2005 was always going to be a difficult act to follow. With all the publicity that nowadays surrounds the top Bordeaux vintages, sometimes excellent succeeding vintages don’t stand a chance at the start: ’86, ’90, ’96, ’01, ’04. All of these seem to have started life dwarfed by their illustrious predecessors, then gained in reputation with time, and the first three, by many accounts, ended up even surpassing them.
Nobody in Bordeaux is saying that the ‘06s will surpass the ‘05s, but, until well into the vineyard year, it was a distinct possibility. One of the coldest winters of the last 25 years resulted in a late healthy budding; then a warm, dry, almost frost-free spring quickly caught up all the lateness and accounted for a regular, efficient, on-time flowering; an almost brutally hot, dry 95-like June and July gave the naturally spaced-out bunches the initial concentration that would certainly account for the high sugars we would see at the end.
We were well on our way to another great vintage. Then the year started falling to pieces. A cool, damp August swelled and weakened the grapes, so that, when the searing heat-wave returned for the first 10 days of September, whilst the dry whites and some earlier reds could be picked in an ideal phase of re-concentration, for the rest, this combination of extreme heat after the dampness gave rise to an acute rot-risk.
All would have been well again, as in ’02 and ’04, if the weather had stayed good for the rest of the harvest. But when the rains arrived mid-September, picking decisions had to be made quickly, especially for the main body of the Merlots and Cabernet Francs, and many missed the final touches to the ripening process that we had so enjoyed the prvious year during the leisurely halcyon September days of 2005, whilst for others, the ripening was induced by the rain rather than the sun and that brought with it a slight pinch to the tannins. Those Cabernet Sauvignons that safely got through this rainy period could take advantage of the very fine final week of September that re-concentrated them for the second time – as in ’86 and ’96 – so that they could mostly be harvested at full ripeness in early October.
So, we may have missed making another ’05 but in the end some very good wines have been made, the reds, especially those Merlots that were ripe when the rot-risk arrived and those Cabernets that could wait until October, fine yet sturdy, gentle but high in alcohol, and displaying very early on a tensile kind of minerality which presumably will become the fruit character of the vintage.
Of course, until August, the vintage was looking far, far superior to ’02 and ’04, and all the benefit from such a fabulous early summer could not just be wiped out by a few problems at the end. Without it, we would have had a weak vintage rather than a weakened strong one.
Winter 2005/2006
The last two vintages have shown the same double phenomenon of unusually cold winters and very hot summers. The return of El Niño doubtless has something to do with it: 1998, the last El Niño year, bears many similarities to 2005 and ‘06: the prolonged winter frosts and the long summer heat-wave. Global warming has something to do with it too: the French Meteorological Office reckons that it will cause it to become a new trend, especially so in the south-west of France. Whatever, it is quite clear that the mid-Atlantic high pressure systems have shifted further north over the past two years, bringing in Arctic air from the north-east in the winter, and sucking in hot easterly air originating from Africa in the summer. And anyway, if this does indeed become a trend, it can only be good for Bordeaux!
The average temperature was below average for every winter month except March, especially in December when it was a full 3°C less than the 85-year average for the month. We didn’t have a white Christmas but it never unfroze over the whole Christmas period (the Sauternes for Chrsitmas pudding did fine on the doorstep rather than in the fridge) and we did have a white 29th December. Altogether there were 44 days of frost, compared with the average 31, most of them, as in ’85 and ’87, were over long periods, especially 18th – 29th December and 23rd January – 6th February. However, unlike ’85 and ’87, the frosts were not nearly as deep, only getting to a minimum of -5°C twice. This meant that pruning could go on pretty much all through the winter, stopping only on days when the canes got too brittle. The quality of the wood was said to be excellent anyway.
There was never a danger of the sap rising before the second half of March, thus avoiding all the dangers of an early start to the vegetation. It also helped to kill off all the pests that, during more humid and warmer winters, lie latent, ready to attack the first foliage.
Anyway, the main cause for concern was not about the temperature, it was about rainfall. Bordeaux had been suffering drought conditions since spring 2003, exacerbated by the record heat of 2003 and the record lack of rainfall of 2005. The winter of 2005/6 was a little wetter than the previous four, with rainfall just over the average for November, December and February, but seriously deficient for the dry, cold February. This was better than the two previous winters but not enough to replenish the water-tables.
If it didn’t rain before the vines started up, there would clearly be problems of nourishment at this critical time and over the summer. Longer term, what if, in addition to the new cycle of colder winters and hotter summers, we were also to go back to the mid-80s long cycle of extreme drought? This in fact looks to be already the case: ’85 -’90 was a drought period, ’91 – ’02 a wet period, and ’03 – ’06 has so far been a drought period. A suivre…
Spring, budding and early growth
At least for the immediate future, March rainfall was to allay these fears. It rained heavily and persistently over the first ten days of the month. There followed a week of cool nights, even several light frosts, that kept the swellings at bay. But when it rained again 18th – 24th, followed by a burst of afternoon temperatures in the mid-twenties, all the whites and earlier Merlots swelled immediately, with the rest close on their heels.
However, the actual bud-burst was delayed and spun-out by the cold night-time temperatures of April. Some had burst around 1st-2nd, others not until 18th or so. By now, the earlier parts of the vineyard (Pessac-Léognan, Pomerol) were up to 3 or 4 leaves, looking like an impressionist painting of fields of little flecks of light green, whilst others (N Médoc, Barsac) still looked like a dark brown winter-scape.
Late budding need not mean a lower quality vintage. Other late buddings were ’96, ’00 and ’04. However, dramatically early-budding vintages like ’03, ’89 or ’90 all seem to have needed to bud early in order to ensure the very early harvest that was the prerequisite for that warm style of fat, forward wines. So the implication now was that we were unlikely to be going towards that kind of vintage, more likely a traditional style.
It is at this time that the embryo bunches can be approximately counted, and it quickly became clear that the “sortie” was not going to be very big, just like in ’02 and ’03. Vignerons say that the vine makes its mind up about its next year’s bunches in the August of the previous year, and in August 2005, they said that the vine was thinking about its own survival against the drought, not about its reproduction for the following year. Hence the low “sortie”.
During this laborious budding period, there was a localised frost on the morning of 11th April. The thermometer went down to -1°5 in parts of the Médoc and Right Bank and even to -4°5 in parts of the Graves. At first, nothing much was thought of it. The buds seemed to be not sufficiently burst to be that fragile. But later, it was noticed that certain parcels had some shrivelled embryo clusters, especially in N Médoc, S Graves, and parts of the Right Bank. It was most unusual to note that the freezing had occurred in little channels running over the higher points of the vineyard, not in the more usual low-lying bits by the woods.
It was not a very significant frost, but was just one more factor towards the naturally low yield of 2006.
Early summer and flowering
May began and continued regularly quite warm. With the vines nicely watered by heavy showers on 8th and 9th, the growth spurted ahead, with good, firm shoots, bright vigorous leaves and long, aerated bunches. Now that the afternoon temperatures were well into the 20s, all it needed for the flowering to start was the storms of 18th – 20th and, after a little pause during the cool nights of 24th – 25th, it started with a bang, accelerating as the temperature increased into a heatwave. It was quickly over in the 30°+ totally fine weather up to 15th June.
There was some “coulure” caused by the cold end-of-May nights on the early Merlots and whites, estimated at around 20%, but in the end was not cried over too much as it seemed to aerate the bunches even more and to be a further contribution to the naturally low yield of the year, making costly crop-thinnings even less necessary. Anyway, it was now known that the permitted maximum yields would continue as low as last year, so why complain about an economy of labour if you can get there naturally?
The lateness of the budding had been neatly caught up, with less than 2 months between it and the flowering. The mid-flowering was variable from region to region but generally was situated at about 7th-8th June – (the average is 14th June, the earliest (’90) end May, and the latest (’91) early July). So we were now back on time and plans could be made for the red harvest to start around 11th – 15th September.
But first, we had to get through what seemed to be becoming another arid, hot summer. April had produced no more than four showers, for a total of 27 mm (the average is 80), then May, with its seven showers totalled 47 mm (the average is 84) and June, with its six thunderstorms ranged from totally dry in the Médoc to only half the normal in Sauternes. By now, the spring rainfall had been absorbed – this is the time of year when the foliage expansion of all plants and trees, not just vines, makes tremendous demands on the soil’s moisture. The vines looked great, but the grass by the side of the roads was going yellow like savannah.
So, while the US East Coast was battling Agnes’s flooding, the French West Coast was bracing for another arid summer. The second half of June remained resolutely dry, apart from five very localised storms of 16th,17th, 19th, 24th and 26th which only really brought any water to Graves, Sauternes and the Entre-Deux-Mers, the rest of the region, especially the Médoc, remaining totally dry. And temperatures, although slightly less than the first half, remained resolutely in the upper 20s.
July and August
In extreme years, such persistent heat often generates violent hailstorms – remember the storms of Whitsun ’85 and Vinexpo ’03, both in similar furnace conditions? The surface temperature of the Gulf of Gascogne heats up (in July 2006 to 23°C off Bordeaux, 1°5 more than usual); this in turn heats the air over the sea; this air reaches the shore and heats the already warm ground even more, which creates hotter air etc etc, until a migrant depression arrives on all this and creates violent convection.
But strangely, although there were seven isolated thunderstorms during the month, none of them carried any hail, so for some reason this violent convection never materialised. No doubt we will find out why.
So the main feature of July was just pure heat and drought. The month’s mean temperature, at 25.2°C, was a massive 4.4°C over the 85-year average. This put it as the hottest July since 1950 and the second-hottest single month, after August ’03, since records began. The heatwave started right from 1st July with a two-day blast of 35°C. It then “cooled” to the high 20s for seven days, after which it resumed for 19 consecutive days, the longest heatwave on record (August ’03 had been “only” 13 days, ’76: 15 days). The July single day record of 38.8°C in 1990 (when four people died of heat exposure while watching the tall ships coming in) was almost smashed on 21st (37.7°C).
We have experienced very hot Julys before (’76, ’83, ’88, ’90, ’95, ’03, ’05) but each time the heat has continued into August….and unfortunately not always into September. So we have very little comparison with the unusual ’06 combination of heatwave in July and coolness in August. 1996 is the closest recent comparison. In that similarly yo-yo year, there was a heatwave from 10th July to 10th August, followed by a cooler rest of August. But this cool remainder of August ’96 was much finer than the cloudy, damp coolness we were to have in August ’06, and the harvest was consequently much less fragile going into September.
The whole of July had seen a broad very high pressure system (often up to 1032 mb) permanently centred over England (There it created unusually (!) hot conditions as well as in northern Europe where records were broken at 36°C on 19th). Against this defence, the attacking Atlantic depressions didn’t stand a chance for the whole month.
Then, as from 1st August, the entire situation changed. The defence collapsed, the high pressure subsided, allowing a whole series of weak depressions to come in over Europe. They were so weak that they brought with them all sorts of little fronts, not strong enough to produce much rain, just a series of dull, drizzly, cool days. And that was the way the whole of August would turn out to be.
It felt colder than it was: in fact August finished only 1°6C below the average. It was just like in July ’98, when similar gloomy conditions seemed worse than they were. And it felt wetter than it was: the rainfall figures were finally not exceptional: 72 mm at Mérignac, compared to an average of 60. There were only three storms with hard local rainfall. The rest was just a drab, dull, drizzly month, as expressed in the sun hours: 225 against the average 242.
The effect on the bunches was very obvious. The grapes swelled from the tiny hard berries they had been at the end of July, especially after the locally heavy rain of 27th – 29th August. If the wind hadn’t got up when the sunny days returned 30th – 31st, there would have been rot problems right then. From now on, rot was to be never far away. After a relatively disease-free year, this was a new problem. Mildew had been present in May but the Summer of 2003 had pretty much scotched the big mildew presence of ’99,’00, ’01 and ’02 and the attacks didn’t return. There had been a few incidences of oidium end April, a bit of pressure after the early June rain, and some small problems in August but nothing really serious. There had been some “ver de la grappe” mid-June which had generally been caught by the sprays. So, up until now, it had not been a disease year. However, it would now become very risky, especially for those who had deemed it unnecessary – and uneconomical – to spray during the dry days of July.
So, finally, what conclusion can we draw from this unusual combination of July heat and August cool? We have had dull July’s before: ’98 and ’04 for example, both of which would have made more luxuriously ripe wines if it had been hot instead, but never such a dull August after such a searingly hot July. We have no recent references here.
What is the effect of early heat? Before the grapes have changed colour, they are not yet able to build up sugar in the same way that they would in a hot August, and some even say that only the skin-tannins can be transformed at this stage. Many were heard to say that this July heat was therefore a bit wasted. But how do we explain the extremely high sugar levels at harvest time, which cannot have come from their slowed progress in August or merely from the 10-day early September heatwave? One possible answer was explained to me by a thoughtful grower this way: Pre-véraison heat is indeed too early to concentrate the grapes directly; however, at this stage it is the vine’s trunk that stores the sugars, which are then transmitted to the bunches at the “aoûtement” in the second half of August. Seems plausible enough.
The véraison started during the dog days of the second half of July, and then proceeded to get blocked by the heat. So these earlier bunches had to wait to finish changing colour until the rest started in the cooler first days of August. This created (1) some inconsistency which was to stay for the rest of the year and which, at harvest time, would account for the presence of slightly less ripe fruit in amongst the ripe bunches and (2) two or three days of lateness in the final vineyard cycle that accounted at this time for putting back harvesting plans by half a week or so.
What is certain is that, with all that July sunshine, the by now common practice of summer de-leafing got put off from June – July to August. With July’s longer days, the sun would have burnt the skins. There was also a second reason, that usually the de-leafing job is done at the same time as the crop-thinning, which, with a naturally medium-sized expected yield, was not a priority this year: no crop-thinning, no de-leafing. However, as August wore on ever duller and damper, with the grapes loosening and the foliage re-starting (at one stage, there was concern that this may turn into a ’97 type situation, when the vine was more interested in producing leaves than grapes), everyone was out in the vineyards, snipping and culling. There were so many people amongst the rows, it looked like harvest time.
In the space of a month, the mood had gone from making another ’05 to saving the vintage. It was now essential to get some good dry weather.
September–October and the harvest
Just like in ’96, the change of month brought a dramatic change of weather conditions, with our old ally the Azores high pressure system ballooning out over Europe and bringing warm air into the region first from the south, then the east. Temperatures immediately spiralled from the low 20s to the 30s, creating the hottest series of September days since the scorcher year 1921, with the afternoons a whole 7.4°C over normal.
In ’96, the August nights had been cooler and the rot-risk minimal, and in ’21 it is said the place was as dry as a desert. This year, August had taken its toll and the grapes had been weakened, even deteriorated by rot in certain less professionally handled generic vineyards. So the effect of this sudden heat, although essential for re-concentrating the grapes, was not always beneficial. Unlike the dry July heat, these September nights were warm and damp, with morning fog and dew, so, although in most serious estates, the rot was kept at bay, it was never far away.
After all, it was not for nothing that this was the beginning of the best cèpe (boletus mushroom) season ever. There was also some wrinkling of the skins as the grapes re-concentrated.
Some early Pessac-Léognan whites had started their harvest at the end of August. Most started now in these first fabulous days of September and most had finished by the time the weather changed on 11th. This dry white harvest came in regularly ripe, with little rot, good sugar levels and nicely refreshing acidities. It was felt this would be a banner year for whites. Sauternes, on the other hand, as in 1997, was battling to eliminate the “aigre” bad rot, grapes and even bunches that had gone beery after the wine-flies, which love such hot damp conditions, had punctured the skins and sent the juice volatile. At the same time, many good botrytis grapes were harvested, but only once the vineyard was clean of the “aigre” could they get on with a normal “trie”.
Now it was time for the red harvest to start. It had been advanced at the flowering, then put back when things slowed down in August, and now, nicely re-concentrated but susceptible to rot, it was to be advanced again. The early-ripened estates, especially in Pomerol and Graves, harvested their Merlots the week of 11th, most of the St-Emilion and Médoc Merlots the following week of 18th.
Alas, the high pressure system chose this critical time to subside and allow a series of storms and fronts to come through, and 11th to 25th accounted for all of Mérignac’s 98mm rainfall, of the Médoc’s 100-120 mm and the Right Bank and Sauternes’ 150 mm. This was just what we didn’t need. Up until this time, the effect of August could well have been compensated by a fine September and the vintage could still have been a great one. But this rain was just too much, ripening the Merlots without concentrating them and pinching their tannins. It also obliged many estates to harvest Merlots that were just short of optimum phenolic ripeness. Bordeaux has been so blessed recently with September weather that it would not have been decent to wish for just one week of sun, but it was that one extra week that would have set everything right again. We have had rainier Septembers than this: 275 mm in ’93, 183 mm in ’98, 175 mm in ’94, but these 98 mm did more to compromise the vintage than those three.
In fact, there were only six main showers during this 14-day period from 11th to 24th, and some of them were only heavy locally, with generally more inland than in Médoc and Graves. They also fell mostly at night, allowing picking to continue in relatively good conditions during the day-time. Some of the Merlots continued to be picked well into the improving weather of the month’s end, and some of the Right Bank habitual late-pickers, by dint of their exceptionally manicured vineyards, managed to push through even to early October. But generally, it was a race against time to get the Merlot harvest in before it rotted, followed quickly by some rapidly deteriorating Cabernet Francs around 25th.
Strangely, it was not the deeper gravel soils that escaped the rot risk the most. The heavier clay soils seemed to do better, a real exception to the rule. The water seems to have gone straight into the deeper root system of these more filtering soils and expanded the grapes more than on the shallower systems. And, as usual in rainy Septembers, those that had ploughed early or made their vines too vigorous with too much nitrogen fertiliser got more water absorption and splitting grapes.
Now it was time to think about the Cabernet Sauvignons, which, in line with their sturdy reputation in difficult autumns, had generally resisted all these problems quite well. Of course, they had swelled in August, but had nicely re-concentrated early September. Now they had swelled again but were not really showing signs of wanting to rot. Most had planned to start on Monday 2nd October but several didn’t want to take the risk and considered they were ripe enough anyway, and started in the warm sunshine of the last week of September.
It was during this week also that Sauternes enjoyed its best botrytis, some still on their first “trie”, some on their second, others on both at once.
The first week of October continued fair and relatively warm (20 – 25°C) with night-time temperatures down in the low teens as a safeguard, so the main body of the Cabs could come in at a comparatively leisurely pace during the whole week. There was one heavy storm on the night of Tuesday 3rd which we could have done without, but the effect did not seem too bad with the grapes now at optimal ripeness, especially North of Pauillac where it stayed dry. In other parts, where the harvest was over, the only effect was to knock out the electricity supply, so fermentations had to be done the old way for a day or two.
Most Sauternais had finished their harvest by now, with a very satisfactory third and sometimes fourth “trie”, whilst others continued through the adequately fine second week of October to produce decent but not the best musts of the vintage. October ended up setting a record for warmth, at 17.9°C, beating the previous record-holder 1995 by 0.1°C. If the final week of October, permanently and uselessly sunny and warm (20-25°C), had come in place of the middle week of September, we would have had a stunning vintage. It was not to be. Bordeaux had had a lot of luck with its Septembers in the 2000s and one day that luck had to run out.