*3.1. Differences in Resistance to Antibacterial Agents in E. coli between Sampling Points as Well as between Trials*

Enrofloxacin-resistant *E. coli* isolated from cloacal swabs and manure samples were found at the beginning of trial 1 (T1) and showed significantly higher mean resistance rates than in the other trials (Table 1). In contrast, in trials 2 (T2) and 3 (T3), none of the *E. coli* isolates during the BT stage were resistant to enrofloxacin. There were no significant differences between trial 2 and trial 3 during this stage.

Significant differences could be found between the trials during the AT and ET stages (Table 1). Isolates from the cloacal swabs and manure samples from trial 2 showed the significantly highest resistance to enrofloxacin of the isolates after administering Baytril®, followed by mean values of trial 1 and trial 3 (cloacal swabs: 2.90, 1.98, and 1.00, respectively; manure samples: 2.63, 2.00, and 1.08, respectively; Table 1). Also, at the ET stage, the results of mean enrofloxacin resistance in trial 2 showed the same relationship to the other experiments (Table 1).


**Table 1.** Means of enrofloxacin-resistant *E. coli* isolates from cloacal swab and manure samples from turkeys.

A, B, C means in the same column differ significantly between the experiments (*p* < 0.05); a, b means differ significantly between the stage of sampling within one experiment (*p* < 0.05); \* MICs were summarized and reported as susceptible (S), intermediate (I), and resistant (R). Afterwards the results were classified as 1 = S, 2 = I, or 3 = R and means thereof were calculated; \*\* BT = before treatment; AT = after treatment; ET = end of trial. T1 = untreated antibiotic trial, T2 = treated antibiotic trial, T3 = trial with simulated water spillage containing antibiotic; \*\*\* Cloacal swabs: N = 648; per trial BT: n = 24, AT: n = 96, ET: n = 96; poultry manure: N = 216; per trial BT: n = 24, AT: n = 24, ET: n = 24. G1 = entire floor pen covered with litter; G2 = floor pen covered with litter and having floor heating; G3 = partially (50:50) slatted flooring including an area that was littered; G4 = fully slatted flooring with a sand bath (900 cm2).

When comparing the sampling stages (Table 1), the means in enrofloxacin resistance were significantly different between trial 1 and trial 2 regarding the *E. coli* isolated from the cloacal swabs (Table 1). For the medicated group (T2), the number of samples with isolation of resistant *E. coli* in materials (cloacal swab and manure sample) significantly increased from the BT to AT stages upon exposure to enrofloxacin. Nevertheless, the *E. coli* from all samples showed no significant differences in the resistance between the AT and ET stages (Table 1).

The results of means in resistance of ampicillin resistant *E. coli* isolates in trials 1, 2, and 3 are presented in Table 2. *E. coli* isolates from cloacal swabs were 100% susceptible to ampicillin during the BT stage except in trial 1. In this trial, isolates showed a significantly higher resistance to ampicillin (G1 = 1.33, G2 = 1.00, G3 = 1.00, respectively; Table 2).


**Table 2.** Means of ampicillin-resistant *E. coli* isolates from cloacal swab and litter/excreta samples from turkeys.

A, B, C means in the same column differ significantly between the experiments (*p* < 0.05); a, b means differ significantly between the stage of sampling within one experiment (*p* < 0.05); \* MICs were summarized and reported as susceptible (S), intermediate (I), and resistant (R). Afterwards the results were classified as 1 = S, 2 = I, or 3 = R and means thereof were calculated; \*\* BT = before treatment; AT = after treatment; ET = end of trial. T1 = untreated antibiotic trial, T2 = treated antibiotic trial, T3 = trial with simulated water spillage containing antibiotic; \*\*\* Cloacal swabs: N = 648; per trial BT: n = 24, AT: n = 96, ET: n = 96; poultry manure: N = 216; per trial BT: n = 24, AT: n = 24, ET: n = 24; G1 = entire floor pen covered with litter; G2 = floor pen covered with litter and having floor heating; G3 = partially (50:50) slatted flooring including an area that was littered; G4 = fully slatted flooring with a sand bath (900 cm2).

During the AT stage, a significant difference between the three trials occurred in isolates from the cloacal samples. At this point in time, isolates in T3 showed the significantly highest means in enrofloxacin resistance in cloacal swabs, whereas in manure samples, T2-samples had the highest means. During the ET stage in trial 2, the means of ampicillin-resistant *E. coli* isolates from cloacal swabs were significantly higher than in the other trials (Table 2). In manure samples, no more ampicillin resistance was found in T1 during the ET stage.

The results of mean ampicillin resistance in T2 (animals treated twice with enrofloxacin) differed significantly between the sampling days regarding the *E. coli* strains isolated from the cloacal swabs and manure samples. There was a significant increase in means from the BT to the AT stage (Table 2). In trial 3, there was also a significant increase in means of resistance. The percentage of susceptible isolates changed from 100% susceptible isolates to 46% after simulation of water losses with water containing antibiotic. The significance between the AT and ET stages could not be found in all trials (Table 2).

#### *3.2. Testing the Effect of Different Flooring Designs on the Resistance to Antibacterial Agents in E. coli*

The mean values of resistance of *E. coli* isolates to enrofloxacin and ampicillin depending on sampling stage and flooring design are presented in Figures 2a–d and 3a–d.

**Figure 2.** *Cont.*

**Figure 2.** Means of susceptible (=1); intermediate (=2); and resistant (=3) *E. coli* isolates concerning enrofloxacin resistance in (**a**) cloacal swabs and (**b**) poultry manure samples as well as ampicillin resistance in (**c**) cloacal swabs and (**d**) poultry manure samples before treatment (BT), after treatment (AT) and at the end of trial (ET; cloacal swabs: N = 648; per trial BT: n = 24, AT: n = 96, ET: n = 96; poultry manure: N = 216; per trial BT: n = 24, AT: n = 24, ET: n = 24). T1 = no treatment with antibiotic; T2 = treatment of enrofloxacin via drinking water; and T3 = water (containing enrofloxacin) loss simulation trial. G1 = entire floor pen covered with litter; G2 = floor pen covered with litter and having floor heating; G3 = partially (50:50) slatted flooring including an area that was littered; and G4 = fully slatted flooring with a sand bath (900 cm2). A, B means differ significantly between the groups at one sampling (*p* < 0.05).

**Figure 3.** *Cont.*

**Figure 3.** Percentage of frequency of enrofloxacin minimum inhibitory concentration (MIC) distribution in commensal *E. coli* isolates from (**a**) cloacal swabs after treatment (AT) and (**b**) end of trial (ET) as well as in (**c**) poultry manure samples during AT and (**d**) ET of untreated antibiotic (T1), treated twice with enrofloxacin via drinking water (T2) and simulated water spillage with water containing enrofloxacin (T3) in turkeys (cloacal swabs: N = 576; per trial AT: n = 96, ET: n = 96; poultry manure samples: N = 144; per trial AT: n = 24, ET: n = 24). Rectangle on the x-axis: Clinical Laboratory Standard Institute (CLSI) has determined a veterinary specific breakpoint of ≥2 μg/mL enrofloxacin for *E. coli* from chickens and turkeys.

#### 3.2.1. Development of Enrofloxacin Resistance Depending on Group

Enrofloxacin resistance in *E. coli* isolates from all samples did not show any differences between the groups during the BT stage. During the AT stage in trial 2 (Figure 2a), *E. coli* isolated from cloacal swabs in G3 showed significantly lower means in resistance rates to enrofloxacin than the isolates collected from animals in other groups showing highest possible means (G3: 2.58; mean values of enrofloxacin resistance in *E. coli* for each group in detail in Supplementary Table S1a). During the ET stage in trial 1, G2 showed significantly higher means in resistance values of enrofloxacin in manure samples (Figure 2b). The *E. coli* isolates in T3 acquired from cloacal swabs and manure samples (Figure 2a,b) were susceptible to enrofloxacin and showed no significant differences between groups; 98% and 96%, respectively.

#### 3.2.2. Development of Ampicillin Resistance Depending on Group

Regarding ampicillin resistance in all trials (Figure 2c,d), there were no significant differences between the groups concerning resistance in isolates from cloacal swabs and manure samples during the BT stages. In contrast, *E. coli* isolates from cloacal swabs during the AT stage in trial 2 (Figure 2c) from G1 demonstrated higher resistance means than in the other groups (G1: 1.92; mean values of ampicillin resistance in *E. coli* for each group in detail in Supplementary Table S1b). In trial 2, the results of means of ampicillin resistance from the manure samples during the AT stage also showed higher values for G1 than observed in either G2, G3, or G4 (2.33, 1.00, 1.00, and 1.00, respectively; Figure 2d; mean values of ampicillin resistance in *E. coli* for each group in detail in Supplementary Table S2b). There was no difference in means of ampicillin-resistant *E. coli* isolated from cloacal swabs and manure samples during the ET stage in all trials (Figure 2c,d).
