Rapid Prediction Methods And Models For The Nutrient Contents Of Fattening Pig Manure

Volumes of manure from large scale livestock farm can cause pollution problems, and therefore an estimation of nutrients in manure is necessary with intensification of protecting environment consciousness and development of modern precision agriculture. This experiment studied three different rapid prediction methods and models for the nutrient contents of fattening pig manure.In order to test methods of rapidly estimating nutrient contents of pig manure based on physical and chemical analysis, 216 manure samples that contained 2.2496%~9.1827% dry matter were selected. One part of each manure sample was analyzed for physical and chemical properties, such as electrical conductivity (EC), specific gravity (SG) and pH; the other part of each manure sample was dried for laboratory analysis to measure nutrient contents. This paper studied the correlativity of physical and chemical properties and nutrient contents. Analytical results showed that: (1) EC could be evaluated for estimating ammonium nitrogen (R2=0.825) and total potassium (R2=0.737). (2) SG could be evaluated for estimating organic matter (R2=0.922), total nitrogen (R2=0.763), total phosphorus (R2=0.806), magnesium (R2=0.885) and zinc (R2=0.743). (3) SG and pH could be evaluated for estimating total nitrogen (R2=0.858) and available phosphorus (R2=0.749). (4) EC and SG could be evaluated for estimating total potassium (R2=0.810) and natrium (R2=0.665).In order to investigate the feasibility of near-infrared reflectance spectroscopy (NIRS) technology for rapidly analysis of the nutrient contents of pig manure, diverse pig manures (n=108) were collected from pig farms in Beijing. Samples were scanned from 12000 to 4000 cm"1 in polyethylene bags on a NIRS System. This study developed calibrations using partial least-squares regression. Analytical results showed that: (1) Organic matter, total nitrogen, total potassium, zinc and moisture content of fatting pig manure had good predictability by NIRS (R2>0.95, RPD>3). (2) Total phosphorus, available phosphorus, calcium and magnesium could be predictable by NIRS (R2>0.85, RPD>2.5). (3) The feasibility of predicting ammonium nitrogen content deserved further research.In order to study methods of rapidly estimating nutrient contents of pig manure based on nutrients intake, six growing pigs were subjected to five diets containing varied quantities of crude protein, phosphorus and Potassium. 30 feces and 30 urine samples were collected. Nutrients intake and nutrient content of pig manure were measured, and the correlativity was established. Analytical results showed that: (1) Nitrogen intake and weight could be evaluated for estimating total nitrogen (TN) in pig feces (R2=0.655), and nitrogen intake could be evaluated for estimating TN in pig urine (R2=0.657) and TN in pig manure (R2=0.825). (2) Nitrogen intake and weight could be evaluated for estimating ammonium nitrogen (AN) in pig feces (R2=0.551), and nitrogen intake could be evaluated for estimating AN in pig urine (R2=0.575) and AN in pig manure (R2=0.586). (3) Phosphorus intake and weight could be evaluated for estimating total phosphorus (TP) in pig feces (R2=0.706) and TP in pig manure (R2=0.846). (4) Potassium intake could be evaluated for estimating total potassium (TK) in pig feces (R2=0.455), TK in pig urine (R2=0.692) and TK in pig manure (R2=0.743). (5) Nitrogen intake could be evaluated for estimating feces (R2=0.478).