Pak. J. Weed Sci. Res. 12(1-2): 79-82, 2006

STUDIES ON SHAPE, SIZE AND WEIGHT OF CERTAIN WEED SEEDS BURIED IN THE SOIL SEED BANK

Allah Bakhsh1, Altaf A. Dasti2, Asia Munir1, Imran Khaliq1,

Muhammad Amin ud Din1 and Muhammad Shaheen Akhtar3

 

ABSTRACT

In agriculture the identification and classification of weed seeds are technically and economically important. In this work, seeds of 15 different weeds species were collected from different crop fields located in the periphery of Dera Ghazi Khan City, Pakistan. The size and weight of the collected seeds were recorded. These parameters may serve as a convenient method for identification and classification of weeds on the basis of their availability in the soil seed bank available in the soil.

Key words: Seed identification, Machine vision, seed bank
 
INTRODUCTION

In agriculture, the identification and classification of weeds is technically and economically important because these would be helpful to identify the seed bank in soil of expected competitor weeds. The standard ocular identification is mostly based on the knowledge of seed size, shape, colour and texture. Machine vision seems a suitable technique to automate this task (Granitto et al. 2000). In addition these characteristics are important to know the quality of seeds (Jansen, 1995; Ahmad et al. 1999).

In most of the studies, identification on bases of these morphological characters and machine vision technique was concentrated on cultivated varieties (Chen et al. 1989; Draper et al.1984; Neuman et al.1987; 1989; Symons et al. 1988).  However, few studies are available in case of weeds (Chtioui et al. 1996; Petersen et al. 1992 and Granitto et al. 2003).[1]

In this paper we are firstly reporting the size, weight, color and shape of weed seeds in Pakistan. This study will be beneficial in future for weed identification, classification and in their management.

MATERIALS AND METHODS

             During February to May 2005, five hundred normal and healthy partially buried seeds of each species of following 15 annual weeds i.e. Anagallis arvensis Linn.,   Avena fatua Linn., Brassica campestris Linn., Chenopodium album Linn., Euphorbia helioscopia Linn., Fumaria indica Hassk., Galium aparine Linn., Ipomoea eriocarpa R.Br., Lathyrus aphaca Linn., Lolium temulentum Linn., Melilotus albus Desr., Medicago denticulata Willd., Phalaris minor Retz., Rumex obustifolious Linn and , Vicia sativa Linn., were collected from the surrounding soil of respective weeds , growing in  different crop fields located in the periphery of Dera Ghazi Khan City.

             The 500 seeds of each weed species were randomly divided into five replicates. Each replicate was weighed with the help of Sartorious Electric Balance (Germany). Similarly out of 500 seeds; ten seeds of each species of weed were randomly selected for measuring the size. The seed size (diameter) was measured with the help of size measuring apparatus. For seed images or shapes, Software Computing Software Package was used.

Statistically, mean values with standard deviations of each weed was computed by using the SPSS Software package, 2003 Version-13.0.

Figure 1. Steel Sieve with holes of different sizes (in mm) in linear order.

 

RESULTS AND DISCUSSION

        In this study, the characteristics like size, weight and shape for identification were used as suggested by various workers in case of cultivated varieties (Chen et al. 1989; Draper et al.1984; Neuman et al. 1987, 1989 and Symons et al. 1988) The average size per seed and weight per 100 seeds with standard deviations of 15 weeds are presented in Table.1. The photographs to show details of seeds of each weed species are also presented in Figure 2.

We found that the single character is not enough to distinguish the species because the seeds of more than one species posses same mean value, however, their standard deviations vary. But the consideration of these three characteristics collectively was found unique in this study.

So these characteristics may serve as a convenient method for identification and classification of weeds on the basis of their seed bank available in the soil.

 

Table-1.Average Size per Seed and Weight per 100 Seeds of 15 Weed Species.

Species

Size

Weight

Species

Size

Weight

Anagallis arvensis

1.6±0.223

0.0384 ±0.004

Lathyrus aphaca

3.5±0.353

1.035±0.054

Avena fatua

2.4±0.707

1.295 ±0.025

Lolium temulentum

2.5±0.223

0.835±0.022

Brassica campestris

2.5±0.418

0.170 ±0.001

Medicago denticulata

2.5±0.418

0.336±0.006

Chenopodium album

1.5

0.0523 ±0.003

Melilotus albus

1.5

0.160±0.008

Euphorbia helioscopia

2.2±0.274

0.194±0.003

Phalaris minor

1.00

0.178±0.005

Fumaria indica

1.5

0.279±0.015

Rumex obustifolous

1.7±0.274

0.113±0.001

Galium aparine

2.1±0.223

0.716±0.036

Vicia faba

4.1±0.223

2.349±0.111

Ipomoea eriocarpa

3.2±0.273

1.301±0.019

 

 

 

 

Figure 2. Photographs of Seeds belonging to 15 different annual weeds

REFERENCES CITED

Ahmad, I.S., J.F. Reid., M.R. Paulsen and J.B. Sinclair. 1999. Color classifier for Symptomatic Soyabean seeds using image processing. Plant disease 83:320 327.

Chen, C., P. Chiang and Y. Pomeranz.1989. Image analysis and characterization of cereal grains with a laser range finder and camera contour extractor. Cereal Chem. 66 (6):466-470

Chtioui,Y., D. Bertrand., Y. Dattee and M. F. Devaux. 1996. Identification of seeds by color imaging: Comparison of discriminant analysis and artificial neural networks. J.Sci.Food Agric. 71:433-441

Draper, S. R. and A. J. Travis 1984. Preliminary observations with a computer based system for analysis of the shapes of seeds and vegetative structures. J. Nat. Inst. Agri. Bot.16:387-395.

Granitto, P. M., P. F. Verdes.,  H.D.Navone and H.A. Ceccato. 2000. Automatic Identification of Weed seeds by color image processing. V1- Argentine Congress on Computer Science (Ushuaia Argentina).

Granitto,P. M., P. A. Garralda, P.F. Verdes and H.A., Ceccato. 2003. Boosting Classifiers for Weed Seeds
                 Identification. JCS&T .3(1).

Jensen,P. I. 1995. Seed production quality in Trifolium balansae and T.resupinatum: The role of Seed color. Seed Sci.Technol. 23:353-364.

Neuman,M. R., H.D. Sapirstein E. Shwedyk and W. Bushuk.1989. Wheat grain color analysis by digital image processing 1. Methodology. J. Cereal Sci.10:175-182.

Neuman,M. R., H.D. Sapirstein., E. Shwedyk and W.Bushuk. 1987.Discrimination of wheat class and variety by digital image analysis of whole grain samples. J.Cereal Sci.6:125-132.

Peterson, P.E.H. and G.W.Krutz. 1992. Automatic identification of weed seeds by color machine vision. Seed Sci.Technol.20:193-208.

Symons, S.J. and R.G.Fulcher. 1988. Determination of wheat kernel morphological variation by digital image analysis: 1.Variation in Eastern Canadian Milling Quality Wheat. J.Cereal Sci.8:211-218.

[1]. Bio. Sciences Dept. Govt College, D G Khan, Pakistan.

2. Institute of Pure & Applied Biology. BZU, Multan,Pakistan.3. Bio. Dept, F.G. Degree College, H/9, Islamabad, Pakistan.
    *Correspondence: saqlan@uaar.edu.pk

 

Copyrights© Weed Science Society of Pakistan®

Department of Weed Science
NWFP Agricultural University Peshawar, 25130 Pakistan