Growth and Performance of Guar (Cyamopsis tetragonoloba (L.) Taub.) Genotypes under Various Irrigation Regimes with and without Biogenic Silica Amendment in Arid Southwest US

Guar is a potential crop that can be grown as a forage or as a seed crop in arid to semi-arid regions due to its low water requirements and tolerance to heat. Optimizing irrigation water use is important for making alternative crops such as guar a sustainable option. Amendments such as biogenic silica, a sedimentary rock from a biogenic source such as fossils, may help plants tolerate water stress due to reduced irrigation. The objective of the current study was to evaluate seed yield and attribute components and agronomic and physiological parameters for four guar genotypes (Matador, Kinman, Lewis, and NMSU 15-G1) under five drip irrigation regimes (I1-normal irrigation, I2-no irrigation at 75% pod formation, I3-no irrigation at 50% and 75% pod formation, I4-terminate irrigation at flowering, and I5-terminate irrigation at flowering + biogenic silica amendment) at Las Cruces in southern New Mexico, USA, from 2016 to 2018. On average, the I1 irrigation regime produced the highest guar seed yield (2715 kg ha−1) followed by I5 (2469 kg ha−1) from 2016 to 2018. As compared to the I1 regime, the I2 and I3 regimes resulted in a 20.8% and 23.4% decline in guar seed yield, respectively, on average from 2016 to 2018. The results suggest that the addition of biogenic silica might help to improve guar seed yield under reduced irrigation conditions and can produce comparable yields with an average of 300 mm of irrigation during the growing season in the southern New Mexico region of the Southwest US.


Introduction
Water scarcity is becoming an important issue worldwide, including in the arid and semi-arid environments of the Southwest US. Agriculture in major parts of the Southwest US, including New Mexico, depends on the water extracted from the Ogallala aquifer (the primary aquifer in the south-central US) for irrigation. For instance, in New Mexico, a total of 3740 thousand acre-feet of fresh water are extracted from the Ogallala aquifer per year [1]. High-volume withdrawals from the Ogallala aquifer, combined with low rainfall, have led to decreasing groundwater levels in regions including New Mexico [2,3].
Poor, sandy soils with low organic matter in the region make it difficult for growers to sustain crop production. High salt buildup is another issue facing agriculture in arid and semi-arid environments, including the desert in the Southwest US. Including alternative crops that can be successfully grown in limited water and high salt environments is an important strategy to sustain agriculture in such regions, including the Southwest US.
Guar (Cyamopsis tetragonoloba L.), also referred to as clusterbean, is a drought tolerant crop that is typically grown on sandy loam soils under hot and dry conditions [4]. Guar There is no information available in the literature about the role of silica on guar growth under limited water conditions. We hypothesize that the addition of biogenic silica as an amendment would help guar plants improve their growth and yield under reduced irrigation.
The objective of the current study was to evaluate selected guar genotypes for their growth, physiological parameters, seed yield, and yield attributes under reduced irrigation with and without the addition of biogenic silica amendment in the southern New Mexico region of the Southwest US.

Growth and Biomass
In general, the growth and biomass of guar plants were affected by irrigation regimes, with slightly varying results in different years. At maturity, in 2018, plant height under I 1 was higher than plant height under any other irrigation regime. Additionally, I 1 had a higher plant height than I 2 in 2016 and higher than I 4 in 2017 (Table 1). In 2016, I 5 had a higher plant height than I 4 , I 3 , or I 2 . Among genotypes, Lewis and NMSU-15-G1 had significantly higher plant heights compared to Kinman in 2016 (Table 1). Table 1. SPAD leaf chlorophyll content at 50% and 100% pod formation stages and plant height and above-ground dry biomass (AGDB) at maturity stages of four guar genotypes under five irrigation regimes from 2016 to 2018, Las Cruces, NM.  Means within a column and particular effects not including letters or sharing a letter do not differ at α < 0.05 (Fisher's F-protected LSD). I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50%, 75%, and 100% pod formation stages at 57, 72, and 100 days after planting, respectively; maturity stage at 120 days after planting. NS = Not significant at α < 0.05.
Above ground biomass measurements at maturity also varied among the irrigation regimes. Among irrigation regimes, the numerically highest above ground dry biomass at maturity was recorded under I 1 (2017 and 2018) and I 5 (2016) ( Table 1), although I 1 and I 5 did not differ significantly in any year. I 5 had higher above-ground dry biomass than I 4 in 2016 and 2017. Among genotypes, in 2016, above-ground dry biomass was higher for NMSU-15-G1 compared to Kinman and Matador (Table 1).

Physiological Parameters
Varying results were observed in the physiological parameters among different irrigation regimes when measured at different stages of plant growth during the study. For instance, when measured at 50% pod formation, the SPAD chlorophyll values did not differ among irrigation regimes in 2016 (Table 1). In 2017, SPAD chlorophyll values at 50% pod formation were higher for I 1 compared to I 3 and I 4 regimes (Table 1). In 2018, I 5 had the highest SPAD value at the 50% pod formation, while I 2 had a lower SPAD than I 3 , I 4 , or I 5 . The I 5 regime resulted in numerically higher SPAD chlorophyll values than I 4 at 50% pod formation from 2016 to 2018, with the difference significant only in 2018 (Table 1).
Among genotypes, during 50% pod formation in 2016, the highest SPAD value was recorded under Matador, which was significantly higher than Kinman and NMSU-15-G1 (Table 1). In 2018, Lewis had significantly higher SPAD values compared to Matador and NMSU-15-G1 (Table 1).
At 100% pod formation, among irrigation regimes, I 5 had higher (2016 and 2017) or similar (2018) SPAD values as compared to I 1 (Table 1) and had higher SPAD than I 4 in all three years.
Physiological parameters, including photosynthetic rate, stomatal conductance, and transpiration rate, also showed varying results among the irrigation regimes. For instance, at the 75% pod formation stage in 2018, I 1 and I 2 had a higher photosynthetic rate than I 3 and I 4 , with I 5 having an intermediate value that did not differ significantly from any other irrigation regime ( Table 2). Among genotypes, the lowest photosynthetic rate was observed for Kinman compared to other genotypes in 2016 ( Table 2). Means within a column and particular effects not including letters or sharing a letter do not differ at α < 0.05 (Fisher's F-protected LSD). I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50% and 75% pod formation stages at 57 and 72 days after planting, respectively. NS = Not significant at α < 0.05.
Results for stomatal conductance indicated significant differences among irrigation regimes only in 2018, with the I 1 regime resulting in higher stomatal conductance compared to other irrigation regimes except for I 2 ( Table 2). In 2018, at the 75% pod formation stage, the transpiration rate was higher under the I 1 regime than the I 3 or I 4 regimes ( Table 2). Among genotypes, Kinman was lower compared to other genotypes only for 2016 (Table 2).
In the current study, biogenic silica seemed to have a positive effect on guar physiological parameters. In 2018, SPAD under I 5 was higher than under I 4 at the 50% pod formation stage in 2018 and in all three years at the 100% pod formation stage. At 75% pod formation in 2018, I 4 had a lower photosynthetic rate than I 1 , while I 1 and I 5 did not differ significantly.

Water Use Efficiency
Intrinsic water use efficiency (WUE i ) did not differ among irrigation regimes and genotypes from 2016 to 2018 (Table 3). Instantaneous water use efficiency (WUE inst ) was greater under the I 5 regime in 2017 compared to the other irrigation regimes except for I 4 ( Table 3). Results could possibly be attributed to the capacity of guar to withstand drought.
Means within a column and particular effects not including letters or sharing a letter do not differ at α < 0.05 (Fisher's F-protected LSD). I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50%, 75%, and 100% pod formation stages at 57, 72, and 100 days after planting, respectively; maturity stage at 120 days after planting. NS = Not significant at α < 0.05.
Agronomic water use efficiency was significantly higher under I 5 compared to all other irrigation regimes in 2016 ( Table 3). The I 5 regime also had a significantly higher WUE y compared to the I 4 regime in 2018 (Table 3). In 2016, the genotype NMSU-15-G1 had significantly lower WUE y compared to other genotypes at maturity (Table 3).

Yield Attributing Characteristics and Seed Yield
While significance patterns varied for different yield attributing characteristics and years when recorded under different irrigation regimes and genotypes during the study, I 1 and I 5 had numerically higher values than other regimes for clusters per plant, pods per plant, seeds per plant, and seed yield in all years. For instance, a higher number of clusters per plant was observed under I 5 as compared to other irrigation regimes in 2016, while no significant differences were observed among irrigation regimes in 2017 and 2018 (Table 4). Among genotypes, Kinman and NMSU-15-G1 had a significantly higher number of clusters per plant as compared to Matador in 2016, while clusters per plant did not differ among genotypes in 2017 and 2018 (Table 4).
In 2016, the number of pods per plant under I 5 was higher than under I 2 , I 3 , or I 4 , while in 2018, I 1 had a higher number of pods per plant than I 2 , I 3 , or I 4 ( Table 4). In 2017, I 3 had a lower number of pods per plant than I 1 and I 5 ( Table 4).
The number of seeds per pod did not differ among irrigation regimes from 2016 to 2018 (Table 4). Among genotypes, Kinman produced the lowest number of seeds per pod, while Matador produced the highest number of seeds per pod from 2016 to 2018 (Table 4).
The total number of seeds per plant was higher under I 5 compared to I 4 and I 3 irrigation regimes in 2016 and 2017, while differences were not significant in 2018 (Table 4). The total number of seeds per plant did not differ among genotypes from 2016 to 2018 (Table 4). Means within a column and particular effects not including letters or sharing a letter do not differ at α < 0.05 (Fisher's F-protected LSD). I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50% and 75% pod formation stages at 57 and 72 days after planting, respectively. NS = Not significant at α < 0.05.
The 1000-seed weight did not differ among irrigation regimes in 2016 and 2018 (Table 5). In 2017, I 5 had a higher 1000-seed weight than I 2 , I 3 , or I 4 ( Table 5). Among genotypes, NMSU-15-G1 had a significantly higher 1000-seed weight compared to other genotypes in 2016 and 2017 ( Table 5). The 1000-seed weight under Matador was significantly lower compared to other genotypes in 2018 (Table 5). Means within a column and particular effects not including letters or sharing a letter do not differ at α < 0.05 (Fisher's F-protected LSD). I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50% and 75% pod formation stages at 57 and 72 days after planting, respectively. NS = Not significant at α < 0.05.
The harvest index (HI) was lower under I 5 compared to other irrigation regimes in 2016 (Table 5). In 2017, however, the I 5 and I 4 regimes recorded higher HI compared to I 1 and I 2 ( Table 5). Among genotypes, the highest HI was recorded under Kinman compared to Lewis and NMSU-15-G1 in 2017 and 2018 (Table 5).
Guar seed yields varied among the irrigation regimes (Table 5). In general, higher levels of guar seed yield were obtained under I 1 (Table 5) as compared to I 2 , I 3 , and I 4 ). Results were mixed for the I 1 -I 5 comparison, with I 5 having a numerically higher but not significantly different seed yield than I 1 in 2016 but a lower seed yield than I 1 in 2018. When compared to I 4 , the addition of biogenic silica in I 5 resulted in a significant increase in seed yield in 2016 (Table 4). Similar trends were observed in 2017 and 2018, with numerically (but not significantly) higher seed yields under I 5 than I 4 ( Table 5). Among genotypes, no significant differences were observed in their seed yield in 2017 and 2018 in Las Cruces. In 2016, NMSU-15-G1 resulted in a significantly lower seed yield as compared to other genotypes.

Experimental Site and Design Description
The study was conducted at the Fabian Garcia Science Center (32 •  The experiment was laid out in a split-plot design with five irrigation regimes as main plot factors and four guar genotypes as sub-plot factors. The four genotypes were randomly assigned to four sub-plots contained within each irrigation regime plot within each block, for a total of four replications (blocks). The four genotypes included three released varieties (Matador, Kinman, and Lewis) and one test line (NMSU-15-G1) that were selected based on best yield performance in previous studies at the same location [12].
The I 5 regime included a total of three applications of biogenic silica (SiO 2 ). The first application of biogenic silica was side-dressed at the trifoliate stage at a rate of 2500 ppm using a water pot. The second and third applications were applied as foliar sprays with a backpack sprayer at 50% pod formation and 75% pod formation using a rate of 2800 ppm.

Field Preparation
Raised beds of 1.02 m width were prepared and irrigated prior to planting. Guar seeds were manually planted at a depth of 2.5 cm in two rows on the raised beds. Hand thinning was conducted one week after emergence to maintain a plant population of 264,000 plants/ha. Weeds were removed manually using hand hoes. An insecticide was used to control leaf miners in 2016: MustangMax 279-3327 (FMC Agricultural Products, Philadelphia, PA, USA) at a rate of 25 g of active ingredient per ha in June, followed by a spray of Radiant 62719-545 (Dow AgroSciences, Indianapolis, IN, USA) in July at a rate of 40 g of active ingredient per ha.

Data Collection
Physiological measurements, including photosynthetic rate, stomatal conductance, and plant transpiration rate, were recorded on the youngest mature leaf during the 75% pod formation stage using a portable photosynthesis system (LI-COR 6400, Lincoln, NE, USA, 68504). The LI-COR 6400 system was calibrated, and the light source used was 6400-02b. Quantum flux photosynthetic active radiation (PAR) used was 1000 µmol photons m −2 s −1 and a CO 2 concentration of 400 ± 10 µmol mol −1 . All of these observations were taken on a clear, sunny day between 10:00 am and 1:30 pm.
Water use efficiencies (WUE) were calculated from data derived from the portable photosynthetic system. Such data included CO 2 assimilation (A N ), transpiration rate (E), and stomatal conductance (gs). Instantaneous water use efficiency was calculated as the ratio between A N and E (A N /E), while intrinsic water use efficiency was calculated as the ratio between A N and gs (A N /gs) [35]. Intrinsic water use efficiency (WUE i ) refers to the photosynthetic water use efficiency or carbon assimilation in comparison to the stomatal conductance [36]. Instantaneous water use efficiency (WUE inst ) is related to the comparison between the rate of photosynthesis and the rate of transpiration [37]. Agronomic water use efficiency (WUE) is defined as the amount of water used per season for the total seed yield during the season for the guar crop [38].
A Field Scout SPAD 502Plus chlorophyll meter (Spectrum Technologies, Inc., Aurora, IL, USA, 60504) was used to determine chlorophyll content in the leaves at 50% and 100% pod formation stages in 2016, 2017, and 2018.
Plant height, above-ground dry biomass, clusters per plant, seeds per pod, pods per plant, seeds per plant, 1000 seed weight (g), harvest index, and seed yield (kg ha −1 ) were recorded at maturity stage. Plant height was measured from the soil surface to the top of the plant using a meter scale. Above ground dry biomass weight was recorded after drying the samples at 55 • C for 72 h. Clusters per plant, seeds per pod, and pods per plant were counted on two randomly collected plants from each sub-plot. The total number of seeds per plant was calculated by multiplying seeds per pod by pods per plant. In 2016, for a 1000-seed weight, seeds were counted using a seed counter (Seedburo Equipment Co., Des Plaines, IL, USA, 60018). In 2017 and 2018, the seed counter used was the SLY-A automatic seed counter (Zhejiang Top Instrument Co., Ltd., Hangzhou, China, 311000).
For aboveground dry biomass measurements, guar plants were harvested from one-m 2 sections in each plot and then oven dried at 55 • C for 72 h, and dry weight was recorded. Seed weight was also recorded from these samples, and a harvest index was calculated by dividing the seed weight per m 2 by the aboveground dry biomass per m 2 . Pods were manually threshed to collect the seeds using a screen to avoid any seed losses. Finally, the threshed samples were cleaned using a clipper office tester (Clipper Separation Technologies, Bluffton, IN, USA, 46714).
The total seed weight from each sub-plot was also collected by manually harvesting the whole plants and collecting all the pods. The collected pods were then threshed using a Large Vogel plot thresher (LVPT, ALMACO Inc., Nevada, IA, USA, 50201).

Statistical Analysis
Data from the four replications were analyzed using a mixed model analysis for a split-plot design with whole plot factor irrigation regime and subplot factor genotype. The mixed model included fixed effects for irrigation regime, genotype, and their interaction. The random effects included block and the whole plot factor experimental unit (i.e., the block*irrigation regime interaction). Data were analyzed for each individual year using SAS PROC MIXED software version 9.4 (SAS Institute Inc., Cary, NC, USA, 2016). When the F-test was found to be significant (p = 0.05), a pairwise means separation test was conducted using Fisher's protected LSD with PDMIX MACRO [39].

Discussion
Water scarcity in semi-arid environments, including the Southwest US, needs to be addressed. Crops such as guar that can grow under limited water conditions can help sustain agriculture in such regions. The current study focused on investigating the response of selected guar genotypes to various irrigation regimes resulting in water stress at different stages during the reproductive phase with and without an amendment of biogenic silica.
Most of the growth and yield parameters of guar were affected by irrigation regimes, with slightly varying results in different years. In general, growth parameters such as plant height and above-ground dry biomass at maturity were higher under the normal irrigation (I 1 ) regime, where no stress was experienced by guar plants, as compared to the reduced irrigation regimes (I 2 , I 3 , and I 4 ) ( Table 1). These results are in accordance with previous studies that reported similar results [19,40]. It is interesting to note that although the irrigation regimes differed in imposing water stress during the reproductive stage, the water stress impacted plant height and above-ground biomass. This is probably due to the indeterminate growth type of guar plants, which can grow vegetatively even after the reproductive stage has set in [14].
Application of biogenic silica in the regime with early termination of irrigation (I 5 ) resulted in higher plant height and above-ground biomass as compared to the regime with early termination of irrigation that received no biogenic silica (I 4 ), indicating a positive impact of biogenic silica on plant growth (Table 1). Similar results indicating the benefits of biogenic silica have been reported on other plants as well [26,27] Physiological parameters such as SPAD chlorophyll, photosynthetic rate, stomatal conductance, and transpiration rate of guar showed varying results among different irrigation regimes when measured at different stages of plant growth during the study. In general, SPAD chlorophyll values were higher under the normal irrigation (I 1 ) regime, where no stress was experienced by guar plants, as compared to the reduced irrigation regimes (I 2 , I 3 , and I 4 ) ( Table 1). The early termination irrigation regime receiving the application of biogenic silica (I 5 ) resulted in higher SPAD values than when no biogenic silica was applied to the early termination irrigation regime (I 4 ). Moreover, SPAD chlorophyll values under I 5 were either higher or similar to the I 1 regime, indicating beneficial effects of biogenic silica under water stress conditions. Previous studies suggested that the application of silicon to cucumbers under drought helped increase levels of chlorophyll [41]. Similarly, other studies have reported positive impacts of biogenic silica application on physiological parameters such as osmotic adjustment [30], interception of light and leaf water potential [26], and photosynthetic activities [42].
Intrinsic and instantaneous water use efficiency indicate the photosynthetic efficiency of water use in relation to stomatal conductance and transpiration rate, respectively. There were either no differences (intrinsic) or slight differences (instantaneous) in water use efficiency among varied irrigation regimes, indicating that guar plants were able to adjust stomatal conductance and transpiration rates under water stress and, therefore, were able to withstand water stress or drought (Tables 2 and 3).
The agronomic water use efficiency (WUE y ), which is a measure of seed yield produced per unit of water used, however, was higher under I 5 compared to all other irrigation regimes in 2016 and as compared to I 4 in 2018, indicating the positive impact of biogenic silica on guar growth under water stress ( Table 3). Application of silicon was also reported to increase the amount of water uptake and antioxidant enzymes in wheat [43].
The seed yield-attributing characteristics contribute to the final seed yields and therefore can impact the overall productivity of a crop. In the current study, the seed yield attributing characteristics, including clusters per plant, pods per plant, seeds per pod, and seeds per plant, were recorded under different irrigation regimes. Although differences were not always found to be significant among irrigation regimes, I 1 and I 5 generally had higher numerical values of these attributing characteristics than other irrigation regimes (Table 4). Previous studies have also reported pods per plant to be an important seed yield-attributing characteristic of guar under water stress conditions [44]. The I 5 regime had a higher number of clusters per plant, pods per plant, and total number of seeds per plant than the I 4 irrigation regime, showing the positive impact of biogenic silica on these yield-attributing characteristics.
The results showed that among irrigation regimes, in general, higher levels of guar seed yield were obtained under I 1 (Table 5) as compared to I 2 , I 3 , and I 4 , which received an annual average of 552.4 mm of irrigation plus 146 mm of precipitation from 2016 to 2018 (Figure 1). Results were mixed for the I 1 -I 5 comparison, with I 5 having a numerically higher but not significantly different seed yield than I 1 in 2016 but a lower seed yield than I 1 in 2018. In southwestern New Mexico, the rain will most likely begin after the flowering stage if it is planted in mid-June for optimum guar development [12]. Earlier studies also showed that guar can achieve its optimum growth in non-irrigated areas with an annual rainfall ranging from 254 to 1016 mm [3], which is less or comparable to the total water needed per season for alfalfa (800-1600 mm), cotton (700-1300 mm), and potato (500-700 mm) [30]. While comparing the data from different years, guar seed yields obtained in 2018 were lower than in 2016 and 2017 ( Figure 1). This was even though the total amount of water applied through irrigation was 160% greater in 2018 than in 2016. There was some rain received along with irrigation at emergence, but no rain or irrigation was received during the unifoliate in 2018. This relatively longer dry spell during the unifoliate stage could have affected the overall development of guar plants in 2018. Additionally, guar plants experienced a longer spell of dryness and heat, resulting in wilting symptoms at the 50% pod-formation stage, due to which extra irrigation at the 50% pod-formation stage was applied to all treatments in 2018. This was a key observation, indicating the importance of water availability at the pod-formation stage. This is in agreement with previous research indicating irrigation at mid-pod filling is critical for guar seed production [20].
The distribution of precipitation also seemed to play an important role in the growth and development of guar plants in the current study. For instance, although the total amount of rainfall received during the guar growing season was similar in 2016 and 2018, a lower guar seed yield was obtained even when the amount of irrigation was higher in 2018 than in 2016. This could potentially be due to the distribution of rainfall and the availability of sufficient amounts of water at regular intervals. A previous study conducted in Iran also showed that regular irrigation every three days was most efficient for guar seed production, indicating the importance of the distribution of available water through the growing season [45].
When averaged from 2016 to 2018, I2 resulted in a 20.8% decline in seed yield as compared to I1 (Figure 1) Figure 1. Amount of total water received (Rain + irrigation) vs. guar seed yield from 2016-2018 in Las Cruces, NM. I 1 -normal irrigation, I 2 -no irrigation at 75% pod formation, I 3 -no irrigation at 50% and 75% pod formation, I 4 -terminate irrigation at flowering, and I 5 -terminate irrigation at flowering + biogenic silica. The 50% and 75% pod formation stages at 57 and 72 days after planting, respectively.
While comparing the data from different years, guar seed yields obtained in 2018 were lower than in 2016 and 2017 ( Figure 1). This was even though the total amount of water applied through irrigation was 160% greater in 2018 than in 2016. There was some rain received along with irrigation at emergence, but no rain or irrigation was received during the unifoliate in 2018. This relatively longer dry spell during the unifoliate stage could have affected the overall development of guar plants in 2018. Additionally, guar plants experienced a longer spell of dryness and heat, resulting in wilting symptoms at the 50% pod-formation stage, due to which extra irrigation at the 50% pod-formation stage was applied to all treatments in 2018. This was a key observation, indicating the importance of water availability at the pod-formation stage. This is in agreement with previous research indicating irrigation at mid-pod filling is critical for guar seed production [20].
The distribution of precipitation also seemed to play an important role in the growth and development of guar plants in the current study. For instance, although the total amount of rainfall received during the guar growing season was similar in 2016 and 2018, a lower guar seed yield was obtained even when the amount of irrigation was higher in 2018 than in 2016. This could potentially be due to the distribution of rainfall and the availability of sufficient amounts of water at regular intervals. A previous study conducted in Iran also showed that regular irrigation every three days was most efficient for guar seed production, indicating the importance of the distribution of available water through the growing season [45].
When averaged from 2016 to 2018, I 2 resulted in a 20.8% decline in seed yield as compared to I 1 (Figure 1). Similarly, I 3 resulted in an average decline of 23.4% in guar seed yield in Las Cruces from 2016 to 2018. The lowest seed yield was obtained under I 4 (Figure 1), which resulted in a 26.4% decline in seed yield as compared to I 1 based on average yield.
It is interesting to note that I 5 resulted in only a 9.1% decline in seed yield as compared to the highest yielding I 1 , when averaged from 2016-2018. When compared to I 4 , the addition of biogenic silica in I 5 resulted in a significant increase in the seed yield and yield-attributing characteristics, including clusters per plant, pods per plant, and seeds per plant, in 2016 (Table 4). Similar trends were observed in 2017 and 2018, with numerically (but not significantly) higher seed yields under I 5 than I 4 ( Figure 1) ( Table 5). The positive impact of biogenic silica in improving the guar seed yield under I 5 as compared to I 4 could be attributed to its role in improving processes including metabolism and altering physiological activities such as osmotic adjustment [30], interception of light and leaf water potential [26], photosynthetic activities [42], and mineral uptake [46].
For the years in this study, it appears that if irrigation is terminated after flowering, adding biogenic silica may completely mitigate the effects of the reduced irrigation in some years but not in others. The results obtained from the current research using drip irrigation showed that guar can grow well in the region with an average of 300 mm of irrigation during the growing season.

Conclusions
This study recorded that reasonable guar seed production can be obtained under reduced irrigation conditions after flowering in southern New Mexico. On average, the I 1 irrigation regime produced the highest guar seed yield (2715 kg ha −1 ), followed by I 5 (2469 kg ha −1 ) from 2016 to 2018. As compared to the I 1 regime, the I 2 and I 3 regimes resulted in a 20.8% and 23.4% decline in guar seed yield, respectively, on average from 2016 to 2018. Among genotypes, Kinman and Matador seemed to perform slightly better with higher growth and yield parameters than other genotypes; the seed yield differences, however, were not significant.
The results of this study suggest that the addition of biogenic silica might help to improve seed yield production under reduced irrigation conditions, such as early termination at flowering, in arid irrigated agriculture in southern New Mexico. The addition of biogenic silica seemed to help alleviate water stress and have a positive impact on guar seed yield. Under water stress, biogenic silica appeared to improve the chlorophyll content and yield attributing characteristics.  Acknowledgments: Acknowledgements are due to Anthony and the farm crew for assistance with field operations and set-up, and to student aides for their help in managing the study and collecting samples and data.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.