Abstract

This research work sought to explore dry storage and regulated dehumidification for
commercial seed storage in the humid tropics. Some of the specific objectives of the
study were to evaluate the potentials of containerized-dry storage for preservation of seed
quality and enhancement of seed longevity; and also to calibrate open seed warehouse
storage environments in the humid tropics and model seed deterioration under them. Five
maize genotypes obtained from IAR&T and Premier Seeds were used: TZPBSR,
DMRESRY, OBASUPER2, SUWAN1-SRY and ART-98/SW1. The seeds were cleaned
and their initial quality determined by recommended methods; after which three seed
storage experiments were set up, concurrently spanning for 24 months. Experiment1 was
carried out in 2005 and 2006; where 100g and 200g of seeds were packaged and stored in
containers containing different weights of silica gel to attain four and five different
gel/seed ratios by weight in each year respectively. Seeds were removed after four
months in storage, subjected to controlled deterioration test and evaluated for seed
moisture, viability and vigour attributes. PROC GLM statement of SAS was used to
separate means among treatments; and PROC PROBIT statements of SAS were used to
estimate seed deterioration and potential longevity variables for each treatment, using
serial germination data. Results provide evidence estimates that containerized storage at
1:1 and 1:2.5 gel/seed ratios maximized storage life of maize seeds. Experiment2 was
carried out at UNAAB and IAR&T seed stores from 2005-2007. About 1kg of seed lots
of the five maize genotypes was weighed and packaged in moisture-proof polythene bags
placed in the inner chamber of the seed stores. Seeds were taken at 3-month interval and
evaluated for seed viability and vigour attributes. PROC GLM statement of SAS was
 used to separate means of variables evaluated in both locations and PROC PROBIT
statements of SAS were used to estimate seed deterioration and potential longevity. The
seed survival pattern under the humid tropical conditions fitted the normal distribution
curve well, and the initial seed quality enhanced seed storability under the storage
locations. In Experiment3, two of the five maize genotypes (Suwan1-SRY and
Obasuper2) were used. The seeds were stored in 2L-capacity, air-tight plastic containers
conditioned to eight different RH levels (range 10 to 80%) and kept inside an incubator
set at 330C with mean inside-container temperature of 31±40C. Storage was carried out
between 2005 and 2006, during which seed moisture, viability and vigour attributes were
evaluated. PROC PROBIT statements of SAS were used to estimate seed deterioration
and potential longevity; while sorption data collected at interval were fitted into five
different sorption models (Henderson, Henderson-Thompson, Chung-Pfost, modified
Halsey and GAB). Maize seed storage at RH between 15.6 and 28.7% (RH conditioned
with Zinc Chloride and Calcium Chloride solutions respectively) kept seeds below 5.0%
SMC for minimum of 12 months. The Henderson-Thompson and Chung-Pfost models
used in modeling sorption data gave best fits for Oba Super2 and Suwan1-SRY seed lots
respectively; justifying the suitability of these models for evaluating moisture sorption
data of stored maize seed.