I have little knowledge on time series or ananlysis of economic data. But I am interested to do (all kinds of) plots in R. I find it quite rare that one given plot cannot be implemented in R. There must be a reason that the author of the packages make it impossible to add another axis in the plot yet I am not able to give any hint.
With all that said, if you do want to do the plot in your way you may need to construct it piece by piece all by YOURSELF. The following code will produce something that is quite similar to what you will get from plot.xts (at least to me). There is no time series structure involved as I am not familiar with it. BTW, the "Canada" data is provided with vars package not with R base installation.
Here I have mapped all the data into the range [0, 6] and the quarterly time to [0, 83]. The result follows:



I am sure there are simpler/better solutions. This just serves as an exmaple showing the super flexibility of R graphic.

I have not tried Rsweave but would probably expect little change in the codes. As far as I am aware of there is no limit on number of packages loaded in older version of R.

Here is my session info:
Good luck.

表示顶一下~如果能够帮助解答的可以再追加200论坛币~真的是急求~

Actually what you've posted is virtually the same with the one shown on the stackoverflow or other forums which I've seen before.

I knew the methodology and would take a try. But the major worry is that it is so delicate and arduous to calculate the distance in order to draw the scale of x or y axis. If you have an irregular time series data, it is really difficult and time-exhausted to do that. So that's the reason why I want to find other solutions beyond that and why I choose specific time series packages like xts to draw the time series plot....

不知道这么久过去了楼主解决了这个问题没有。
只要在plot中加上ylim=c(a,b)就可以。其中a b 是你想要的上下限。