Growth enhancement in broiler chickens receiving CHEMEQ^RTM polymeric antimicrobial

Abstract

This study investigated the use of CHEMEQ^RTM polymeric antimicrobial as a means of enhancing health and growth rates in broiler chickens. One-day-old Cobb-500 broilers were reared to 42 days. In trial 1, 100 birds received CHEMEQ^RTM polymeric antimicrobial via water, and 100 birds were untreated. Treated birds had significantly greater (P<0.0001) average daily gain (ADG), significantly lower (P<0.0016) feed conversion ratios, significantly lower (P<0.05) mortality, and a significantly lower (P<0.0042) viscosity of their ileal contents. In trial 2, three groups of 40 birds received reducing dosages of CHEMEQ^RTM polymeric antimicrobial, and a fourth group remained untreated. Groups receiving CHEMEQ^RTM polymeric antimicrobial had significantly greater (P<0.05) ADG, and two groups had significantly lower (P<0.05) feed conversion ratios than the controls. In trial 3, groups of 40 birds received either CHEMEQ^RTM polymeric antimicrobial, Salinomycin or Lasalocid in-feed, or were untreated. Birds receiving CHEMEQ^RTM polymeric antimicrobial had significantly greater (P<0.05) ADG than those receiving Salinomycin. CHEMEQ^RTM polymeric antimicrobial helped to maintain the health and to improve the growth performance of broiler chickens.

1 Introduction

For over 50 years antibiotics have been administered to poultry to promote feed efficiency and improve rates of weight gain (Gustafson & Bowen, 1997). Usually the antibiotics are administered at subtherapeutic levels. Although most growth-promoting antimicrobials target Gram-positive bacteria (Butaye et al., 2003), some, such as the quinoxalines, target Gram-negative bacteria (Suter et al., 1978). The mechanisms by which these agents promote growth are not exactly known, but are believed to be associated with their antibacterial effects. Suggested mechanisms include protection of nutrients from bacterial destruction, improved absorption of nutrients due to thinning of the small intestinal wall, decreased production of bacterial toxins, and reduced incidence of subclinical infections (Butaye et al., 2003).

Resistance of bacterial pathogens to antibiotics is recognized as an increasing problem, and is being widely studied (Lucy et al., 2002; Vandemaele et al., 2002). As a result of fears about transfer of antimicrobial resistance from animal bacteria to human bacterial pathogens, and of human acquisition of resistant zoonotic bacteria from food animals (Van den Bogaard & Stobberingh, 1999), the use of antibiotics in animals is undergoing increasing regulation (Tanner, 2000). To compensate for removal of antimicrobial agents, and to maintain health and productivity, various compounds have been introduced for use by the intensive animal industries. Recently, a novel antimicrobial agent, CHEMEQ^RTM polymeric antimicrobial, has been developed (Melrose & Huxham, 1999), and has been shown to be useful for controlling post-weaning diarrhoea in piglets (Hampson et al., 2000; Hampson & Murdoch, 2003). In vitro studies have shown that CHEMEQ^RTM polymeric antimicrobial exerts biocidal activity against a range of bacteria, bacterial spores, fungi and viruses, and furthermore the molecular weight of CHEMEQ^RTM polymeric antimicrobial limits its absorption through the gut wall (unpublished data). CHEMEQ^RTM polymeric antimicrobial has a different mode of action to antimicrobials in current use; it contains reactive aldehyde groups that act by initially adsorbing to, then denaturing, surface proteins of microorganisms, thereby killing them in a non-selective way. The non-specific mode of action means that its use is unlikely to contribute to further antibiotic resistance in enteric pathogens of food-producing animals or human beings.

The purpose of the current work was to evaluate the efficacy and safety of CHEMEQ^RTM polymeric antimicrobial in broiler chickens.

2 Materials and Methods

The work was conducted with the approval of the Murdoch University Animal Ethics Committee.

2.1 Trial 1

Two hundred Cobb-500 one-day-old broiler chickens were purchased from a commercial hatchery. They were transported to Murdoch University, identified with leg tags, and allotted in a randomized complete block design to two groups of 100 chickens (groups 1 and 2). Each group had four replicates (pens 1 to 4) of 25 chickens per pen, housed in adjacent pens. Feather sexing was used to ensure an even distribution of male and female chickens in the pens. Artificial light was provided for 12 h/day. The chickens were weighed on admission to the trial, then every 7 days until day 42. Group 1 was allocated as the negative control, and received no treatment. Group 2 was allocated as the treatment group receiving 0.1% (w/v) CHEMEQ^RTM polymeric antimicrobial (Chemeq Ltd, Bentley, Australia; 30 mg/kg body weight/day) via a medicated watering system. Dosages for CHEMEQ^RTM polymeric antimicrobial were calculated every 7 days for each pen in accordance with the trial protocol, and increased accordingly as the group's weight increased. The chickens receiving treatment commenced medication on day 1. Each day, following consumption of the medicated water, fresh water was supplied ad libitum. Water to the untreated negative control group was also available ad libitum. Feed was available ad libitum, changed daily, and pen feed consumption and feed conversion ratios (FCRs) were recorded. The initial diet was a commercial chick starter ration, followed after 14 days by a grower ration (18% crude protein) containing 125 parts/10⁶ dinitolmide (D.O.T. 250 Premix, CCD Animal Health, Australia) for the prevention of coccidiosis. No other antibiotics were included in the diet for the chickens. The birds were housed on pine sawdust, and each pen was provided with a commercial infrared heat lamp. Visual assessments and clinical examinations of the birds were made on a daily basis. All birds that died were subjected to a full postmortem examination conducted by experienced veterinary pathologists who were blinded to the nature of the experiments being conducted.

The remaining chickens were euthanized by cervical dislocation on day 42 of the trial, and then weighed before necropsy. All chickens underwent a thorough gross examination. The gastrointestinal tract was divided into portions, and the following portions weighed: total gastrointestinal tract (GIT) (glandular stomach to terminal colon); small intestine (proximal duodenum to ileum); and large intestine (caecae and colon). For bacteriology, within each pen samples from subgroups of five birds were pooled and thoroughly mixed, so that overall 20 pooled samples from each intestinal site were available from each of the two experimental groups. Sterile bacteriology swabs were inoculated from these pooled digesta samples from the crop, glandular stomach, muscular stomach (gizzard), proximal duodenum, mid-jejunum, ileum, caeca, colon and cloaca. Pooled samples from the terminal ileum were collected for measurements of digesta viscosity.

2.2 Trial 2

Trial 2 utilized 120 Cobb-500 one-day-old broiler chickens from the same source as in trial 1. The birds were allotted in a randomized complete block design into four (identified as groups 1 to 4) replicates of 30 chickens per pen, housed in adjacent pens. Group 1 received 0.1% (w/v) CHEMEQ^RTM polymeric antimicrobial (30 mg/kg body weight/day), group 2 received 0.033% (10 mg/kg body weight/day), and group 3 received 0.01% (3 mg/kg body weight/day). Group 4 was an untreated negative control. The rest of the trial was conducted in an identical manner to trial 1, except that bacteriology swabs were taken from the cloacae of birds in groups 1 and 4 once a week from weeks 2 to 6. At postmortem the same procedures were followed as in trial 1; however, only six pools of digesta samples were available from each of the four groups. Pooled samples from both the ileums and caecae were tested for their viscosity. For bacteriology, pooled digesta samples from the mid-jejunum, caecum and colon were used for bacterial counts.

2.3 Trial 3

Trial 3 was conducted in the same way as the two previous trials. One hundred and twenty birds were divided into four groups. Group 1 received 0.01% (w/v) CHEMEQ^RTM polymeric antimicrobial in water (3 mg/kg body weight/day). Group 2 received 120 parts/10⁶ Salinomycin in their feed (60 g/t) (Sacox 120 powder; Intervet, Australia), Group 3 received 95 parts/10⁶ Lasalocid in their feed (633 g/t) (Avatec 150 Premix; Alpharma, Australia), and group 4 acted as an untreated negative control group.

The chickens were euthanized on day 42 of the trial, weighed and then necropsied. At necropsy all chickens underwent a thorough gross examination. No other measurements were made.

2.4 Assessment of coliform and clostridial numbers

In trial 1, a semi-quantitative estimation of coliform numbers in the digesta along the GIT was made at the time of postmortem examination. Swabs loaded with approximately equal quantities of digesta were used to inoculate Eosin Methylene Blue Agar Base plates (EMB), according to the technique AS 1766.1.4–1991 (Oxoid, UniPath, UK), and incubated at 37°C for ≤24 h. The extent of coliform growth on the plates was estimated in a semi-quantitative fashion, with a score of 1 for the organisms present only in the inoculum, 2 for organisms also present in the first streak, 3 for organisms also present in the second streak, and 4 for organisms also present in the third streak and beyond.

In trial 2, the cloacal swabs collected at weekly intervals were used for assessment of Clostridium perfringens numbers. Swabs were broken into 1 ml of 0.1% peptone water, and the tubes thoroughly shaken. Serial 10-fold dilutions were then made in peptone water. From each dilution, l00 μl was inoculated in duplicate on Tryptose Sulfite Cycloserine Agar Base plates (TSC), according to technique AS 1766.2.8–1991 (Oxoid), and spread evenly. The liquid was allowed to soak through the agar and then an overlay of 10 ml molten TSC was used to cover the plates. These were incubated at 37°C for ≤24 h in jars in an anaerobic environment consisting of 94% H₂ and 6% CO₂, generated using anaerobic Gaspak plus sachets (BBL; Becton Dickinson Microbiology Systems, USA). For estimating coliform numbers at postmortem, the pooled digesta samples were weighed and prepared in a one in 10 dilution with 0.1% peptone water. The samples were vortexed for 30 sec. Serial dilutions of the sample were prepared and aliquots of 100 μl were placed in duplicate on EMB Agar Base plates and incubated at 37°C for ≤ 24 h. Coliform colony counts were then made.

2.5 Viscosity of the ileal and caecal digesta

Measurements of the viscosity of the pooled contents from the ileum (trials 1 and 2) and caecum (trial 2) were performed using a Brookfield LVDV-II⁺cone plate (CP40) rotational viscometer (Brookfield Engineering Laboratories, Inc., Massachusetts, USA). The samples were diluted 1:1 (v/v) with distilled water, mixed and centrifuged at 12 000 g for 8 min. The viscosity of 0.5 ml supernatant fraction was measured at 25°C, applying a shear rate of 60/s (McDonald et al., 2001).

2.6 Statistical analysis

FCRs were calculated as the weight of feed consumed divided by the weight gain of the birds. Bird weights, average daily gain (ADG) and FCRs per pen and group were compared using a post-hoc analysis; the Tukey mean separation test and an analysis of variance (ANOVA). Bacterial scores between the two groups in trial 1 were compared using the Mann–Whitney test, and bacterial counts in trial 2 between the four groups were log transformed and compared by ANOVA, The mortality rates between groups were compared using a chi-squared analysis, and the viscosity results were analysed with a Mann–Whitney test and an unpaired t-test.

3 Results

3.1 Trial 1

The results for the treatment versus control groups were blocked on gender and by pen. There was no significant interaction between gender and treatment (F=0.44, degrees of freedom (d.f.) =1, P=0.51).

There was a significant effect of treatment on the ADG of the chickens over the 42-day period (ANOVA with blocking F=63.7; d.f.=1.6; P<0.0001). This analysis allows for the influence of gender on weight gain. The ADG of the group treated with CHEMEQ^RTM polymeric antimicrobial was 25.2% greater than the untreated control group over the 42-day period (Table 1).

ADG, FCR, mortality, mean viscosity of ileal contents, and median scores for coliforms along the GIT of broiler chickens at postmortem following either the administration of 0.1% w/v CHEMEQ^RTM polymeric antimicrobial or no treatment

	0.1% (w/v) CHEMEQ^RTM polymeric antimicrobial group	Untreated control group	P value^a
Average daily gain (g)^b	30.75±2.55	24.57±1.53	<0.0001
Feed conversion ratio^b	2.5±0.18	2.9±0.31	<0.0166
Mortality	1	10	<0.05
Viscosity (mPa s)(ileum)^b	2.42±0.172	10.63±2.687	0.0042
Median (range) coliform scores along the GIT
Crop	4 (3 to 4)	3 (2 to 5)	NS
Glandular stomach	2 (1 to 4)	1 (0 to 5)	NS
Muscular stomach	1 (0 to 2)	1 (0 to 1)	NS
Proximal duodenum	1 (0 to 1)	1 (0 to 2)	NS
Mid-jejunum	1 (0 to 2)	1 (1 to 3)	NS
Ileum	3 (1 to 4)	3 (1 to 4)	NS
Caecum	4 (3 to 5)	5 (2 to 5)	NS
Colon	3 (1 to 5)	4 (2 to 5)	NS
Cloaca	3 (2 to 4)	3 (2 to 4)	NS
^aNS, not significant.^bValues are mean±standard deviation.

The chickens in the group treated with CHEMEQ^RTM polymeric antimicrobial had a significantly lower FCR than did the untreated control group (F=10.82; d.f. =1.6; P=0.0166). The FCR of the group treated with CHEMEQ^RTM polymeric antimicrobial was 13.8% lower than the untreated control group at the end of the 42-day period (Table 1).

The group treated with CHEMEQ^RTM polymeric antimicrobial had significantly lower mortality rates over the 42-day period in comparison with the untreated control group (χ²<0.05). In the treatment group one chicken was euthanized, on humane grounds, due to a right periorbital swelling that contained a caseous material on postmortem. In the untreated control group 10 chickens died. All chickens underwent postmortem examination. Nine chickens died from a generalized septic colibacillosis, and one from a localized infection with Erysipelothrix rhisiopathiae causing a pneumonia. No chickens in the group treated with CHEMEQ^RTM polymeric antimicrobial died from colibacillosis.

The group treated with CHEMEQ^RTM polymeric antimicrobial had significantly lower ileal viscosity compared with the untreated control group (P<0.0042). Coliform numbers throughout the GIT of the two groups of birds were similar (Table 1).

3.2 Trial 2

Overall, treatment with CHEMEQ^RTM polymeric antimicrobial at different dose rates had a significant effect on weight gain (F=7.04; d.f.=3.111; P=0.0002). The ADGs were significantly different (P<0.05), with the three groups treated with CHEMEQ^RTM polymeric antimicrobial on average gaining between 17.3% and 21.1% more body weight per day than the untreated control group (Table 2). However, there was no significant difference in the ADG between the three groups receiving different concentrations of CHEMEQ^RTM polymeric antimicrobial.

ADG, FCR, digesta viscosity and average scores for coliforms along the GIT of broiler chickens at postmortem following the administration of 0.1%, 0.033% or 0.001% (w/v) CHEMEQ^RTM polymeric antimicrobial, or no treatment

	0.1% (w/v) CHEMEQ^RTM polymeric antimicrobial group	0.033% (w/v) CHEMEQ^RTM polymeric antimicrobial group	0.01% (w/v) CHEMEQ^RTM polymeric antimicrobial group	Untreated control group	P value^a
Average daily gain (g)^b	39.07^A±8.25	40.57^A±11.69	40.31^A±11.26	33.30^B±11.69	<0.05
Feed conversion ratio^b	2.46^A±0.17	2.65^AB±0.22	2.46^A±0.19	2.76^B±0.25	<0.05
Viscosity (mPa s)^b
Ileum	2.36±0.36	2.26±0.35	2.31±0.76	2.08±0.46	NS
Caecum	36.46±9.58	28.75±9.61	40.16±8.15	26.02±11.38	NS
Mean coliform counts along the GIT
Mid-jejunum	0.67×10^8	0.70×10^8	1.10×10^8	1.00×10^8	NS
Colon	6.27×10^8	8.40×10^8	9.10×10^8	1.30×10^9	NS
Caecum	1.10×10^9	1.00×10^9	1.60×10^9	1.50×10^9	NS
^aNS, not significant.^bValues are mean±standard deviation.Within a row, groups with different upper case superscripts differ at the 5% level of significance.

The treatment groups receiving 0.1% and 0.01% CHEMEQ^RTM polymeric antimicrobial, but not that receiving 0.033%, had significantly lower FCR over the 42-day period compared with the untreated control group (P<0.05) (Table 2).

C. perfringens numbers in the faeces were all less than 100 colony forming units/g with only a few colonies cultured in all groups. At postmortem, coliform counts at the small and large intestinal sites in the groups receiving CHEMEQ^RTM polymeric antimicrobial tended to be lower than those of the untreated control group, although not significantly so (Table 2).

No statistically significant differences in digesta viscosity were found between the groups at either intestinal site (Table 2).

One death was recorded in the treatment group receiving 0.01% w/v CHEMEQ^RTM polymeric antimicrobial. The diagnosis, on postmortem, was intestinal dysbacteriosis probably originating from a subacute yolk saculitis. In the treatment group receiving 0.033% w/v CHEMEQ^RTM polymeric antimicrobial, two chickens were euthanized, on humane grounds, due to genetic angular limb deformities. This diagnosis was confirmed on postmortem. No other mortalities occurred.

3.3 Trial 3

There was a significant difference in ADG between all the groups in this trial (F=8.61; d.f. 3. 113; P=0.000). The groups had equal variances (Table 3).

ADG, FCR and mortality rates following the administration of 0.01% (w/v) CHEMEQ^RTM polymeric antimicrobial, 120 parts/10⁶ Salinomycin, 95 parts/10⁶ Lasalocid or not treatment

	0.01% (w/v) CHEMEQ^RTM polymeric antimicrobial group	120 parts/10^6 Salinomycin group	95 parts/10^6 Lasalocid group	Untreated control group	P value^a
Average daily gain (g)^b	41.36^A±5.99	33.62^B±5.99	38.23^B±8.72	38.96^A±11.25	<0.05
Feed conversion ratio^b	2.15±0.17	2.23±0.18	2.38±0.21	2.28±0.21	NS
Mortalities	0	0	1	2	NS
^aNS, not significant.^bValues are mean±standard deviation.Within a row, groups with different upper case superscripts differ at the 5% level of significance.

Using a post-hoc analysis (mean separation test—Tukey), the ADG of the group treated with 120 parts/10⁶ Salinomycin was significantly lower than that of the other groups (P<0.05). Overall, the group receiving 0.01% w/v CHEMEQ^RTM polymeric antimicrobial had the greatest ADG (Table 3). There were no significant differences in FCR between groups.

No mortalities were recorded in the groups receiving 0.01% w/v CHEMEQ^RTM polymeric antimicrobial and 120 parts/10⁶ Salinomycin. One chicken in the group receiving 95 parts/10⁶ Lasalocid died, and two chickens died in the group receiving no treatment. At postmortem the diagnosis for all mortalities was septicaemic colibacilliosis.

At postmortem at the completion of all the experimental studies, none of the chickens receiving CHEMEQ^RTM polymeric antimicrobial showed clinical or pathological changes, or signs of toxicity or drug intolerance.

4 Discussion

The three trials described here were intended to test, in a controlled experimental environment, whether CHEMEQ^RTM polymeric antimicrobial gave growth-enhancing effects when offered to healthy broiler chickens. The conditions in which the birds were held were generally superior to those in commercial production, and it is reasonable to assume that CHEMEQ^RTM polymeric antimicrobial might have resulted in better comparative growth performance under more crowded and less hygienic conditions where there are greater microbial challenges. On the other hand, the chickens were only exposed to 12 h of light per day, and this represents 40% less feeding time compared with birds held under commercial conditions. Hence their full growth potential may not have been met.

The birds used were from a commercial source, and in the first trial there was a considerable amount of ill-health recorded in younger untreated birds, particularly associated with a generalized septic colibacillosis that involved the yolk sac. The commercial suppliers attributed this high incidence of disease to a hatchery problem, with other commercial broiler producers supplied at the same time also reporting increased broiler mortalities. Colibacillosis in broiler chickens is typically a secondary localized or systemic disease that occurs when host immunity is compromised and/or when hatchery management is suboptimal, so that yolk sac infections develop. Besides mortalities, such infections can lead to reduced growth rates, increased feed conversion ratios, reduced flock uniformity, increased susceptibility to other pathogens, and increased treatment and management costs (Barnes & Gross, 1997). Lesions consistent with colisepticaemia are frequently observed in chicken carcasses at processing plants (Yogaratnam, 1995).

Antibiotic growth promoters (AGP) are currently used extensively in the broiler industry to improve growth rates as well as to control enteric disease and colibacillosis. Increasingly, enteric bacteria have become resistant to a wide range of antibiotics, and as a result treatment is becoming problematic. Furthermore, Regulatory Authorities worldwide are concerned about the use of antibiotics in all food-producing animals, and consequently many AGP have been withdrawn from use. The results of trial 1 suggest that CHEMEQ^RTM polymeric antimicrobial may be an effective substitute for some of these AGP. Over the whole of trial 1, birds receiving CHEMEQ^RTM polymeric antimicrobial had significantly greater ADG and a reduced FCR compared with the untreated birds. Furthermore, treatment with CHEMEQ^RTM polymeric antimicrobial assisted in the control of colibacillosis, with no mortalities occurring due to colibacillosis in the treated birds compared with 9% mortality due to colibacillosis in the control birds.

In light of the antimicrobial properties of CHEMEQ^RTM polymeric antimicrobial, and by analogy with other growth promotants, it seems likely that the mechanism(s) involved in the improved growth performance associated with CHEMEQ^RTM polymeric antimicrobial was related to it reducing total bacterial numbers or altering the balance of bacterial populations in the GIT of the treated birds. In future work it would be useful to investigate this, particularly focusing on the Gram-positive microflora that are frequently suppressed by many of the other growth promotants in common use.

In this trial, coliform numbers in the GIT were examined at postmortem in order to determine whether the protection from colibacillosis was due to an effect of CHEMEQ^RTM polymeric antimicrobial on these organism within the GIT. Somewhat unexpectedly, no significant differences in coliform numbers were found between the two groups. It is possible that the cases of colibacillosis were associated with an initial overgrowth of clonal groups of Escherichia coli in the GIT that preceded the septicaemia, and that a similar overgrowth was suppressed by the CHEMEQ^RTM polymeric antimicrobial in the treated birds. Such an effect would not necessarily be reflected by coliform numbers in the GIT of birds examined at 6 weeks of age.

Interestingly, at postmortem the viscosity of the ileal contents in the treated birds in trial 1 was significantly reduced compared with that of the control birds. It is known that reduction of digesta viscosity by the use of exogenous enzymes enhances the growth rate in broilers (Choct et al., 1995). The mechanism(s) by which CHEMEQ^RTM polymeric antimicrobial may reduce digesta viscosity is uncertain, but this observation provides another possible explanation for the enhanced performance of birds receiving the CHEMEQ^RTM polymeric antimicrobial.

Trial 2 was conducted to determine whether lower dose rates of CHEMEQ^RTM polymeric antimicrobial also have a growth enhancing effect. All three dose rates again significantly increased the ADG. Significant improvements in the FCR were also seen with both 1% and 0.1% CHEMEQ^RTM polymeric antimicrobial. These results therefore support the finding that CHEMEQ^RTM polymeric antimicrobial can enhance growth rates in broiler chickens. In this trial overall production was increased compared with in trial 1, and there was little mortality. Unexpectedly, no significant effects were found for digesta viscosity, with the values for the untreated control group resembling those in the birds receiving treatment in both trials. This difference may relate to the overall improved health of the birds in trial 2, and the absence of colibacillosis following improvements made to hatchery management. Coliform numbers in the GIT were again examined in this trial, to determine how they compared with numbers in the first trial. Although not significantly different, numbers tended to be lower in the treated birds than in the controls, and were similar to those found in trial 1. There was no evidence of proliferation of C. perfringens in any of the groups, so the effects of CHEMEQ^RTM polymeric antimicrobial on these Gram-positive anaerobes could not be determined.

In trial 3, the effectiveness of CHEMEQ^RTM polymeric antimicrobial was compared with that of two commercial AGP, the carboxylic ionophore polyether antibiotics Lasalocid and Salinomycin. These are both Streptomyces products that are primarily registered for their anticoccidial activity, but they also are used extensively as growth promoters. They behave as an alkali metal ionophore in altering bacterial cell permeability, complexing with sodium in the cell membrane to cause passive transport of potassium ions out of the cell and their replacement by hydrogen ions. These kill the cell by lowering intracellular pH.

In trial 3 all groups showed good ADG and FCR, but the group receiving CHEMEQ^RTM polymeric antimicrobial performed best. Differences were only significant for ADG, with the group receiving Salinomycin performing worse than the other three groups.

Together the findings from the three trials suggest that CHEMEQ^RTM polymeric antimicrobial will be a useful alternative to current AGPs. The polymer had a positive effect on the health and growth performance of the broilers and, due to its mode of action, it is unlikely to encourage further antimicrobial drug resistance. More extensive field testing of the new antimicrobial polymer is now being undertaken.

Acknowledgments

Thanks are due to Vinnie Wycoco for assistance with the bacteriology, and to Associate Professor Ian Robertson for undertaking the statistical analyses.