The A×S interactions on millet DM and N uptakewere due to response differences in pots containing2262.tex; 9/06/1999; 14:02; p.4
219Table 2.F ratios for treatment effects on cumulative millet DM, N and P uptake, and soil propertiesSource ofDMN uptakeP uptakeTotalSoilBray1-TotalHeavyN heavyP heavyN lightP lightvariation(g pot-1) (mg pot-1) (mg pot-1) soil-NNH4–NPSOMSOMSOMSOMSOMSOM(mg kg-1) (mg kg-1) (mg kg-1) (mg kg-1) (g pot-1) (mg pot-1) (mg pot-1) (g kg-1) (g kg-1)Species (S)19.318.104.22.168NSNSNS45.041.30.31.83.5Amendment (A)6.19.669.19.3NS24.1NS4.32841368.2146Fertilizer-N (N)37458538.018.1NSNS4.5NS30.8NSNSNSS×A22.214.171.124.7NSNSNS27.02.9NSNSNSS×N3.12.6NSNSNSNSNSNSNS5.6NSNSA×NNSNSNS5.0NSNSNSNS19.7NS19.5NSS×A×NNS2.5NSNSNSNSNSNSNSNSNS3.5MSE5.69602126.96.36.1990.200.232.011188.8.131.52.01Figure 1.Cumulative dry matter production by milletthe leaves of PG and VU. Whereas the feces derivedfrom all plant species, on average, produced similaramounts of DM and N uptake over the 4 crop cycles,pots containing the leaves of PG produced less DMand had lower N uptake than pots containing the leavesof VU (Figures 1 and 2).Total-P uptake by millet was most affected byamendment type followed by fertilizer-N (Table 2).The effect of amendment type on millet P uptake de-pended to some extent on plant specie (i.e. relativelylow F ratio for S×A interaction). On average, mil-let in pots containing feces accumulated about 37%2262.tex; 9/06/1999; 14:02; p.5
220Figure 2.Cumulative N uptake by milletmore P than those amended with leaves. Addition ofleaves generally reduced millet-P uptake relative to thecontrol (Figure 3). Millet grown in pots containing theleaves of GS, PE and PG had less (P <0.05) P uptakethan millet grown in control pots. Whereas additionof most leaves apparently immobilized P, addition offeces derived from PE, GS and VU, increased (P <0.05) P uptake by millet.Relationships between amendment composition andyield and nutrient uptakeThe amounts of leaf DM and cell wall (NDF) addedto pots, and the NDF:N ratio of the leaves added topots accounted for about 60% (i.e. single correlationcoefficients squared) of the variation in cumulativemillet DM yield and N and P uptake (Table 3). Inpots that received fertilizer-N, the negative effects ofthese leaf additions were most pronounced during thefirst millet growing cycle, and diminished thereafter.The opposite response trend was observed in pots thatdid not receive fertilizer-N: the negative effects of leafDM, NDF and NDF:N application became more pro-nounced during the 2nd, 3rd and 4th millet growingcycles. Millet DM and N uptake was positively relatedto the total N concentration of leaves. Fractionation oftotal-N into soluble- and insoluble-N did not improveprediction of millet DM and N uptake over predic-tions using total-N as a dependent variable. Neitherleaf lignin, polyphenol, nor IPAC (or their ratios withN) proved to be related to millet response to eitherleaf or feces additions, in the presence or absence offertilizer-N.