# README
Flexible nutrient oxidation during desiccation enhances climate stress responses of an invasive fly

---

---
METHODOLOGICAL INFORMATION

Fly collection, handling and rearing
Ceratitis capitata was selected as a desiccation-tolerant species and Ceratitis rosa was selected to represent a desiccation-sensitive species. Mean body size did not differ significantly between the species but females were significantly larger than males in both (Appendix S1). Ceratitis capitata and C. rosa were field obtained from infested fruit collected in the Stellenbosch and Nelspruit areas of South Africa, respectively, and reared with standard methods (Appendix S1). All measurements were performed on laboratory reared, F2 generation fruit flies.

Stable isotope (13C) tracer molecules
2.1 Tracer molecules and diet preparation
For both species, the diets of three parallel groups of larvae reared under otherwise identical conditions (Appendix S1) were enriched with either U-13C6-glucose (Glu), 1-13C palmitic acid (Palm), or 1-13C- leucine (Leu) (Cambridge Isotope Laboratories, Tewksbury, MA, USA) at a concentration of 2.0 g.L-1. Assuming that all were equally digested, assimilated, and mobilized from the flies’ body, the 13C labels from each ingested nutrient would be incorporated into body tissues. Labelled glucose, palmitic acid or leucine result in more 13C in carbohydrate, lipid, or protein body stores, respectively. By sampling gas (exhaled, oxidised breath) from flies of each species fed different 13C labelled larval diets, we were then able to determine which body store was utilized during desiccation stress. If 13C was more prominent in the gas samples from one of the groups with enriched Glu, Palm or Leu, we could then infer that the targeted storage nutrient was being metabolised.
2.2 Sampling intervals
All measurements were on five-day old flies reared under benign conditions (25°C, 76 %RH). Following the experimental design of [29], three sampling intervals were selected: 100% survival (T0; no desiccation stress), immediately following desiccation pre-treatment for 90% survival (T1), and flies subjected to desiccation pre-treatment until 90% survival, then allowed to recover for 24 hours without food and water (T2) (Figure 1). The time in hours (h) that 90% of C. capitata and C. rosa were predicted to withstand desiccation were taken from previously published estimates [27-28]. From this data, the 90% survival time for C. capitata were taken as 36 h and as 24 h for C. rosa while the 50% survival time were taken as 48 h for C. capitata and as 33 h for C. rosa.

Respirometry and breath samples
The concentration of labelled 13C isotopes (molar flux rates) oxidised during desiccation stress were calculated in both fly species with repeated alternating measures of gas ‘breath’ samples and respirometry trials; the protocols of which were adapted from the tsetse fly experiments of [25] and adjusted to the mass and metabolic rate of individual tephritid flies determined in [30].
3.1 Respirometry
Respirometry equipment were set up as described in S1. To determine how metabolic rate and nutrient utilisation changes over a period of desiccation stress in two species of flies with different desiccation physiology, individuals from each species and treatment group were subjected to an alternating regimen of metabolic rate measurement followed by a rest (incubation) period. Respirometry measurements were taken under desiccating conditions (25°C and 0% RH) in CO2-free air with flow through respirometry systems (LiCor 7000 CO2/H2O gas analyser,) at a flow rate of 100ml.h-1, using similar methods as in [27]. Metabolic rates were recorded as carbon dioxide production (ml.h-1) using flow-through respirometry.
Ten five-day old males and females were randomly selected from each species and treatment, and weighed to the nearest 0.1 mg (Mettler Toledo, MS104S/01). Metabolic measurements were recorded from females and males (n = 7 each) at 25°C in separate 5 ml chambers with a push mode 8-channel multiplexing respiratory system (RM8 Intelligent Multiplexer, V5, Sable Systems). Each of the seven fly-containing channels were measured for 15 min per channel, and the remaining empty channel recorded a baseline CO2 reading, later used to drift-correct metabolic measurements of flies if necessary. These measurements were repeated for each of the three sampling intervals, with new individuals used in each measurement. The remaining three individuals were weighed along with the rest of the flies and their mass specific metabolic rates estimated based on the scaling relationships determined from the seven flies tested and existing data for the species (e.g., as in [26]).
A electronic activity detector (Sable Systems AD-1) was connected to at least one of the respirometry chambers per multiplexer run, to qualitatively identify resting or active periods. To keep the flies quiescent in the respiratory chamber, natural light was excluded by wrapping the system in aluminium foil and closing the bath lid.Gas concentrations were recorded on LiCor software and Sable Systems DATACAN V software (Expedata) was used to and calculate metabolic rate (V ̇_(CO_2 )) using standard equations as per [29]

3.2 Stable isotope gas ‘Breath’ samples
Immediately following the 2h respirometry trial the individual 20 ml respiratory chambers were sealed with a luer cap, thus sealing in the dried air and continuing the desiccation exposure treatments.
The respiratory chambers were then placed in an incubator under standard conditions (25°C) for 16 h to allow CO2 concentrations to accumulate to a detectable amount [i.e., >2%], after which the breath sample was collected. The duration (h) required for flies to produce this amount of CO2 depends on the metabolic rate (ml.h-1). Basson et al. (2012) reported the metabolic rate of C. capitata to be between 0.01 - 0.014 ml.h-1, while another study found a metabolic rate of 0.084 ml.h-1 for the same species (pilot trial, 2018). Using these values, we estimated that an individual fly would take between 5 h (using values from pilot trails) and 33 h (using values form Basson et al. 2012) to produce enough CO2. Since three bouts of breath samples were collected to measure changes in nutrient metabolism over time, the total time to collect these samples were between 15 and 99 h. However, both C. capitata and C. rosa have average survival times of <50 h when exposed to desiccating conditions (Weldon et al. 2016), therefore a breath sampling collection time of 16 h was used, as this falls between the two values calculated while also permitting fly survival for the full duration of the experiment.
Breath samples of individual flies were collected by ejecting 15 ml of the gas from each individual plastic syringe into evacuated Exetainer vials (Labco Limited, Lampeter, UK). After a breath sample was taken, the respiratory chambers were flushed with room air with standard atmospheric concentrations of CO2 (i.e., 0.04%), and reconnected to the multiplexer.
The amounts of labelled 13C (δ13C values in terms of international VPDB standards) in each breath sample were measured with a Helifan Plus (Fischer, ANalysen, Instrumente, Germany) non-dispersive infrared spectrometer interfaced with a FanAS auto sampler (as described by 29). The spectrometer was calibrated by running vials containing CO2 with known δ13C before each batch of unknown sample.
Three rounds of respirometry (at sampling intervals 1-3), each followed by a 16-hour rest period were conducted to gain an overview of which nutrients were utilized by the flies during desiccation. After the final round of respirometry the flies were flash frozen in liquid nitrogen and stored at -80°C for later biochemical analysis. Macro nutrient composition (body water, lipid, protein and carbohydrate stores were determined for each species, tracer and timepoint (30; S2).

3.3.3 Data analysis
Lower metabolic rates are linked to reduced respiratory water loss, therefore we compared the mass specific (V ̇_(CO_2 )) (ml.h-1.mg-1) between C. capitata and C. rosa at different time points and acclimation treatments with repeated measures ANOVA, to test if differences in metabolic rate drive previously observed differences in desiccation physiology [24] in these two species.
We expected metabolism of specific macronutrients during desiccation stress would differ between the two species, therefore the amounts of 13C tracers detected in the gas samples was predicted to differ between C. capitata and C. rosa. To test this prediction, we compared the 13C traces in the carcasses and gas (breath) samples between the two species at each time point (Figure 1) for the different acclimation treatments, with unpaired t-tests. The 13C tracers present in treated flies were reported as the amounts relative to control flies, and therefore the atom fraction excess (AFE) was modelled for each species, treatment and sex from the equation of [19]. In this equation the VPDB is a constant (IAEA, 2000). At each time point the instantaneous rates of tracer oxidation (T) for each of the species, sexes and acclimation treatments were calculated as the product of the metabolic rate (V ̇_(CO_2 )) of each fly with the related AFE (for that specific sample), divided by the molar mass of each tracer (m) and volume of CO2 produced per gram of mixed substrate oxidized using a value of 1.01 g-1 (K) [19]. The amounts of each tracer metabolised over time were calculated as the cumulative T over the sampling points and compared between species, sex, and acclimation treatment with unpaired t-tests.
If the metabolism of specific nutrients such as lipids are linked to the desiccation physiology of fruit flies, we would expect that the flies will either store different amounts of these nutrients or metabolise them at different rates. The amounts of protein, lipid, and carbohydrate present before, during, and after recovery from desiccation stress were compared between C. capitata and C. rosa with repeated measures ANOVAs, to determine how much of each nutrient is stored by each species under benign conditions and which of these nutrient stores are metabolised once flies are required to mount a stress response (S2-S3).

Statistical analyses were performed using R (R core team, 2013) and Statistica version 13 (Statsoft, Tulsa, Oklahoma,s possible.
DATA-SPECIFIC INFORMATION FOR: SI_T-C_rosa

[](https://)
Number of variables: 10

Number of cases/rows: 378

Variable List: Species, Tracer, Time_point, Sex, Round, Replicate, Percentage_CO2, T-C_13C, Individual_CO2_ml_h, VCO2 ml/min pMol/hour

Missing data codes: NA

Specialized formats or other abbreviations used: T-C (Treatment value - control value)

DATA-SPECIFIC INFORMATION FOR: SI_T-C_capitata

Number of variables: 10

Number of cases/rows: 316

Variable List: Species, Tracer, Time_point, Sex, Round, Replicate, Percentage_CO2, T-C_13C, Individual_CO2_ml_h, VCO2 ml/min pMol/hour

Missing data codes: NA

Specialized formats or other abbreviations used: T-C (Treatment value - control value)

DATA-SPECIFIC INFORMATION FOR: Bosua et al_Metabolic_tracer_all data

Number of variables: 22

Number of cases/rows: 694

Variable List: Species, Tracer, Time_point, Sex, Round, Replicate, Percentage_CO2, T-C_13C, Delta At%, VCO2 ml/h, Individuals, Individual_CO2_ml_h, bkgd 13C, bkdg At%, APE, AFE, K, VCO2_ml/min, molar_mass, nMol/min, nMol/h, pMol/hour,

Missing data codes: NA

Specialized formats or other abbreviations used: T-C (Treatment value - control value)

---
## Sharing/access Information

Links to other publicly accessible locations of the data:

Was data derived from another source?No
If yes, list source(s):