in Papallacta, Ecuador.
Fig 2.
Water temperature during the microcosm experiment.
HOBO data loggers recorded temperatures (1162 sampling points) from chambers in the control/ambient [S1] scenario (blue), +2.5°C [S2] scenario (yellow), and +5°C [S3] scenario (red) throughout the 10-day experiment. The X-axis shows recorded temperatures from the HOBO data logger with date and hour (12:00 pm). The Y-axis represents the temperatures in degrees Celsius (°C).
Fig 3.
Experimental facility located in the Papallacta region, Napo Province, Ecuador.
Insect collection site in the Cayambe Coca National Park and location of the microcosm facility. The rivers and streams in this area are part of the Napo subbasin, Amazon Basin.
Fig 4.
Larvae of aquatic insect species.
(A) Mayfly larvae, Andesiops peruvianus (Ulmer, 1920). (B) Caddisfly larvae, Anomalocosmoecus illiesi (Marlier, 1962). Reprinted from [www.ex-situphotography.com/es/proyectos-publicaciónes/proyectos-publicaciones] under a CC BY license, with permission from [Vieira, J. ExSitu], original copyright [©ExSitu, all rights reserved].
Table 1.
One-way ANOVA estimates (mean ), standard deviation (SD), and F values of physical and chemical parameters analyzed in water from chambers of the mesocosm at three temperature scenarios: S1 (control), S2 (+2.5°C), and S3 (+5°C).
p value (95%). η2 generalized effect size (eta-squared).
Fig 5.
Fine particulate organic matter in three temperature scenarios.
Boxplots of the fine particulate organic matter collected during the ten-day mesocosm experiment. The central line represents the median FPOM content. The lines above and below the box are the highest and lowest scores, respectively. The first and third quartiles are represented in the box. S1 (control), S2 (+2.5°C) and S3 (+5°C).
Table 2.
Coefficients from the GLM (family = quasibinomial) of mortality rates for Andesiops peruvianus and Anomalocosmoecus illiesi in different temperature scenarios.
Estimates were in link = logit with confident high (conf.high) and low (conf.low) values. Temp: S1 (control) S2 (+2.5°C) and S3 (+5°C). Significance coefficients (C I = 95%) are in bold.
Table 3.
Analysis of deviance of the GLM with a binomial distribution (link = logit.) of larval survival (%) influenced by dissolved oxygen (O2) and mean temperature (T) during the mesocosm experiment.
Significant terms are in bold (CI = 95%).
Table 4.
Mean (), standard deviation (SD), and standard error (SE) of body length, wing-pad length and dry mass from Andesiops peruvianus and Anomalocosmoecus illiesi head capsule width and dry mass.
S1 (control), S2 (+2.5°C) and S3 (+5°C). n, number of individuals measured.
Table 5.
LMM performances of growth rates of Andesiops peruvianus and Anomalocosmoecus illiesi with mean FPOM as fixed factors and chamber as a random factor.
Species [SP] was added as a covariate. The models tested were the null model, SP+FPOM without interaction, and SP x FPOM with single and interaction effects. The AIC best model is in bold.
Fig 6.
Aquatic insect growth rates subjected to rising temperatures.
Growth rates of two aquatic insect species during a ten-day experiment. (A). Andesiops peruvianus growth rates (upper plot). (B). Anomalocosmoecus illiesi growth rates (bottom plot). Black dots are the mean growth rate. Lines represent confidence intervals (CI = 95%). The X-axis represents the estimated mean growth rate (mg mg mg-1 day-1). The Y-axis represents the temperature scenarios of climate change. Temp: S1 (control), S2 (+2.5°C), and S3 (+5°C).
Table 6.
The LMM was used to compare the growth rates of Andesiops peruvianus [A.p] and Anomalocosmoecus illiesi [A.i] with the mean FPOM during the microcosm experiment.
Estimates were extracted from the best candidate model with fixed and random effects. Significant coefficients are in bold (CI = 95%). Temperature scenarios S1(control), S2 (+2.5°C), and S3 (+5°C). LRT: likelihood ratio test statistic; and logLik: logarithmic likelihood of random effect (chamber).