AU2012201839B2 - Detection of Tunicamines - Google Patents

Abstract

A method of detecting a tunicamine in a sample said method comprising the step of using a near infrared reflectance spectrum of the sample to detect the tunicamine.

Description

1 DETECTION OF TUNICAMINES FIELD OF THE INVENTION The present invention relates to methods for detecting and quantifying tunicamines and to systems for generating and delivering the results of analyses carried out using the methods. The present invention also relates to methods of preparing a tunicamine calibration for use in NIR spectroscopy. BACKGROUND To THE INVENTION Tunicaminyl-uracil compounds, including corynetoxins, are potent hepato- and neuro toxins because they specifically inhibit, at a cellular level, protein glycosylation in mammals, The contamination by corynetoxins in hays, straws, pastures and grains is a major concern for livestock industries. Corynetoxins infect some annual grasses and are produced when the nematode, Anguina spp., carries the bacterial organism, Rathyibacter toxicus ("R. toxicus") or similar species, into the developing seed heads of the grasses and fonns galls. Under some conditions, the bacterium out-competes the ncmatode in these galls in what are referred to as 'bacterial galls' and further under some conditions the bacterium produces corynetoxins. Products which contain seed heads contaminated with corynetoxins, such as hays, straws, pastures or grains can cause annual ryegrass toxicity (ARGT). Corynetoxins cause potent neurological disruption and often can be fatal. Similar toxins are involved in flood-plain staggers and Stewart's Range syndrome. At present there are several assays a) for the bacterial galls (visual assessment), b) for the causative bacterium (immunoassay), and c) for the corynetoxins (high pressure liquid chromatography (HPLC) or immunoassay. The current most common method for detection is immunoassay for the causative bacterium, R. toxicus, although the presence of R, toxicus is not always an indicator of the presence or the amount of corynetoxins. The present invention seeks to improve on the current approaches or at the very least seeks to provide an alternative thereto. RECEIVED TIME 22. FEB. 12:37 2 SUMMARY OF THE INVENTION The present invention provides a method of detecting and quantifying a tunicamine in a sample from a feedstuff said method comprising the step of using a near infrared reflectance spectrum of the sample to detect and quantify the tunicamine. The present invention also provides a method of detecting and quantifying a tunicamine in a sample from a feedstuff said method comprising the steps of: (i) producing a near infrared reflectance spectrum of the sample; and (ii) using said spectrum to detect and quantify the tunicamine. The present invention also provides a method of preparing a quantitative tunicamine calibration for use in NIR spectroscopy, the method comprising the steps: (i) collecting NIR spectra of a suitable number of samples from a feedstuff with a known tunicamine value; and (iii) using the spectra collected in (i) to prepare the quantitative tunicamine calibration. The method may further comprise the step of using the spectra in step (i) above to prepare a prediction (validation) equation. The present invention also provides a system comprising (i) means for electronically receiving a request to detect and quantify the amount of at least one tunicamine in a sample from a feedstuff, wherein the request includes a NIR reflectance spectrum of the sample; (ii) means for comparing the spectrum to a database that correlates the presence and amount of the tunicamine to known near infrared reflectance spectra; RECEIVED TIME 22. FEB. 12:7 3 (iii) means for detecting and calculating the amount of the tunicamine; and (iv) means for electronically reporting the results. It is to be recognised that other aspects, preferred forms and advantages of the present invention will be apparent from the specification including the detailed description, examples, drawings and claims provided below. BRIEF DESCRIPTION OF DRAWINGS In order to facilitate a better understanding of the present invention, preferred embodiments are described herein with reference to the accompanying drawings, in which: Figure 1 is a graph showing near infrared reflectance spectra (NIRS) of an oaten hay to which different amounts (0, 2, 10, 20 and 3 0 tg) of tunicamycin were added, indicated by the different lines; Figure 2A is a graph depicting the results of PLSR analysis of the NIR spectra concerning corynetoxins extract, tunicamycin, and corynetoxins extract plus tunicamycin mixture (50:50 mixture) (total tunicamines) in Table 5 and in which the solid line depicts the calibration regression, and the dotted line depicts the prediction (validation) regression; Figure 2B is a graph depicting the results of PLSR analysis of the NIR spectra concerning corynetoxins extract and tunicamycin (50:50 mixture) in Table 5 and in which the solid line depicts the calibration regression and the dotted line depicts the prediction (validation) regression; Figure 2C is a graph depicting the results of PLSR analysis of the NIR spectra concerning tunicamnycin in Table 5 and in which the solid line depicts the calibration regression and the dotted line depicts the prediction (validation) regression; Figure 2D is a graph depicting the results of PLSR analysis of the NIR spectra concerning corynetoxin extract in Table 5 and in which the solid line depicts the calibration regression and the dotted line depicts the prediction (validation) regression; RECEIVED TIME 22. FEB. 12:37 4 Figure 3A is a graph depicting the results of prediction of the concentrations of corynetoxins in "spiked" hays in Table 6; Figure 3B is a graph depicting the results of prediction of the tunicamycin concentrations in "spiked" hays in Table 6; and Figure 3C is a graph depicting the results of prediction of the concentrations of corynetoxins and tunicamycin concentrations (50:50 mixtures) in "spiked" hays in Table 6. DETAILED DESCRIPTION OF THE INVENTION Method of detecting a tunicamine The present invention provides a method of' detecting a tunicamine in a sample said method comprising the step of using a near infrared (NIR) reflectance spectrum of the sample to detect the tunicamine. A range of tunicamines can be detected with the method of the present invention. Preferably, the tunicamine comprises Formula I: 0 O OHHH H ~ OH OH O C N O H4 H H H HO OH

NHCOCH

3 0

HOH

2 C OH OH (I) RECEIVED TIME 22. FEB. 12:37 5 wherein R is an aliphatic chain. Preferably the aliphatic chain is selected from the group consisting of: (i) S15a, H16i, Ul6i, Ul6n, H17i, Sl6i, S16n, U17i, Sl7a, H18i, U18i, H19a, U18n, S18i, Ul9a, S19a, H17a or Ul7a; where S = saturated fatty acid; H = p-hydroxy fatty acid; U= ap unsaturated fatty acid; the numeral refers to the chain length including the carbonyl; and a = anteiso (EtCH(Me)CH 2 -; i = iso (Me 2

CHCH

2 -); and n = normal chain; (ii) a fatty acid selected from p-Hydroxy a13-Unsaturated Saturated

C

16 , C 1 (preferably iso C 16 -CIS (preferably iso Ci6, Cis (preferably configuration) configuration) iso configuration)

C

17 , C1 9 (preferably anteiso Co, C 19 (preferably anteiso C 15 , C 17 , C 19 configuration) configuration) (preferably anteiso configuration)

C

14 , C16 (preferably normal configuration C 1 5 (preferably iso configuration)

C

13

-C

17 (preferably iso configuration)

C

15 , C 17 Preferably the tunicamine is a tunicaminyl uracil toxin or a tunicamycin. The tunicamine can also be a corynetoxin capable of causing ARGT. For the purposes of the present invention the terms "spectrum" and "spectra" when used in relation to near infrared reflectance includes data across a span of RECEIVED TIME 22. FEE, 12:37 6 wavelengths e.g. every 2nm and data concerning one or more single wavelengths. In this regard, reflectance data at certain wavelengths in the NIR spectrum will be more useful for detecting tunicamines than others and thus measuring the reflectance at a predetermined number of wavelengths as opposed to measuring across the entire or large sections of the near infrared spectrum may be preferable. Preferably, the near infrared reflectance spectrum of the sample comprises at least one wavelength from the list of wavelengths comprising: 1102, 1148, 1226, 1342, 1390, 1450, 1508, 1580, 1650, 1680, 1836, 1920, 2060, 2090, 2110, 2160, 2224, 2270, 2354, 2368 and any of the preceding wavelengths +/- 1-12nm. Preferably, the near infrared reflectance spectrum of the sample comprises at least 2-5, 2-8, 2-13 or 2-19 of the aforementioned wavelengths. NIR spectra may be produced using available NIR spectroscopy instrumentation. Thus, the present invention also provides a method of detecting a tunicamine in a sample said method comprising the steps of (i) producing a NIR reflectance spectrum of the sample; and (ii) using said spectrum to detect the tunicamine . Preferably, the spectrum is used to detect the tunicamine by comparing it with a reference source or calibration such as a database of NIR spectra indicative of at least one characteristic of the tunicamine such as concentration or quantity. In this regard, development of a set of calibration samples and application of calibration techniques such as multivariate calibration techniques are important for NIR analytical methods. Preferably, the comparison involves a multivariate analysis such as partial least squares regression (PLSR), principal components analysis (PCA), principal components regression (PCR), discriminant analysis (DA) and artificial neural network (ANN). In addition to detecting its presence, the method of the present invention can also quantify the tunicamine in a sample. Thus the present invention also provides a RECEIVED TIME 22.EB. 12:37 7 method of detecting and quantifying the amount of a tunicamine in a sample said method comprising the step of using a NIR reflectance spectrum of the sample to detect and quantify the tunicamine. Preferably, the tunicamine is quantified in tenns of concentration or amount in the sample. The present invention seeks to offer a sensitve method for detecting and quantifying tunicamines. In this regard, prior to developing the subject invention, it was not possible to determine whether NIR could be used to reliably detect tunicamines with sufficient sensitivity to represent a viable detection method. Preferably, the method is capable of detecting and/or quantifying between at least 1-5ppb (1-5ug/kg), at least 4-5ppb (4-5ug/kg), at least 1-25ppb (1-25ug/kg), at least 5-20ppb (5-20ug/kg), 5 50ppb (5-50ug/kg), 5-100ppb (5-100pjg/kg) or 5-200ppb of the tunicamine. The sample may originate from a variety of sources. For example, the sample may be derived from a foodstuff such as animal feed (including premixed feeds or pellets), fodders, hays, straws, pastures, grains, screenings from grains or hays or food intended for human consumption. Preferably, the sample is derived from hays, grains or pastures. The sample may require pre-treatment, mathematical or otherwise, to ensure the NIR reflectance spectra yields useful results. Thus, the present invention may further comprise the step of pretreating the samples prior to producing the NIR reflectance spectra The sample may be a solid or liquid sample. When the sample is solid it may have a maximum particle size of 1, 2, 3, 4 or 5mm. Preferably the particle size will be uniformly 1, 2, 3, 4 or 5mm. When the sample is a liquid it may be in an extract such as water or an alcohol and includes ethanol and methanol extracts as well as aqueous ethanolic or methanolic extracts. RECEIVE[ TIME 22. FEB 12:U 8 Method of preparing a tunicamine calibration The present invention also provides a method of preparing a tunicamine calibration for use in NIR spectroscopy, the method comprising the steps: (i) collecting NIR spectra of a suitable number of samples with a known tunicamine value; and (ii) using the spectra collected in (i) to prepare the tunicamine calibration. The suitable number of samples may be varied and is preferably at least 50 but may comprise 100, 200, 500 or 1000. The tunicamine value can be any useful value such as quantity, amount, concentration or type/class of tunicamine. Furthermore, it can be predetermined using any available analytical technique. Preferably, the spectra collected are used to prepare the tunicamine calibration by applying multiple linear regression (MLR), PLSR or PCR to prepare a calibration model. Data concerning the spectra may be pretreated using mathematical transformation to enhance one or more features and/or remove or reduce one or more sources of unwanted sources of variation to create a more robust calibration model. The calibration model/set may be tested using a validation set comprising samples, with known tunicamine values not included in the calibration set, or by a cross validation procedure. Thus the present invention also provides a method of testing a tunicamine calibration for use in NIR spectroscopy, the method comprising the steps: (i) collecting NIR spectra of a suitable number of validation samples with a known tunicamine value; and (ii) comparing the spectra collected in (i) with the tunicamine calibration. RECEIVED TIME 22. FEB. 12:37 9 Systems for remote processing and reporting NIR reflectance information The method of the present invention can facilitate the remote processing and reporting of NIR reflectance information. Thus, the present invention also provides a system comprising: (i) means for electronically receiving a request to detect and/or quantify the amount of at least one tunicamine in a sample, wherein the request includes a NIR reflectance spectrum of the sample; (ii) means for comparing the spectrum to a database that correlates the presence and/or amount of the tunicamine to known NIR reflectance spectra; (iii) means for detecting and/or calculating the amount of the tunicamine; and (iv) means for electronically reporting the results. Preferably, the request is received over the internet and the means for electronically reporting the results also utilises the internet. In one form of the invention the request is submitted to a website and the results are report via a website. In another form at least one of the request and the results are exchanged via electronic mail. EXAMPLE S Example 1 - Detecting tunicamycin using NIRS Materials/Methods (a) Tunicamycin The tunicamycin was from Streptomyces sp. CAS-No 11089-65-9, Product code T7765, from Sigma-Aldrich Pty. Ltd. 12 Anella Ave, Castle Hill, NSW 2154 and was dissolved in 70% methanol (aq) for use in the Example. (b) Sample Preparation RECEIVED TIME 22. FEB. 12:37 10 Known amounts of the solution of tunicarnycin were added to 2 0g hay samples, ground to pass a 1mm mesh, prior to NIRS analysis. (c) NIR machine and operating parameters The ground hay samples were contained in sealed polyethylene bags and analysed using a Unity Scientific SpectraStar, a stand-alone bench-top near-infrared (NIR) diffuse reflectance spectrometer designed for rapid, non-destructive, multi-constituent analysis of a wide range of products. The sample bags were placed directly onto the viewing window for scanning. The NIRS spectrophotometer was from Unity Scientific, LLC, 10240 Old Columbia Rd, Columbia MD 21046. The data were analysed using 'UnScrambler @ X' (Camo Software AS). Partial least squares regressions (PLSR) were fitted to the spectra. Calibration and prediction (validation) regressions equations were derived using the full data set with frill cross validation of the calibration set to obtain a prediction regression. Results An example of the results of the NIRS analysis of the spectra are presented graphically in Figure 1. Estimates of the goodness of fit of the PLSR models of the data are provided by a number of statistics including R2 (square of the correlation between the reference value and either the calibration value or the prediction (validation) value), RMSEC (Root Mean Square En-or of Calibration) and RMSEP (Root Mean Square En-or of Prediction (Validation)) and. These are summarised in Tables 1, 2 and 3 below. These indicate that a tunicamine can be detected reliably in a solid matrix by NIRS and appropriate multivariate analysis of the NIRS spectra. RECEIVED TIME 22. EB. 12:37 11 Table 1. Amount of tunicamycin predicted by NIRS in oaten hays Amounts of Correlation for tunicamycin (pg) prediction forSEC RMSEP added to hays ditioncalibration (RC2 (reference values) (validation) (R 2 0 -0.5 0.82 0.94 0.06 0.08 0 - 5.5 0.84 0.86 0.57 0.62 0-30 0.97 0.98 0.82 1.04 Table 2. Error expected at low and high concentrations of tunicamycin ~2*0.08, 50 0 -0.5 0.06 0.08 0.16 8 or 0.16 2*0.62, 75 0 -5.5 0.57 0.62 1.24 62 or 1.24 2*1.04, 105 0-30 0.82 1.04 2.08 104 or 2.08 RECEIVED TIME 22. FEB. 12:37 12 Table 3. Relative error expected across a range of tunicamycin concentrations Amounts of tunicamycin Concentration of (pg) added to hays tunicamycin (pg/20g hay) Relative error (%) (calibration range) in a prediction. 0-5.5 2.2± 0,62 0.28 0 -5.5 5.5 ± 0.62 0.11 0-30 2.2 ±1.04 0.47 0-30 30± 1.04 0.03 Example 2 - NIRS detection of tunicamines Materials/Methods Three tunicamine solutions and mixtures thereof were analysed by near infra-red spectroscopy (NIRS): (i) tunicamycin (Sigma T7765); (ii) corynetoxins extract; and (iii) 50:50 mixture of (i) and (ii). The corynetoxin solution was prepared by soaking grain screenings which had been ground to pass a 1mm screen in 70% methanol (aq.) for seven days. The ratio of ground screenings (air dry weight) and 70% methanol (aq.) was 0.17. The screenings had previously been analysed to contain 34mg corynetoxins/kg DM confinedd by high pressure liquid chromatography (HPLC) and enzyme-linked immunosorbent assay (EIA). Solutions of (i), (ii) and (iii-) were made in 70% methanol (aq.) such that amounts of 250p.L contained 0.052, 0.105, 0.525 or 1.050ug of the tunicamines. With 250pL of these solutions added to lOg ground hay this would yield 5.2, 10.5, 52.5 and 105.0pg tunicamines/kg hay. The tunicamine solutions were scanned in triplicate using a FOSS XDS NIRS machine and the spectra are reported as 1/log reflectance from 1100 to 2500nm at 2nm intervals. RECEIVED TIME 22. FEB. 12:37 13 The NIRS spectrophotometer was from FOSS NIRSystems Inc., Macquarie View Park, Unit 2 112-118 Talavera Rd, North Ryde NSW 2113, Australia. The data were analysed using 'UnScrambler @ X' (Cano Softvare AS). The solutions were scanned in polystyrene vials. Correction for scatter was made using the Extended Multiplicative Scatter Correction (EMSC) algorithm in 'UnScrambler' because the polystyrene in the vials also is detectable in NIRS spectra. Partial least squares regressions (PLSR) were fitted to the corrected spectra. Calibration and prediction (validation) regression equations were derived using the full data set with full cross validation. Twenty wavelengths for a smaller prediction model were selected as important predictors from an analysis of the regression coefficients of the entire spectrum of 700 wavelengths (1100 to 2500nm at 2nm intervals), to detennine in which parts of the spectrum the regression coefficients were different from zero, A regression coefficient of zero adds nothing to prediction. The regression coefficients that were significantly different from zero were considered to represent the wavelength regions that are important for predicting Y, the concentration or amount of tunicamine. These regression coefficients were significantly different from zero at P<0.05, as judged by a jacknife procedure in UnScrambler® to determine the variance of each regression coefficient and thus the confidence interval around each regression coefficient for each wavelength in the prediction equation. Results The wavelengths in Table 4 were identified as most useful for identifying the compounds of interest. RECEIVED TIME 22J EB, 12:31 14 Table 4 1102 1148 1226 1342 1390 1450 1508 1580 1650 1680 1836 1920 2060 2090 2110 2160 2224 2270 2354 2368 The PLSR calibration and prediction parameters at the 20 wavelengths listed in Table 4 are detailed in Table 5 and graphically illustrated in Figures 2A-2D. The data in those Figures are expressed as [tg/250gL solution. RECEIVED TIME 22. FEB. 12:37 15 Table 5 PCalibratronsion2Number of PLRS regression R 2 RMSE Intercept Slope sames on prediction samples (n) Corynetoxins extract, tunicamycin and corynetoxins extract plus tunicamycin 50:50 Calibration 0.97 0,077 0.01 0.97 37 mixture (total tunicamines) (Figure 2A Corynetoxins extract, tunicamycin and corynetoxins extract plus Prediction tunicamycin 50:50 (validation) 0. 92 0,115 0.03 0.094 37 mixture (total tunicamines) (Figure 2A) Tunicamycin plus corynetoxins extract (50:50 mixture) Calibration 0.99 0.030 0.00 0.99 12 (Figure 2B) Tunicamycin plus corynetoxin extract Prediction (50:50 mixture) (validation) 0.98 0062 -0,003 0.99 12 (Figure 2B) Tunicanycin Calibration '.00 0,012 0.00 1 00 12 (Figure 2C) Tunicamycin Prediction 0.98 0.058 0 00 0 96 12 (Figure 2C) (validation) Corynetoxins extract Calibration 0.99 0.029 0.00 0.99 12 (Figure 2D) Corynetoxins Prediction extract 0.97 0.075 0.01 1,00 12 (Figure 2D) The results demonstrate that near infrared spectroscopy can be used to accurately and quantitatively measure a range of tunicamines with high sensitivity, A common set of wavelengths can be used to predict a range of concentrations of tunicarmines as is shown Table 5 and Figures 2A to 2D. RECEIVED TIME 22. FEB. 12:37 16 Example 3 N1RS detection of tunicamines in hay Materials/Methods Varying amounts (0.105, 0.525 or 1.050tig) of the tunicamines from Example 2 in 205pL of 80% ethanol (aq.) were added to 10.Og samples of oaten hay that had been tested and confirmed as R. toxicus free, prior to use. The hays were ground to pass a 1mm mesh. The amounts of tunicamines added were equivalent to final concentrations of tunicamines of 10.05, 52.50 and 105.0 ttg/kg hay. The hay sample "spiked" to give an equivalent of 105.0pg/kg hay was assayed in triplicate in an EIA specific for corynetoxins. The "spiked" hay samples were placed in borosilicate glass specimen vials (Wheaton, cat no. 225466) and scanned using the same FOSS XDS NIRS machine as in Example 2. The data were analysed using 'UnScrambler @ X' (Camo Software AS). No corrections were made to the spectra, and regressions (PLSR) were fitted to them. The significant predictor wavelengths identified in Example 2 were used to predict the concentrations of tunicamines in the "spiked" hay samples. As well, calibration and prediction (validation) regressions were fitted to the spectra using PLSR and all 700 wavelengths of the spectra at 2nm intervals fi-om 1100 to 2500nm. The calibration regression was validated using a fill cross-validation. Results The results of the PLSR analysis using 20 significant predictor regression wavelengths (see Table 4) are summarised in Table 6 and the prediction equations are graphically illustrated in Figures 3A-3C. The results of PLSR analysis using all wavelengths at 2nm intervals between 1100 and 2500nm are summarised in Table 7. Comparison between the data in Table 6 and Table 7 indicate that prediction of the concentration of tunicamines relative to the reference concentrations is more precise and more sensitive using the 20 significant predictor wavelengths in the range 1100 to 2500nm than using all 700 wavelengths at 2nm intervals in the range 1100 to 2500nm. The "spiked" hay containing a calculated 105.0 tg corynetoxins/kg hay was assayed by an EIA specific for tunicamines to contain 98 pg corynetoxins/kg hay. The comparison with the NIRS analysis of the same sample is shown in Table 8. The two analyses yield the same result; the confidence intervals overlap (where the confidence interval for NIRS is RECEIVED TIME 22. FEB. 12:37 17 calculated as 2*RMSEP). The NIRS analysis is more accurate and precise than is the EIA. Table 6 Lowest concentration Number of detectable, not equal to zero Prediction equation R RMSEP Slope Intercept o bero da observations (n) (pig/kg) ( (2*RMSEP)+ intercept)) Corynetoxins extract 0.99 4210 0.99 0.66 9 9.08 (Figure 3A) Tunicamycin 1.00 0.520 1,00 0.01 9 1.05 (Figure 3B) Corynetoxins extract plus tunicamycin 1.00 0252 100 0.00 9 0.05 (Figure 3B) Table 7 PLSR regression Calibration or R 2 RMSEP Intercept Slope prediction Corynetoxins Prediction (valdation) 0.92 12.520 1.26 0.99 extract (pg/kg hay) Tunicamycin (pg/kg hay) Prediction (validation) 0,73 22.618 11.72 0.72 Corynetoxins extract plus tunicamycin (50:50 mixture) Prediction (validation) 0.89 14,343 13.52 0,77 (pg/kg hay) RECEIVED TIME 22. FEB. 12:37 18 Table 8 Reference Corynetoxins Coefficient of Confidence Method of determination of concentration of predicted (pg/kg variation between interval corynetoxins extract added corynetoxins hay) replicate samples to ground hays (= 2* standard (pg/kg hay) (average value) M deviation) EIA for corynetoxins for three 105 0 replicates of the same 96 10.3 77.8 to 118.9 sample NIRS using 20 wavelengths 105,0 for three replcates of the 103.9 11.4 92.5 to 115.2 same sample The results demonstrate that near infrared spectroscopy can be used to accurately measure a range of tunicamines in hay with high sensitivity. As would be apparent, various alterations and equivalent forms may be provided without departing fi-om the spirit and scope of the present invention. This includes modifications within the scope of the appended claims along with all modifications, alternative constructions and equivalents. In the present specification, the presence of particular features does not preclude the existence of further features. The words "comprising", "including" and "having" are to be construed in an inclusive rather than an exclusive sense. RECEIVED TIME 22. FEB, 12:37

Claims (31)

1. A method of detecting and quantifying a tunicamine in a sample from a feedstuff said method comprising the step of using a near infrared reflectance spectrum of the sample to detect and quantify the tunicamine.

2. A method according to claim 1 wherein the step of using a near infrared reflectance spectrum of the sample comprises using data concerning at least one wavelength from the list of wavelengths in Table 4 and any of the preceding wavelengths +/- 1- l2nm.

3, A method according to claim 2 wherein data concerning at least 2-5, 2-8, 2-13 or 2-19 of the wavelengths from the list of wavelengths is used.

4. A method according to any one of the preceding claims capable of detecting 1-5ppb (1-5ug/kg) of tunicamine.

5. A method of detecting and quantifying a tunicamine in a sample from a feedstuff said method comprising the steps of: (i) producing a near infrared reflectance spectrum of the sample; and (ii) using said spectrum to detect and quantify the tunicamine.

6. A method according to claim 5 wherein the step of producing a near infrared reflectance spectrum of the sample comprises using data concerning at least one wavelength from the list of wavelengths in Table 4 and any of the preceding wavelengths +/- 1- 12nm.

7. A method according to claim 6 wherein data concerning at least 2-5, 2-8 or 2 13 of the wavelengths from the list of wavelengths are used. RECEIVED TIME 22. FEB. 12:27 20

8. A method according to any one of the preceding claims wherein use of said spectrum comprises comparing said spectrum with a reference source to detect and quantify the tunicamine.

9. A method according to claim 8 wherein the reference source comprises a database of near infrared reflectance spectrum information indicative of at least one characteristic of the tunicamine.

10. A method according to claim 9, wherein the near infrared reflectance spectrum information indicative of at least one characteristic of the tunicamine is its concentration or quantity.

11. A method according to claim I wherein the concentration or amount of the tunicamine is quantified.

12. A method according to any one of the preceding claims wherein the sample is a solid sample,

13. A method according to claim 12 wherein the solid has a maximum particle size of 1-2mm.

14. A method according to any one of the preceding claims wherein the sample is a liquid.

15. A method according to claim 14 wherein the liquid is an extract.

16. A method according to claim 15 wherein the extract is an alcohol extract.

17. A method according to claim 16 wherein the alcohol is methanol.

18. A method according to claim 1 wherein the foodstuff is selected from the group comprising: an animal feed, fodder, pasture, grain, grass or hay or a product containing any one or more of the foregoing feedstuffs. RECEIVED TIME 22. FEB. 12: 37 21

19. A method of preparing a quantitative tunicamine calibration for use in NIR spectroscopy, the method comprising the steps: (i) collecting NIR spectra of a suitable number of samples from a foodstuff with a known tunicamine value; and (ii) using the spectra collected in (i) to prepare the quantitative tunicamine calibration.

20. A method according to claim 19 further comprising the step of using the spectra in step (i) to prepare a prediction (validation) equation.

21. A method according to claim 19 or 20 further comprising treating data in relation to NIR spectra using a mathematical transformation to enhance one or more features and/or remove or reduce one or more sources of unwanted sources of variation to create a more robust calibration model.

22. A method according to any one of claims 19 to 21 wherein the samples are hay samples.

23. A method of testing a quantitative tunicamine calibration for use in NIR spectroscopy, the method comprising the steps: (i) collecting NIR spectra of a suitable number of validation samples from a feedstuff with a known tunicamine value; and (ii) comparing the spectra collected in (i) with the quantitative tunicamine calibration.

24. A method according to claim 23 wherein the validation samples are hay samples.

25. A system comprising RECEIVED TIME 22 FEB 12:37 22 (i) means for electronically receiving a request to detect and quantify the amount of at least one tunicamine in a sample from a feedstuff, wherein the request includes a NIR reflectance spectrum of the sample; (ii) means for comparing the spectrum to a database that correlates the presence and amount of the tunicamine to known near infrared reflectance spectra; (iii) means for detecting and calculating the amount of the tunicamine; and (iv) means for electronically reporting the results.

26. A system according to claim 25, wherein at least one of the request and the results are exchanged over the internet.

27. A system according to claim 25 or 26, wherein at least one of the request and the results are exchanged via a website.

28. A system according to claim 25 or 26, wherein at least one of the request and the results are exchanged via electronic mail.

29. A system according to any one of claims 27 to 28 wherein the sample is a hay sample.

30. A report on the results from the system of any one of claims 25 to 29

31. A method for detecting and predicting the amount of at least one tunicamine in a sample, which comprises submitting a spectrum to the system in any one of claims 25 to 29 and receiving a report. RECEIVE) TIME 22. EB. 12:37