April 2013 Issue of Wines & Vines

Enological Research in the East

Scientists find yeast-assimilable nitrogen differs between vinifera and hybrid cultivars

by Andrew G. Reynolds
Several universities in the eastern and Midwestern United States and Canada have active and productive enology programs. By far the largest is Cornell University, which has Food Science, Horticulture and Plant Pathology/Entomology departments on two campuses (in Geneva and Ithaca, N.Y.) in addition to research facilities in western New York state, the Hudson Valley and Long Island. Other noteworthy research programs include those at Pennsylvania State University, Virginia Polytechnic Institute and State University, Ohio State University, Michigan State University, Purdue University and Brock University in Ontario, Canada.

Many eastern programs are carrying out enology research, specifically Dr. Gavin Sacks and his colleagues in Geneva, who are studying odor-active compounds; additionally, two programs (Geneva and Purdue) are looking into yeast-assimilable nitrogen (YAN) concentrations in grapes throughout the East. The descriptions below are based upon edited authors’ abstracts from meetings of ASEV-Eastern Section.

Odor-active compounds
Characterization of odor-active compounds in grapes and wines produced from non-Vitis vinifera species important to grape breeding. Many jurisdictions in the East are almost exclusively growing Vitis vinifera, but New York and some other areas depend on hybrid varieties for wine production as well. The breeding program in Geneva has a long tradition in the introduction of high-quality hybrid grapes such as Cayuga White (1972), Chardonel (1990), Traminette (1996) and most recently Corot Noir and Noiret (2006).

Under the direction of Dr. Gavin Sacks, the enology program at Geneva has concentrated on odor-active compounds in non-vinifera cultivars and species. Qun Sun et al. (ASEV-ES 2011) focused upon non-vinifera species and produced wines from V. riparia, cinerea and vinifera under identical conditions.

Volatiles were extracted from wines by solid-phase-microextraction, and 40 potent odorants determined by quantitative gas chromatography olfactometry (GC/O), with the majority subsequently identified by GC-mass spectrometry (GC-MS). The key odorants previously associated with foxy aromas in V. labrusca grapes, methyl anthranilate and 2-aminoacetophenone were not detected in the V. riparia or cinerea wines by either GC/O or GC-MS. Most odorants (24 of 40) were fermentation-derived (i.e., fatty acid ethyl esters) and did not differ by more than a factor of two in flavor dilution (FD) value. Several grape-derived odorants with vegetative and earthy aromas were at higher FD (> 4 factor of difference) in the non-vinifera wines: eugenol, cis-3-hexenol, 1,8-cineole, 3-isobutyl-2-methoxypyrazine (IBMP) and 3-isopropyl-2-methoxypyrazine (IPMP). These latter two compounds are common in Cabernet Franc, Cabernet Sauvignon and Sauvignon Blanc, and are responsible for their green bean and bell pepper characteristics. Quantitative GC-MS confirmed that these compounds were at higher concentrations in non-vinifera wines.

In 2010, 3-isobutyl-2-methoxypyrazine (IBMP) and 3-isopropyl-2-methoxypyrazine (IPMP) concentrations were measured in accessions of V. riparia, rupestris and cinerea. IBMP concentrations in V. riparia and cinerea were generally above concentrations found in mature V. vinifera, with nine of 14 and six of 10 accessions, respectively, containing >50 pg/g IBMP at harvest (median sugar level: 20.4° Brix), which is considered high. For example, IBMP in 2010 Ontario Cabernet Franc wines ranged from 30 to 76 pg/g, while IPMP ranged from 17 to 36 pg/g (Susanne Kögel, pers. comm.). One V. cinerea accession harvested at 20o Brix contained 353 pg/g IBMP. IPMP was also detectable (>1 pg/g) in seven of 10 V. cinerea accessions, and IBMP and IPMP were modestly correlated across all accessions (r = 0.55, p < 0.05).

Fruit-zone light response curves for sensory compounds in Riesling. We generally think of most aroma compounds as being products of enzymatic reactions and therefore dependent upon berry temperature. However, we also know that compounds such as methoxypyrazines can break down under intense sunlight. This, of course, has implications for canopy management, although optimal cultural influences with respect to cluster exposure timing and intensity have not been established. The work of J.M. Meyers et al. (ASEV-ES 2011) showed that some odor-active compounds in Riesling responded to fruit-zone cluster exposure.

To explore the spatio-temporal relationships between fruit-zone cluster exposure and chemical concentrations in Riesling berries, correlations were measured among eight odor-active compounds that are well-known to be important for Riesling aroma: [glycosylated tridihydronaphthalene (TDN), β-damascenone, vitispirane, linalool oxide, α-terpineol, 4-vinylguaiacol, vanillin and eugenol], five cluster exposure metrics of varying spatial precision, two sites and two phenological stages in two consecutive seasons.

At Site A in 2008, TDN (“petrol”) and vitispirane (“flowery”) were positively impacted by exposure at fruit set as measured by two metrics of cluster exposure: cluster exposure layer (CEL) or cluster exposure flux availability (CEFA) (TDN: log CEL, r2 = 0.50; vitispirane: CEFA, r2 = 0.45). TDN and β-damascenone (enhances “fruity” in a matrix) positively correlated with CEFA at veraison (r2 = 0.76 and r2 = 0.53, respectively), but these relationships were not as strong at Site A in 2009 or at Site B in both years where fruit was more heavily shaded (average CEFA <0.2). Eugenol (“smoke” or “spice”) negatively correlated with veraison cluster exposure (log CEL, r2 = 0.44), even at low light levels.

Relationship of IBMP with 3-isobutyl-2-hydroxypyrazine and removal of IBMP from musts using nonpolar sorbents. Dr. Sacks’ program has also placed focus upon ways of reducing methoxypyrazines in wines, with the ultimate goal of lowering vegetal aromas in cultivars such as Cabernet Franc. Controlling final concentrations of the herbaceous-smelling IBMP is of interest to the wine industry not just in New York but elsewhere in the East. Some of this work has focused on canopy management (e.g., basal leaf removal), while other work has taken place on the enological side—and two experiments were performed toward this goal (Sarah Harris et al.; ASEV-ES 2012).

In one experiment, the behavior of a likely IBMP precursor, 3-isobutyl-2-hydroxypyrazine (IBHP), was characterized during the growing season. Following development of an improved method for IBHP measurement, IBHP and IBMP were quantified from fruit set to harvest at one site in California’s Central Valley and two sites in the Finger Lakes region of New York. IBHP was detectable at the earliest sampling point (four weeks pre-veraison), increased to up to >800 ng/L and started to decrease about two weeks after IBMP decrease began. The highest IBMP was observed at the site with the highest IBHP, suggesting that IBMP accumulation is dependent on IBHP formation.

In a second experiment, they investigated the effects of treating must with a nonpolar sorbent (silicone) prior to fermentation. Silicone is a polymer that contains silicon. Food-grade silicone formulations are sold by home-winemaking companies as anti-foaming agents, while aluminosilicates and silicon dioxide are both permitted under North American and EU regulations for de-foaming and clarification. To the best of my knowledge, silicone is not a permitted additive in wines.

Treatment of four different musts with silicone pre-fermentation resulted in noteworthy decreases of methoxypyrazines in the final wine (53% to 93%), without affecting the majority of other wine volatiles. This has major implications for producers of varieties such as Cabernet Franc—not just in eastern North America but worldwide.

Yeast nutrition
Assessment of yeast nutrient supplements, residual nitrogen in wine and amino acid profile in hybrid varieties. If most winemakers are like me, they probably add a fixed amount of diammonium phosphate (DAP) to their must every year and cross their fingers. I don’t know many wineries that actually analyze must nitrogen or verify that the amount of DAP added was sufficient. This is the situation addressed in work by Amanda Stewart and Christian Butzke (right) of Purdue University (ASEV-ES 2011).

Yeast-assimilable nitrogen (YAN) is essential for growth and fermentation properties of yeasts. Two main sources of YAN are found in grape juices and musts: α-amino acids and ammonium ions. The Purdue program developed recommendations for YAN based on initial sugar content more than 10 years ago, and these have proven successful in most situations. YAN in juice or must ranges between 40 and 559 mg/L based on one vintage on the West Coast of the United States. Commercial yeast nutrient supplements are available to correct nitrogen deficiencies, and analytical tools are available to measure these deficiencies.

This study assessed the amino nitrogen contributions of 20 complex yeast foods at manufacturer-recommended doses and found them to range from 8 to 24 mg/L. Analysis of 128 commercial wines showed residual YAN ranged from 11 to 586 mg/L, with hybrid cultivars on average showing higher residual YAN than Vitis vinifera. Nitrogen utilization varies greatly among native and commercial yeast strains, and juices from hybrid cultivars in the Midwest and East can have both relatively low sugar and high nitrogen concentrations at harvest, which may contribute to high residual YAN. Residual nitrogen available following alcoholic fermentation can provide a significant nutrient base for spoilage organisms, particularly surface yeasts and Brettanomyces spp.

To further investigate differences in yeast nutrient status between V. vinifera and hybrids, amino acid profiles were determined for several cultivars. Amino acid profiles of several hybrid cultivars were substantially different from the generally reported profile for V. vinifera.

Some follow-up work (Stewart and Butzke, ASEV-ES 2012) examined amino acid profiles and YAN in hybrid winegrapes from the eastern U.S. and showed that there were substantially different amino acid profiles in winegrapes common to the eastern United States. Comparison of profiles from hybrid varieties with V. labrusca parentage versus those of V. labrusca varieties suggested that amino acid profile is heritable. This relationship is also being investigated for V. riparia hybrids and V. rotundifolia (Muscadines). Amino acid profile should be considered when designing yeast nutrients for hybrid and native winegrape applications.

They also surveyed YAN in winegrapes across several Midwestern and southern states, and found a range of 89 to 938 mg/L across one vintage, more than 30 grape varieties and four states. For some varieties, average YAN far exceeded previous recommendations based on initial sugar content (200-, 250-, or 300 mg/L YAN at 21o, 23o—or 25o Brix, respectively). Understanding the differences in amino acid profile and total YAN concentration between hybrids and V. vinifera is essential to developing targeted fermentation management strategies.

Predicting harvest concentration of yeast-assimilable nitrogen in Finger Lakes Riesling. The enology program at Cornell (Geneva) also has focused, among many things, on yeast nutrition. Drs. Anna Katharine Mansfield (above) and Ramón Mira de Orduña are the main individuals investigating this issue. This project was presented at the ASEV-ES meeting in Towson, Md., in 2011 and was conducted by Mark Nisbet et al. of Cornell University’s New York State Agricultural Experiment Station in Geneva. It involved a survey of 77 commercial Riesling vineyards to determine the range and variation in YAN concentration in New York and to develop a method for predicting the final YAN concentrations prior to harvest.

YAN, comprised of ammonia and primary amino nitrogen, is a measure of the nitrogen available for yeast biomass production and is often the limiting nutrient in alcoholic fermentation. A deficiency of YAN in grape musts is common in the New York wine industry, but because YAN analysis is difficult and time consuming, many producers make nutrient additions without determining initial must YAN concentration. Over 80% of the Riesling samples obtained in the survey had YAN concentrations less than 140 mg nitrogen per liter, the minimum concentration recommended for healthy fermentation.

Riesling samples collected from 61 commercial vineyards at three time points throughout the growing season (pre-veraison, two weeks pre-harvest and at harvest) were also analyzed to determine whether early YAN determination could be used to predict harvest concentrations. A linear regression was used to estimate the harvest YAN concentration from measurements taken two weeks before harvest. This approach achieved a cross-validated R2 of 0.86, indicating good predictive power. Predicting the harvest YAN concentration two weeks before harvest may allow winemakers time to send samples to external laboratories and evaluate their supplementation needs prior to harvest. This data was based on one year of sampling; two additional years will be conducted.

Understanding the relationship between fermentation-derived aromas and juice nitrogen composition.
Mark Nisbet et al. (ASEV-ES 2012) presented work related to relationships between wine aroma compounds and yeast nutrition. We have known for a long time that fusel alcohols and their acetate esters are important components of a wine’s sensory profile, and that their final concentrations are a function of the nitrogenous compounds in grape must, such as ammonia and primary amino acids (PAN), known collectively as yeast-assimilable nitrogen (YAN). The YAN content of New York grapes varies widely, often not meeting the minimum concentrations required for efficient fermentation. Low nitrogen is associated with the production of sulfur off-odors, so supplementation is common. Inorganic nitrogen, added as DAP, is the simplest and least expensive. Complex nitrogen sources are costly and consist of hydrolyzed yeast extracts containing PAN.

Fusel alcohols can arise from two different pathways: catabolism of amino acids via the Ehrlich pathway and metabolism of sugar through anabolic pathways. The relative contribution of each pathway is currently unknown. This project aims to elucidate the contribution of volatiles from catabolic and anabolic pathways under variable nitrogen concentrations using GC–isotope ratio MS. The results will be modeled using multivariate statistics (partial least squared regression) to predict the concentration of fusel alcohols and esters based on PAN concentration. The model will be validated in Riesling musts gathered from sites around New York state. This method is new to wine applications and has the potential to allow enhancement of aroma compounds through targeted nutrition. A clearer understanding of this system will allow winemakers to fine-tune the amount and types of nitrogen supplements used, reducing costs and enhancing wine quality.

Andrew G. Reynolds is professor of biological sciences and viticulture at the Cool Climate Oenology and Viticulture Institute at Brock University in St. Catharines, Ontario. He has written the Wine East column “What’s New in Research: Summaries of Current Literature on Grapes and Wine” since 1997.

Print this page   PRINTER-FRIENDLY VERSION   »
E-mail this article   E-MAIL THIS ARTICLE   »
Currently no comments posted for this article.