Abstract
Weeds are unwanted plant species growing in ordinary environment. In nature there are a total of 8000 weed species out of which 250 are important for agriculture world. The present study was carried out on weed species composition and distribution pattern with special reference to edaphic factor and farming practices in maize crop of District Mardan during the months of August and September, 2014. Quadrates methods were used to assess weed species distribution in relation to edaphic factor and farming practices. Phytosociological attributes such as frequency, relative frequency, density, relative density and Importance Values were measured by placing 9 quadrates (1 x 1 m(2)) randomly in each field. Initial results showed that the study area has 29 diverse weed species belonging to 27 genera and 15 families distributed in 585 quadrats. Presence and absence data sheet of 29 weed species and 65 fields were analyzed through PC-ORD version 5. Cluster and Two Way Cluster Analyses initiated four different weed communities with significant indicator species and with respect to underlying environmental variables using data attribute plots. Canonical Correspondence Analyses (CCA) of CANOCO software version 4.5 was used to assess the environmental gradients of weed species. It is concluded that among all the edaphic factors the strongest variables were higher concentration of potassium, organic matter and sandy nature of soil. CCA plots of both weed species and sampled fields based on questionnaire data concluded the farming practices such as application of fertilizers, irrigation and chemical spray were the main factors in determination of weed communities. (C) 2016 The Authors.