> Do not worry about the decrease in arable land...
Apply more technology: agroforestry combined with Boston Dynamics-type robots. Look into published material from/by Ernst Götsch, F. H. King, and Masanobu Fukuoka among others on principles. The techniques they worked out and published can be used to rebuild, reclaim, and build de novo arable land. Main criticism of these techniques is they are too labor-intensive. There isn't a law of physics that enforces wheeled farm machinery, so robotics-based farm machinery can assist with capturing scale. It's a new frontier, so no established player wants to pursue it at this time as long as our backs aren't against the wall when it comes to available arable land.
> shortages of drinking water...
Apply more technology: potable water supply is roughly a derivative function of available energy. Push comes to shove, we as a species are going to widespread nuclear power if necessary instead of the leadership dying by dysentery.
> decline in biomass in fisheries...
Apply more technology: mega-scale research and engineering to build fish sanctuaries/breeding grounds in the relative marine deserts of the open oceans. We're still in the relative hunter-gatherer phase of our aquaculture tech. When we have vast farms in the deep open water ocean that grow algae, then plankton, and upwards, raising an entire tower of consumption, instead of catching wild fish to feed domesticated fish, then I'd say we are relatively progressing into true aquaculture. Wouldn't hurt to develop at the same time genetically modified sponges and coral to capture and entomb mercury, and start millennia long filtration while we work out how to stop dumping it into our ecosystem.
> scarce minerals, or any of that stuff. It's all in your head!
Apply more technology: von Neumann mining while at the same time sending out starwisps to scout nearby star systems. Hundreds and even millennia-long timelines here, but not constrained by the known laws of physics, only by our species' own will to overcome our internal dissension.
The question I'm unsure of is will capital allocate to the "more technology" solution space before we as a species run out of time? I'm not entirely convinced capital markets actually understand the timelines, scales, and civilizational risks we are working with here, and even if individual people within the markets understand them, there is an argument to be made that the incentive structures of the markets combined with people's natural lifespans combine to create perverse incentives for funding Long Now-scale tech.
The dominant thesis the markets have embraced thus far is "the tech will become available when the threat becomes clear and present", which is outright faith and not based upon any settled science. Once the timeline to develop and deploy the tech to solve or even mitigate the challenges you outlined above exceed the timeline to catastrophic civilization-wide impacts, the capital markets have failed their anticipation function.