Data Acquisition for Monitoring Vapor Pressure Deficit in a Tropical Lowland Shelter-house Plant Production

The objective of this study was to monitor air Vapor Pressure Deficit (VPD) in a tropical lowland shelterhouse plant production. A custom-designed real-time Data Acquisition (DAQ) system with three independent microcontroller boards and sensors for monitoring aerial parameters was developed, calibrated and tested. Sample temperature and Relative Humidity (RH) data for VPD calculations were continuously collected every 60 sec, for 6 days, inside a 40 m 2 shelter-house located at the Universiti Putra Malaysia agricultural experimental field. Preliminary results showed that VPD values varied from 0.16 to 2.51 kPa, with a mean of 0.83 kPa and standard deviation (Std) of 0.6 kPa. Different regression models were used to describe the nonlinear correlation that existed between temperature and VPD data. Results showed that squared polynomial model produced the maximum coefficient of determination (R 2 ) equal to 0.976. This model was successfully used for VPD prediction based on temperature inputs. The hypotheses that collected data follow normal distribution and have different means in the 6 days of experiment were rejected at any significant level. The result of this study can be used in decision support systems’ database for controlling tropical lowland plant production environments.


INTRODUCTION
Controlled Environmental Plant Production Systems (CEPPS) with fully covered transparent material are widely used in Tropical Lowland (TL) regions; however the excess heat imposed by the direct solar radiation cause substantial amount of increase in the inside temperature and ambient Relative Humidity (RH).Ajwang and Tantau (2005) reported that under humid tropical climatic conditions, increase in temperature can rise to 38°C.In order to take into account both temperature and (RH) measurements, calculating the difference between the actual air moisture and its saturated point, denoted as Vapour Pressure Deficit (VPD), is of interest.Large values of VPD, (approximately higher than 2 kPa), cause high transpiration rates with significantly increases in Evapo Transpiration (ET) demands (El-Sharkawy et al., 1986) and stomatal closure (Pettigrew et al., 1990).Plants transpiration may be suppressed at low VPD which impacts the energy balance of the canopy.According to Pregner and Ling (2000) AEX-804, fungal pathogens and mineral deficiency symptoms appear below VPD value of 0.43 kPa and disease infection can be most damaging below VPD value of 0.2 kPa.It is therefore recommended that the VPD of greenhouse air should be kept above 0.20 kPa.The Ideal range of VPD for most greenhouse crops is between 0.8 to 0.95 kPa (Argus Control Systems Ltd., 2009).Several methods for estimating vapor pressure deficits and RH have been analyzed in the study of Castellvi et al. (1996).
For tropical and subtropical conditions, the strategy in greenhouse climate management and its adaptation to the local environment situations are important in improving resource efficiency (Luo et al., 2005).Evaluating performance of ventilation and pad-and-fan systems based on ideal and acceptable VPD requirements for production of tomato (Shamshiri and Wan Ismail, 2012) showed that proper utilization of natural and mechanical ventilation can be a more effective method of providing acceptable growth conditions in a TL-CEPPS compared with pad-and-fan systems.A comparison between different greenhouse covering materials, including polyethylene film, photoselective red colored film and insect-proof net, for tomato cultivation during summer was carried out by Arcidiacono et al. (2001).It has been reported that by using insect-proof net covered greenhouse, the inside and outside temperature remain similar, while temperature was observed to be rising with the photoselective film during summer.In this regard, shelter houses that are covered with net insect screens have gained more popularity in TLPPS due to the likely potential of air temperature and RH that are naturally close to plants desired levels.These structures are generally used to reduce insect migration on the crop and subsequent crop damage (Teitel,20 of damaged by high rainfall, extreme solar radiation and high wind speeds. Microcontrollers and Personal Computers (PC) based data acquisition systems have been widely used in different specific plant production applications due to their lower development cost, operational flexibility and environmental adaptability (Thomas Boaventura et al., 1997;Mukaro andCarelse, 1999 Ameur et al., 2001;Koutroulis and Kalaitzakis, Rosiek and Batlles, 2008;Junxianga and Haiqing, 2011).The purpose of this project was to monitor air VPD in a TL shelter-house plant production by means of a custom-designed, low-cost accurate real acquisition system that continuously collects and stores temperature and RH data for VPD calculation.The results of this study will be used in the database of a decision support system for TL-CEPPS of tomato.

RESULTS AND DISCUSSION
The DAQ system was installed on a 40 m 2 TL shelter-house (Fig. 7) located at the Universiti Putra Malaysia agricultural experimental field (Latitude = 3°0'9.8094"and Longitude = 101°42'11.2926").The combined temperature and RH sensors were placed inside Okra (Abelmoschus esculentus) plants canopy (Fig. 8) that had been planted in three rows of total 15 polybags in the shelter-house.The VPD sensor was placed in the field, close to the canopy with air around the sensor homogenized by using a plastic shelter.Sample temperature and RH data for VPD calculations were continuously collected every 60 sec, for 6 days, from April, 15 to April 17 and April, 22 to April, 24, 2013.An average of the three readings was then calculated and plotted in Fig. 9 and 10 as the temperature and RH of the air inside plant canopy in the 6 days respectively.The reason for such a long term monitoring was to take into account the effects of solar radiation, climate variations and wind speed.VPD was calculated according to the internationally accepted formulation (Murray, 1967) for saturation vapor pressure that is most widely used by meteorologists and was plotted versus temperature in Fig. 11.The   support that the average daily temperature, RH and VPD values did not vary in these 6 days.In order to study the distribution of data, each sample has been shownby means of box plots (Velleman and Hoaglin, 1981) in Fig. 13.The hypothesis to be tested was that the collected temperature and RH samples follow normal distribution.To test this claim, density distribution and QQ plot was provided as shown in Fig. 14 to 17.It can be seen from plots of normal distribution fit in Fig. 14 and 15 and Q-Q plots in Fig. 16 and 17, that both temperature and relative humidity data have significant deviation from normality on the tails sides.The one-sample Kolmogorov-Smirnov test (Corder and Foreman, 2009) at the 5% significance level was also used to test the hypothesis that the collected temperature and RH data follow a normal distribution.Based on the results, it was concluded with 99% confidence that that neither collected temperature nor RH data follow normal distribution.

CONCLUSION
The high demands for quality agricultural product necessitates practicing various methods of modern technologies including automation and mechanization in different scopes of plant production system.The crop cultivation in TL environments by using the enhanced agricultural machine has not reached the optimal crop production as crop is still subjected to various stresses such as heavy rainfall, insects and extreme solar radiation.The high temperature and ambient RH are major issues in providing ideal environmental condition.This research study provided a better understanding of TL environments by addressing relatively long-term trends in temperature, RH and VPD.Major steps involved in this project included DAQ hardware/software interfacing, laboratory tests and calibrations, monitoring of temperature and RH from the actual shelter-house, VPD calculations and data analyzing.The designed DAQ system was successfully used for data collection and results showed that the average VPD is in the ideal range for plant growth.It can be concluded that perhaps the best way to improve performances of plant production systems in TL environments is to use the local information and the knowledge about crop itself.This is however a hypothesis to be tested, that by only applying appropriate natural and mechanical ventilation techniques, as in shelter houses, the probability of achieving successful production of a particular plant at its different growth stages would be higher that closed environment with evaporative cooling.

Fig. 12 :Fig. 14 :
Fig. 12: Plots of actual and predicted VPD in the 6 days of experiment Teitel, 2001)izchip Fig.2: BizChip PIC development board in TLPPS due to the likely potential of air temperature and RH that are naturally close to plants desired levels.These structures are generally used to reduce insect migration on the cropTeitel, 2001), reduce risk of damaged by high rainfall, extreme solar radiation

Table 1 :
Descriptive statistics summary of the collected data in the 6 days of experiment

Table 4 :
Vapor pressure deficit data statistics in each day of experiment Ajwang, P.O. and H.J. Tantau, 2005.Prediction of the effect of insect proof screens climate in naturally ventilated greenhouse in humid tropical climates.Acta Hortic., 691: 449-456.Ameur, S., M. Laghrouche and A. Adane, 2001.Monitoring a greenhouse using a microcontrollerbased meteorological data acquisition system.